Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils 1119871476, 9781119871477

Citation preview

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils

Edited by Mikhail A. Varfolomeev Kazan Federal University Kazan Russia

Chengdong Yuan Skolkovo Institute of Science and Technology Moscow Russia and Kazan Federal University Kazan Russia

Jorge Ancheyta Instituto Politécnico Nacional Mexico City Mexico and Instituto Mexicano del Petróleo Mexico City Mexico

This edition first published 2023. © 2023 John Wiley & Sons Ltd. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/ permissions. The right of Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta to be identified as the authors of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats. Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Library of Congress Cataloging-in-Publication Data Names: Varfolomeev, Mikhail A., editor. | Yuan, Chengdong, editor. | Ancheyta, Jorge, editor. Title: Catalytic in-situ upgrading of heavy and extra-heavy crude oils / edited by Mikhail A Varfolomeev, Chengdong Yuan, Jorge Ancheyta. Description: Chichester, West Sussex : Wiley, 2023. | Includes bibliographical references. Identifiers: LCCN 2023006627 (print) | LCCN 2023006628 (ebook) | ISBN 9781119871477 (hardback) | ISBN 9781119871484 (adobe pdf) | ISBN 9781119871491 (epub) Subjects: LCSH: Catalytic reforming. | Heavy oil. Classification: LCC TP690.45 .C38 2023 (print) | LCC TP690.45 (ebook) | DDC 665.5/384–dc23/eng/20230403 LC record available at https://lccn.loc.gov/2023006627 LC ebook record available at https://lccn.loc.gov/2023006628 Cover Design: Wiley Cover Images: © Anan Kaewkhammul/Shutterstock Set in 9.5/12.5pt STIXTwoText by Straive, Chennai, India

v

Contents List of Contributors xv About the Editors xix Preface xxi 1

1.1 1.2 1.3 1.3.1 1.3.2 1.3.3 1.4 1.4.1 1.4.2 1.4.3 1.4.4 1.5 1.5.1 1.5.2 1.6 1.6.1 1.6.2 1.7

Properties of Heavy and Extra-Heavy Crude Oils 1 Alexis Tirado, Guillermo Félix, Fernando Trejo, Mikhail A. Varfolomeev, Chengdong Yuan, Danis K. Nurgaliev, Vicente Sámano, and Jorge Ancheyta Introduction 1 Heavy and Extra-Heavy Crude Oils 2 Physical Properties 3 Density, Specific Gravity, and API Gravity 4 Viscosity 4 Pour Point 5 Chemical Properties 5 Elemental Analysis (CHONS) 6 Metal Content 7 Carbon Residue 8 Molecular Weight 9 Composition 9 SARA Analysis 12 TBP Distillation 14 Typical Heavy Crude Oils 19 Properties 19 Relationship Between Properties 27 Concluding Remarks 34 References 34

2

Advanced Characterization of Heavy Crude Oils and their Fractions 39

2.1

EPR Spectroscopy 40 Marat R. Gafurov, Alexander Rodionov, Fadis Murzakhanov, and Georgy Mamin Basic Principles of EPR Spectroscopy for Petroleum Investigation 41 Pulsed EPR Techniques 48 HYSCORE Spectroscopy 51 Pulsed ENDOR 54 References 56

2.1.1 2.1.2 2.1.3 2.1.4

vi

Contents

2.2

2.2.1 2.2.2 2.2.3 2.2.4 2.2.5

2.3

2.4

NMR-Spectroscopy and NMR-Relaxometry 59 Ilfat Z. Rakhmatullin, Sergey V. Efimov, Marsel G. Fazlyyyakhmatov, Ranel I. Galeev, and Vladimir V. Klochkov Phenomenon of Nuclear Magnetic Resonance 60 Technical Aspects of NMR Spectroscopy 62 NMR Spectroscopy in Study of Oil Samples and Their Individual SARA Fractions 65 Technical Aspects of NMR Relaxometry 70 NMR Relaxometry in the Study of Oil Samples, Oil-Saturated Rock Samples, and Their Individual SARA Fractions 71 References 75 FTIR-Spectroscopy 79 Irek I. Mukhamatdinov References 111

Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods 114 Roman S. Borisov, Anton L. Maximov, Anastasiia Yu. Kanateva, and Vladimir G. Zaikin 2.4.1 Introduction 114 2.4.2 Chromatography Methods 115 2.4.2.1 Gas Chromatography 115 2.4.2.1.1 Fingerprinting 116 2.4.2.1.2 Group Analysis and Simulated Distillation 117 2.4.2.1.3 Selective Detection 121 2.4.2.2 Liquid Chromatography 123 2.4.3 Mass Spectrometry Methods 125 2.4.3.1 Ionization Methods 126 2.4.3.1.1 Electron Ionization (EI) 126 2.4.3.1.2 Chemical Ionization (CI) 126 2.4.3.1.3 Atmospheric-Pressure Chemical Ionization (APCI) 126 2.4.3.1.4 Field Ionization (FI) 127 2.4.3.1.5 Photoionization (PI), Atmospheric Pressure Photoionization (APPI), Atmospheric Pressure Photochemical Ionization (APPCI) 128 2.4.3.1.6 Electrospray Ionization (ESI) 128 2.4.3.1.7 Atmospheric Solid Analysis Probe (ASAP) Mass Spectrometry 128 2.4.3.1.8 Laser Desorption/Ionization (LDI), Matrix-Assisted Laser Desorption/Ionization (MALDI), Surface-Activated Laser Desorption/Ionization (SALDI) 129 2.4.3.1.9 Direct Analysis in Real Time (DART) 130 2.4.3.1.10 Other Prospective Desorption Ionization Techniques 130 2.4.3.2 High and Ultrahigh-Resolution Mass Spectrometry (uHTMS) 131 2.4.3.2.1 Fourier-Transform Ion Cyclotron Mass Spectrometry (FT-ICR-MS) 131 2.4.3.2.2 Fourier Transform Orbitrap Mass Spectrometry 131 2.4.3.2.3 Multireflection Time-of-Flight (TOF) Mass Analyzer 132 2.4.3.2.4 Presentation of High-Resolution Mass Spectrometry Data 132 2.4.4 Particular Application of Combined Chromatography-Mass Spectrometry Methods to Analysis of Heavy Oils, Their Fractions, and Petroleum Products 132 2.4.4.1 Capabilities of Gas Chromatography-Mass Spectrometry Techniques in the Analysis of Heavy Oils 132

Contents

2.4.4.1.1 Combined One-Dimensional GC and MS Method (GC-MS) 132 2.4.4.1.2 Combined Two-Dimensional GC and MS (GC × GC-MS) 136 2.4.5 Application of Soft Ionization and Desorption/Ionization Mass Spectrometry to Analyze Heavy Oils, Their Fractions, and Refining Products 138 2.4.6 Conclusion 144 References 145 3

3.1 3.2 3.2.1 3.2.1.1 3.2.1.2 3.2.1.3 3.2.2 3.2.2.1 3.2.2.2 3.2.2.3 3.2.3 3.2.3.1 3.2.3.2 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.4 3.5 3.6 3.6.1 3.6.2 3.6.3 3.7

4

4.1 4.2 4.2.1 4.3 4.3.1 4.3.1.1 4.3.1.2

Applications of Enhanced Oil Recovery Techniques of Heavy Crudes 153 Chengdong Yuan, Mikhail A. Varfolomeev, Mustafa V. Kok, Danis K. Nurgaliev, and Airat H. Gabbasov Introduction of EOR 153 Thermal EOR Methods 154 Steam Injection 154 Cyclic Steam Stimulation (CSS) 154 Steam-Assisted Gravity Drainage (SAGD) 155 Steam Flooding 156 In Situ Combustion (ISC) 156 Traditional ISC 156 Toe-to-Heel Air Injection (THAI) 158 CAtalytic Upgrading PRocess In Situ (THAI–CAPRI) 159 Other Thermal Methods 160 Electromagnetic Heating Methods 160 Electrical Heating Methods 160 Chemical EOR Methods 161 Polymer Flooding 161 Surfactant Flooding 161 Combination Flooding of Surfactant, Alkali, and Polymer 162 Solvent Injection 162 Gas EOR Methods 162 Microbial EOR Methods 163 Hybrid EOR Methods 163 Hybrid Thermal-Solvent Methods 163 Hybrid Thermal-NCGs Methods 164 Hybrid Thermal-Chemical Methods 164 In Situ Upgrading 164 References 165 Fundamentals of In Situ Upgrading 168 Alexey Vakhin, Firdavs Aliev, Galina Kaukova, Ameen A. Al-Muntaser, Muneer A. Suwaid, Chengdong Yuan, Jorge Ancheyta, and Mikhail A. Varfolomeev General Aspects 168 The Initiation of the Hydrothermal Upgrading Process 170 The Water–Gas Shift Reaction 171 The Role of Water (Steam) as Green Hydrogen Donor During Hydrothermal Upgrading 173 Evaluation of Donating Performance 174 FTIR Spectroscopy Measurement of Oil Samples and Their SARA Fractions 174 Isotope Analysis of Oil Samples and Their SARA Fractions 178

vii

viii

Contents

4.3.1.3 4.3.2 4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.4 4.3.3 4.3.4 4.4 4.5 4.6 4.6.1 4.6.2 4.6.3 4.6.3.1 4.6.3.2 4.6.3.3 4.6.3.4 4.6.3.5 4.6.4 4.7 4.8 4.8.1 4.8.2 4.9 4.10

5 5.1 5.2 5.3 5.3.1 5.3.2 5.4 5.4.1 5.4.2 5.5 5.5.1 5.5.1.1 5.5.1.2 5.5.1.3

Elemental Composition of Oil Samples and Their SARA Fractions 179 Evaluation of Upgrading Performance 179 GC Analysis of Evolved Gases 179 Viscosity and Elemental Analysis of Oil 180 SARA Analysis of Oil 181 Distribution of n-Alkanes in the Saturates 181 Evaluation of the Morphological and Structural Changes of Oil-Soluble Catalyst Before and After Catalytic Aquathermolysis 182 Evaluation of the Possibility of Deuterium Exchange Out of Aquathermolysis Window Scope at 120 ∘ C 184 Viscosity and Hydrothermal Upgrading 186 The Role of Minerals as Natural Catalysts 189 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance 199 Material Balance (Products Distribution) and Pressure Changes During the HTU Process 200 Evolved Gas Components Analysis by Gas Chromatography 201 Liquid Products Analysis 202 Viscosity and API Gravity of Oil Before and After HTU 202 Elemental Analysis and Desulfurization of Oil Samples During Thermal Conversion Process 203 FTIR Spectroscopy of Oils Before and After Thermal Conversion 204 Changes in SARA Fractions 205 Analysis of SARA Fractions 206 Analysis of Coke Obtained at 300, 350, and 400 ∘ C using FTIR-Spectroscopy 210 Hydrodesulfurization 211 Evolved Noncondensable Gases 212 CO2 and CO Production 212 Methane and C2+ Generation 215 Field Tests 218 Conclusions and Recommendations 229 References 230 Catalyst for In Situ Upgrading of Heavy Oils 237 Persi Schacht, Pablo Torres-Mancera, and Jorge Ancheyta Introduction 237 General Aspects of Homogeneous Catalysts 238 Water-Soluble Catalysts 239 Synthesis Procedure 240 Activity of Water-Soluble Catalysts 242 Oil-Soluble Catalysts 243 Synthesis Procedure 243 Activity of Oil-Soluble Catalysts 244 Mineral Catalysts 246 Synthesis Procedure 247 Single-Mineral Catalysts 248 Blend of Mineral Catalysts 248 Synthetic Catalysts Obtained from Minerals 249

Contents

5.5.2 5.6 5.6.1 5.6.2 5.7 5.8

Activity of Mineral Catalysts 250 Ionic Liquids 251 Synthesis Procedure 251 Catalytic Activity of Ionic Liquids 253 Catalysts Characterization 253 Concluding Remarks 257 References 258

6

Nanoparticles for Heavy Oil In Situ Upgrading 263 Muneer A. Suwaid, Sergey A. Sitnov, Ameen Al-Muntaser, Chengdong Yuan, Alexey Vakhin, Jorge Ancheyta, and Mikhail A. Varfolomeev General Aspects 263 Synthesis 264 Metallic Nanoparticles 265 Metal Oxide Nanoparticles 267 Coprecipitation Method 267 Sol–Gel Processing 269 Microemulsion Method 270 Preparation of Catalysts in Supercritical Water 271 Carbon Nanotubes (CNTs) 271 Characterization 273 Catalytic Activity 282 Monometallic Catalysts in Oil Upgrading 282 Magnesium 284 Aluminum 284 Silicon 285 Calcium 285 Titanium 285 Vanadium 285 Chromium 286 Manganese 286 Iron 286 Cobalt 288 Nickel 289 Copper 291 Zinc 291 Zirconium 292 Molybdenum 292 Cerium 292 Wolfram 293 Bimetallic, Polymetallic, and Supported Catalysts in Oil Upgrading 293 Nonsupported Bimetallic Catalysts 293 Nonsupported Polymetallic Catalysts in Oil Upgrading 294 Supported-Nanoparticles Coated with Nanoparticles of a Different Metal 295 Biogenic and Complex Organic Supports 299 Conclusions 300 References 300

6.1 6.2 6.2.1 6.2.2 6.2.2.1 6.2.2.2 6.2.2.3 6.2.2.4 6.2.3 6.3 6.4 6.4.1 6.4.1.1 6.4.1.2 6.4.1.3 6.4.1.4 6.4.1.5 6.4.1.6 6.4.1.7 6.4.1.8 6.4.1.9 6.4.1.10 6.4.1.11 6.4.1.12 6.4.1.13 6.4.1.14 6.4.1.15 6.4.1.16 6.4.1.17 6.4.2 6.4.2.1 6.4.2.2 6.4.2.3 6.4.3 6.5

ix

x

Contents

7

7.1 7.2 7.2.1 7.2.2 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.3.4.1 7.4 7.4.1 7.4.2 7.5 7.5.1 7.5.2 7.5.3 7.5.3.1 7.5.3.2 7.6 7.6.1 7.6.1.1 7.6.1.2 7.6.1.3 7.6.1.4 7.6.2 7.6.2.1 7.6.2.2 7.6.2.3 7.7 7.7.1 7.7.2

8

8.1 8.2

Catalytic Mechanism and Kinetics 309 Guillermo Félix, Alexis Tirado, Ameen Al-Muntaser, Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta Introduction 309 Reaction Mechanism During Heavy Crude Oil Upgrading 311 Reaction Mechanism During Hydrocracking of Heavy Crude Oils 311 Reaction Mechanism During Aquathermolysis Process of Heavy Crude Oils 313 Description of Reported Kinetic Models for In situ and Ex situ Hydrocracking of Heavy Crude Oil Applications 314 Four-lump Kinetic Models 316 Five-lump Kinetic Models 318 Six-lump Kinetic Models 322 Detailed-lumping Kinetic Model 323 Continuous Lumping Kinetic Model 325 Description of Reported Kinetic Models for the Aquathermolysis Process of Unconventional Reservoirs 326 Kinetic Models for Gas Prediction 326 Kinetic Models for Liquid Composition and Gas Generation 329 Methods for Calculating Kinetic Parameters and Model Assumptions 337 Hydrocracking Kinetic Models 337 Aquathermolysis Kinetic Models 341 Methodology to Ensure the Optimal Set of Kinetic Parameters 344 Study Case in Hydrocracking of Heavy Oils 345 Study Case in Aquathermolysis of Heavy Oils 345 Results and Discussion 347 Hydrocracking 347 Global Reaction Order for Residue Conversion 347 Calculation of Reaction Rate Coefficients 349 Activation Energies 352 Selectivity of Hydrocracking Reactions 364 Aquathermolysis 365 Reaction Order for Aquathermolysis Reaction 365 Reaction Rate Coefficients and Activation Energies 366 Discussion on Kinetic Studies and Modeling for Aquathermolysis Reaction 373 Conclusion 374 Hydrocracking Kinetic Models 374 Aquathermolysis Kinetic Models 374 References 375

Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms of In Situ Upgrading 382 Nail Khafizov, Vadim Neklyudov, Anastasiya Mikhailova, and Oleg Kadkin Introduction 382 A General View of In Situ Upgrading Processes from the Standpoint of Physical Chemistry 383

Contents

8.3 8.3.1 8.3.2 8.3.3 8.4 8.4.1 8.4.2 8.4.3 8.4.4 8.4.5 8.4.6 8.4.7 8.5 8.5.1 8.5.2

8.6

9 9.1 9.2 9.2.1 9.2.2 9.2.3 9.2.4 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.4

Quantum Chemical Approaches to the Calculation of Thermochemical and Kinetic Parameters of In Situ Upgrading Processes 385 Choice of Model Compounds for Simulating the In Situ Processes 385 Calculation Methods Used for Molecular Modeling of Reactants, Catalysts, and Simulation of Reaction Mechanisms 387 Thermochemical and Kinetic Parameters from the Quantum Chemical Calculation Results 388 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results 389 Initial Reaction Steps of Cleavage of Heteroatomic Bonds in Model Compounds and Their Thermochemical Parameters 389 Mechanism of Total Reaction of Cleavage of Heteroatomic Bonds in Model Compounds CPS, CPE, and CPA 394 Calculation of Thermodynamic and Kinetic Parameters of Aquathermal Decay of Cyclohexyl Phenyl Sulfide 395 Calculation of Thermodynamic and Kinetic Parameters of Aquathermal Decay of Cyclohexyl Phenyl Ether 401 Calculation of Thermodynamic and Kinetic Parameters of Aquathermal Decay of Cyclohexyl Phenyl Amine 407 Aquathermolysis Reactions of Dibenzyl Sulfide Under Conditions of Pyrolysis 412 Molecular Structure of Metal Stearate Catalysts and Simulation of Their Supramolecular Arrangement by MD Methods 415 Mechanisms of Aquathermal Pyrolysis of Asphalthene Fractions and Calculation Results 421 General Approaches to Quantum Chemical Calculations of Aquathermal Pyrolysis of Polycondensed Aromatic Compounds 421 Elucidation of the Mechanisms of Pyrene Oxidation on the Surface of Copper(II) Oxide Nanoparticles by Quantum Chemical Calculation Methods Based on the Results of Laboratory Experiments 422 Conclusions 427 References 430 Behavior of Catalyst in Porous Media 435 Timur R. Zakirov, Rail I. Kadyrov, Chengdong Yuan, and Mikhail A. Varfolomeev Introduction 435 Methods 436 Mathematical Model 436 Artificial Digital Models of Porous Media 440 Validation of Intraparticle Diffusion Model 440 Observation of the Catalyst Distribution in the Pore Space Using 4D Microtomography 441 Results and Discussion 442 Effect of Heterogeneity Coupled with Peclet Number 442 Effect of Heterogeneity Coupled with Damkohler Numbers 444 Effect of Heterogeneity Coupled with Porosity 445 Catalyst Distribution in the Pore Space 4D X-Ray CT 446 Conclusion 450 References 451

xi

xii

Contents

10 10.1 10.2 10.2.1 10.2.2 10.2.2.1 10.2.2.2 10.2.3 10.2.3.1 10.2.4 10.2.4.1 10.2.4.2 10.2.4.3 10.2.4.4 10.2.4.5 10.3 10.3.1 10.3.2 10.4 10.4.1 10.4.2 10.4.3

11 11.1 11.2 11.3 11.3.1 11.3.1.1 11.3.1.2 11.3.1.3 11.3.2 11.3.3 11.4 11.4.1.1 11.4.1.2 11.5 11.5.1 11.5.2 11.6 11.7

Numerical Simulation of Catalytic In Situ Oil Upgrading Process 453 Allan Rojas, Denis Shevchenko, Vladislav Sudakov, Sergey Usmanov, and Michael Kwofie The Reaction Scheme 453 Modeling the Phase Behavior of Oil 454 Oil Characterization 454 Correlations for Property Estimation of Hydrocarbons 455 Critical Parameters 455 Dead Oil Viscosity with Temperature Dependence 457 Special Data Requirement 458 Oil Viscosity with Temperature Dependence 458 The Cubic EoS and Phase Behavior 460 The Peng and Robinson EoS 460 Binary Interaction Coefficients 461 Volume Translation 462 Tuning an EoS 464 Lumping Sensitivity 466 Numerical Validation of Experimental Data 468 Numerical Validation in Static Conditions 468 Validation of Lab-Scale Kinetic Models in Dynamic Conditions 470 Upscaling Laboratory-Scale to Field-Scale 474 Some Approaches for Upscaling Steam Processes 474 Upscaling Laboratory Data 477 Criteria for the Selection of the Optimal Grid Type and Size 479 Nomenclature 482 Indexes 484 References 484 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil 489 Khusain Kadiev, Anton L. Maximov, and Jorge Ancheyta Introduction 489 Features of the Composition and Properties of Heavy Oil Feedstock 490 Main Directions of Processing of Oil Residues, Heavy Oils, and Bitumens 492 HOF Upgrading with Carbon Part Removal 492 Deasphalting 492 Thermal Cracking 492 Catalytic Cracking 493 Hydrogenation Processes for HOF Upgrading 493 Efficiency Analysis HOF Processing 494 Catalysts for HOF Hydroprocessing 499 Morphology of Catalysts 499 Hydroconversion Conditions in Slurry Processes 500 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion 500 Catalysts Derived from Oil-Soluble Precursors 501 Catalysts Synthesized from Water-Soluble Precursors 503 Principles of Sulfidation of Dispersed Molybdenum-Containing Catalysts 508 Formation of Coke-like Polycondensation Products and their Effect on the Structure and Catalytic Activity of MoS2 Suspensions 509

Contents

11.8 11.9 11.10 11.11

Synthesis Conditions and Catalytic Activity of Catalysts Synthesized Ex Situ 510 Behavior of Vanadium and Nickel at HOF Hydroconversion Using Suspensions of Nanosized Catalysts 514 Kinetic Parameters of Heavy Oil Feedstock Hydroconversion in the Presence of a Suspension of MoS2 Nanoparticles 515 Conclusion 516 References 516 Index 521

xiii

xv

List of Contributors Firdavs Aliev Department of Petroleum Engineering Kazan Federal University Kazan Russia

Sergey V. Efimov Institute of Physics Kazan Federal University Kazan Russia

Ameen Al-Muntaser Department of Petroleum Engineering Kazan Federal University Kazan Russia

Marsel G. Fazlyyyakhmatov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Jorge Ancheyta Department of Petroleum Engineering Kazan Federal University Kazan Russia Instituto Politécnico Nacional Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada Ciudad de México Mexico

Guillermo Félix Department of Petroleum Engineering Kazan Federal University Kazan Russia Airat H. Gabbasov PJSC Tatneft Almetyevsk Russia

and Instituto Mexicano del Petróleo Mexico City Mexico Roman S. Borisov A.V. Topchiev Institute of Petrochemical Synthesis Russian Academy of Sciences Moscow Russia

Marat R. Gafurov Institute of Physics Kazan Federal University Kazan Russia Ranel I. Galeev Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

xvi

List of Contributors

Khusain Kadiev A.V. Topchiev Institute of Petrochemical Synthesis Russian Academy of Sciences Moscow Russia Oleg Kadkin Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Rail I. Kadyrov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Anastasiia Yu. Kanateva A.V. Topchiev Institute of Petrochemical Synthesis Russian Academy of Sciences Moscow Russia Galina Kaukova Department of Petroleum Engineering Kazan Federal University Kazan Russia Nail Khafizov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Vladimir V. Klochkov Institute of Physics Kazan Federal University Kazan Russia

Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Michael Kwofie Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Georgy Mamin Institute of Physics Kazan Federal University Kazan Russia Anton Maximov A.V. Topchiev Institute of Petrochemical Synthesis Russian Academy of Sciences Moscow Russia Anastasiya Mikhailova Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Irek I. Mukhamatdinov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Fadis Murzakhanov Institute of Physics Kazan Federal University Kazan Russia

List of Contributors

Vadim Neklyudov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia and

Vicente Sámano Instituto Politécnico Nacional Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada Ciudad de México Mexico

Technion Haifa Israel

Persi Schacht Instituto Mexicano del Petróleo Mexico City Mexico

Danis K. Nurgaliev Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Denis Shevchenko Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Ilfat Z. Rakhmatullin Institute of Physics Kazan Federal University Kazan Russia and Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Alexander Rodionov Institute of Physics Kazan Federal University Kazan Russia Allan Rojas Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Department of Higher Mathematics Kazan Innovative University named after V. G. Timiryasov Kazan Russia Sergey A. Sitnov Department of Petroleum Engineering Kazan Federal University Kazan Russia Vladislav Sudakov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia Muneer A. Suwaid Department of Petroleum Engineering Kazan Federal University Kazan Russia

xvii

xviii

List of Contributors

Alexis Tirado Department of Petroleum Engineering Kazan Federal University Kazan Russia

Mustafa Versan Kok Middle East Technical University Üniversiteler Mahallesi Çankaya Ankara Turkey

Pablo Torres-Mancera Instituto Mexicano del Petróleo Mexico City Mexico

Chengdong Yuan Center for Petroleum Science and Engineering Skolkovo Institute of Science and Technology Moscow Russia

Fernando Trejo Instituto Politécnico Nacional Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada Ciudad de México Mexico

and Department of Petroleum Engineering Kazan Federal University Kazan Russia

Sergey Usmanov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Vladimir G. Zaikin A.V. Topchiev Institute of Petrochemical Synthesis Russian Academy of Sciences Moscow Russia

Alexey Vakhin Department of Petroleum Engineering Kazan Federal University Kazan Russia

Timur R. Zakirov Institute of Geology and Petroleum Technologies Kazan Federal University Kazan Russia

Mikhail A. Varfolomeev Department of Petroleum Engineering Kazan Federal University Kazan Russia

xix

About the Editors Dr. Mikhail A. Varfolomeev graduated in chemistry (2005) from the Kazan State University. He completed his PhD thesis in physical chemistry with focus on thermodynamics of fluids in 2007, also in the Kazan State University. He is a recipient of with more than 20 different national and international awards in education, research, and innovation areas. He is a coauthor of 17 patents and more than 310 papers (H-index of 30). He was invited as a researcher and professor in University of Rostock (Germany), IFP Energies Nouvelles (France), Southwest Petroleum University (China). Nowadays, Dr. Varfolomeev is a chair of Department of Petroleum Engineering and head of Enhanced Oil Recovery Center of the Kazan Federal University. His research interests include petroleum engineering, enhanced oil recovery, catalytic oil upgrading, in situ combustion, gas injection, chemical flooding, phase behavior, gas hydrates, thermodynamics, thermal analysis, and calorimetry. He was supervisor of more than 15 PhD students and 50 BSc and MSc theses. He actively cooperates with petroleum industry. He supervised more than 60 technical projects. A good number of them were introduced to the industrial scale. He actively participated in one of the world’s first successful pilot tests of in situ catalytic upgrading of heavy oil in Russia and Cuba. Dr. Varfolomeev is an associate editor of Journal of Petroleum Science and Engineering, Journal of King Saud University – Engineering Science, and member of Editorial Boards of Petroleum and Energies. He has given more than 40 plenary, keynote, and technical presentations on international conferences. Chengdong Yuan holds a PhD degree in oil and gas field development engineering through a combined master’s-PhD program from Southwest Petroleum University, China (2016). He graduated with a bachelor’s degree in petroleum engineering from Southwest Petroleum University (2011). He worked in Department of Petroleum Engineering of the Kazan Federal University as an associate professor (2019–2022), and in Department of Physical Chemistry of the Kazan Federal University as a senior researcher (2017–2022). Dr. Yuan has worked as Principal Research Scientist and Assistant Professor at Skolkovo Institute of Science and Technology since 2022. His academic interests focus on efficient hydrocarbon recovery. Specific interests include thermal methods for enhanced oil recovery including steam injection, in situ combustion (ISC), new technologies for in situ heavy oil upgrading, catalytic in situ oil upgrading, catalytic oxidation of crude oil, chemical flooding especially for interfacial phenomena (wettability of solids, interfacial tension, foams, and emulsions),

xx

About the Editors

and profile control and water shutoff technologies. He has been authorized 7 patents and is author and coauthor of more than 130 scientific papers (H-index of 24), has been awarded scientific scholarship in the field of research in pharmaceutics, chemistry, and petrochemistry, oil production, and oil and gas geology of the KFU Board of Trustees (2020). He has participated and given presentations in international conferences about 15 times since 2015, including 8 times SPE conferences presenting technical presentations. He was guest editor of the international journal FUEL of the special issue “In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils.” Jorge Ancheyta, PhD, graduated with a bachelor’s degree in Petrochemical Engineering (1989), master’s degree in Chemical Engineering (1993), and master’s degree in Administration, Planning, and Economics of Hydrocarbons (1997) from the National Polytechnic Institute (IPN) of Mexico. He splits his PhD between the Metropolitan Autonomous University (UAM) of Mexico and the Imperial College London, UK (1998) and was awarded a postdoctoral fellowship in the Laboratory of Catalytic Process Engineering of the CPE-CNRS in Lyon, France (1999). He has also been visiting professor at the Laboratoire de Catalyse et Spectrochimie (LCS), Université de Caen, France (2008, 2009, 2010), Imperial College London, UK (2009), Mining University at Saint Petersburg, Russia (2016, 2017), and Kazan Federal University, Russia (2021–2024). Dr. Ancheyta has worked for the Mexican Institute of Petroleum (IMP) since 1989 and his present position is manager of Products for the Transformation of Crude Oil. He has also worked as professor at the undergraduate and postgraduate levels for the School of Chemical Engineering and Extractive Industries at the National Polytechnic Institute of Mexico (ESIQIE-IPN) since 1992, and for the IMP postgraduate since 2003. He has been supervisor of more than one hundred BSc, MSc, and PhD theses. Dr. Ancheyta has also been supervisor of a number of postdoctoral and sabbatical year professors. Dr. Ancheyta has been working in the development and application of petroleum refining catalysts, kinetic and reactor models, and process technologies mainly in catalytic cracking, catalytic reforming, middle distillate hydrotreating, and ex situ and in situ heavy oils upgrading. He is author and coauthor of a number of patents, books, and about 250 scientific papers (H-index of 63), has been awarded the highest distinction (Level III) as National Researcher by the Mexican government, and is a member of the Mexican Academy of Science. He is principal associate editor of the international journal FUEL. Dr. Ancheyta has also chaired numerous yearly international conferences since 2004, namely International Symposium on Hydroprocessing of Oil Fractions (ISAHOF) and International-Mexican Congress on Chemical Reaction Engineering (IMCCRE).

xxi

Preface Heavy and ultra-heavy oil resources account for about 60–70% of total proved oil reserves all over the world, which are concentrated in various countries such as Russia, Mexico, Canada, and Venezuela. Due to the high viscosity and density of heavy oils, their production, transportation, and processing are much more difficult than conventional oils. For effective development, usually thermal methods are required to reduce the viscosity for the easy flow of heavy oils in the reservoir. Currently, steam injection is the most widely used thermal method for heavy oil recovery. However, during its application, various issues have been exposed, such as ● ● ●

Low efficiency with high energy and freshwater consumption for generating steam Environmental issues caused by the large consumption of freshwater and burning of coal or gas The viscosity of the recovered oil is not low enough on the ground, which increases difficulties and cost for its transportation and processing.

To solve these problems, using catalysts to initiate a catalytic aquathermolysis process for achieving a higher level in situ upgrading of heavy oils during steam injection is a promising solution, which, on the one hand, can improve the properties of heavy oils to ease the difficulties in transportation; on the other hand, can reduce the injection volume of steam, thus decreasing the consumption of energy and freshwater, reducing the cost, and improving the efficiency of steam injection. Various efforts have been made to improve the in situ upgrading and efficiency of steam injection by using different catalysts. For these reasons, it was identified that there was the need to have a document to summarize the theoretical aspects and current advances in the main topics related to in situ upgrading of heavy and extra-heavy crude oils. Catalytic In Situ Upgrading of Heavy and Extra-Heavy Crude Oils is organized in the following 11 chapters: Chapter 1 describes general aspects of definition, classification, and properties of crude oils, as well as detailed experimental data of typical crude oils around the world to achieve a better understanding of their composition. Chapter 2 deals with the description of advanced characterization of heavy crudes and their fractions. Particular emphasis is put on electron paramagnetic resonance (EPR) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy and relaxometry, Fourier transform infrared spectroscopy (FTIR), and chromatographic and mass spectrometry methods. The methods for in situ enhanced oil recovery (EOR) methods for heavy crudes recovery are detailed in Chapter 3. Chapter 4 aims at describing the fundamentals of in situ upgrading. Chapters 5 and 6 focus on the catalyst used for in situ upgrading, liquid catalyst, and nanoparticles. Chapter 7 deals with the different kinetic models for in situ upgrading, including noncatalytic aquathermolysis, catalytic aquathermolysis, and using hydrogen. Chapter 8 is devoted to the

xxii

Preface

application of quantum chemical calculations for studying thermochemistry, kinetics, and catalytic mechanisms of in situ upgrading. A general methodology, calculation techniques, and preliminary results of applying quantum chemistry methods for studying complex physicochemical phenomena that accompany the in situ upgrading processes are described. Chapter 9 is devoted to describing the behavior of a catalyst in porous media. A systematical investigation of the effect of pore space heterogeneity on the dynamics adsorption of catalyst dissolved in the water during a single-phase flow is studied, which allows for registering the catalyst distribution in the pore space using 4D tomography. Chapter 10 details the numerical simulation of catalytic in situ oil upgrading process, and Chapter 11 presents the novel technologies for upgrading heavy and extra-heavy oil. It is foreseen that Catalytic In Situ Upgrading of Heavy and Extra-Heavy Crude Oils becomes promptly an outstanding and distinctive book, not only for researchers that conduct investigations in this area, but also for BSc, MSc, and PhD students that need detailed information and explanations on how to carry out experiments and calculations in the topic of upgrading of heavy oils. We would like to thank all our colleagues that contributed with the preparation of chapters and for the support of the Russian Science Foundation related to the Project № 21-73-30023 dated 17 March 2021. June 2023

Mikhail A. Varfolomeev Chengdong Yuan Jorge Ancheyta Kazan, Russian Federation

1

1 Properties of Heavy and Extra-Heavy Crude Oils Alexis Tirado 1 , Guillermo Félix 1 , Fernando Trejo 2 , Mikhail A. Varfolomeev 1 , Chengdong Yuan 1,5 , Danis K. Nurgaliev 3 , Vicente Sámano 2 , and Jorge Ancheyta 1,2,4 1

Department of Petroleum Engineering, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Politécnico Nacional, Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada, Unidad Legaria, Ciudad de México, Mexico 3 Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia 4 Instituto Mexicano del Petröleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, Mexico City, 07730, Mexico 5 Center for Petroleum Science and Engineering, Skolkovo Institute of Science and Technology, Moscow, Russia 2 Instituto

1.1

Introduction

The increase in the population and continuous development of the global economy and industrialization has triggered a sharp growth in energy demand. Petroleum oil is the source of energy most used, and it is also the leading feedstock for various types of industries, among the manufacture of synthetic fibers, plastics, paints, fertilizers, insecticides, soaps, and synthetic rubber. Thus, the uses of petroleum as a source of raw material in manufacturing keep functioning the modern industry. According to the Organization of Petroleum Exporting Countries, oil demand is expected to increase by 16.4 MB/D between 2015 and 2040, reaching 109.4 MB/D by 2040. Conventionally, the energy demand has been covered by the exploitation of light oil resources. Nevertheless, petroleum is a nonrenewable resource that cannot be replaced naturally at the rate that it is consumed and is, therefore, a limited resource. Due to the decrease in light hydrocarbon reserves, it is essential to develop technologies capable of improving the production, transportation, and refinement of unconventional hydrocarbons reserves, such as heavy crude oil, extra-heavy crude oil, bitumen, among others, that represent 60–70% of the proven oil reserves around the world. The high content of high molecular weight hydrocarbon molecules with heteroatoms in their lattices (asphaltenes and resins) in this type of resource hinders their exploitation and utilization since these compounds tend to precipitate out, resulting in deposition and plugging of oil wellbores, pipelines, and surface facilities that cause various costly operational problems to oil producers. It is then of high importance to know the details of oil properties for the proper definition of the processes and catalysts that can be used for its upgrading. The objective of this chapter is to present the definition and classification of crude oils according to their constituents, as well as to describe the main properties used to characterize them. The relevancies of some physical and chemical properties, and standardized methods used for measurement are provided together with detailed experimental data of diverse samples of typical crude oils around the world in order to achieve a better understanding of petroleum composition and its constituents.

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

2

1 Properties of Heavy and Extra-Heavy Crude Oils

1.2

Heavy and Extra-Heavy Crude Oils

Petroleum or its equivalent term “crude oil” covers a wide assortment of materials consisting of a naturally occurring mixture of hydrocarbons and some organic compounds derivatives containing heteroatoms such as sulfur, nitrogen, oxygen, and trace amounts of some metals such as nickel and vanadium. Crude oil is derived from organic matter (dead plant and animal material) decomposed and exposed to certain temperatures and pressure for prolonged periods. This organic matter migrated from the original source beds to more porous and permeable rocks, where it has been buried and accumulated underground at pressure (depending on the depth) in geologic structures called reservoirs. An ensemble of reservoirs within a common rock structure or in separate structures but neighboring formations is currently referred to as an oil field (Abdel-Aal et al. 2003). Characterization of the different chemical species contained in crude oils remains a challenging task due to an immense range of moieties presented in terms of number of molecules and chemical structures. The variation of types of molecules present in petroleum is influenced by the temperature and pressure in the reservoir, the age of the oil field, as well as the origin and the relative amounts of the different constituents that form the original materials. Thus, the composition of petroleum around the world varies from one oil field to another, from one well to another in the same field, and even with the depth of an individual well. Therefore, it is possible that adjacent wells produce crude oils with diverse properties. Under pressure and temperature conditions on the reservoir surface (i.e. at the wellhead), low boiling hydrocarbon compounds (methane, ethane, propane, and butane) emerge from petroleum as gases. Meanwhile, higher boiling hydrocarbon derivatives (pentane and higher molecular weight compounds) remain in the liquid phase. Moreover, higher molecular weight hydrocarbon derivatives occur in the solid phase (i.e. wax derivatives), remaining dissolved in the liquid phase. The high hydrocarbons content in crude oil, with its diverse structures of alkane (paraffins), cycloalkane (naphthenes), and aromatic hydrocarbon derivatives, provides a source of usable products such as waxes, lubricants, diesel, gasoline, and various forms of petrochemicals. Certain crude oil reserves contain high proportions of hydrogen-rich compounds with relatively short hydrocarbon chains and low boiling components (low molecular weight). On the other hand, some oil reserves have been altered by aerobic biodegradation, where meteoric water supplies nutrients, oxygen, and bacteria attack the lighter alkanes, which increase the proportions of higher boiling components (higher molecular weight). This is the case of the deposits of extra-heavy crude oil and tar sands bitumen from the Orinoco Belt in Venezuela and Alberta Basin in Canada in which the oil has accumulated in a low-temperature environment allowing for the growth of bacterial communities feeding on the crude oil. Although most heavy crude oil reserves worldwide are the result of bacterial alteration of conventional oil, other factors can be responsible for the formation of heavy crude oils. For instance, the heavy crude oils deposits of California are explained by the nature of the sedimentary organic matter from which the oil is sourced. This organic matter is thermally labile and releases petroleum at an early stage of the burial history of the source rock, which results in a viscous, sulfur-rich, and thermally immature oil. Heavy crude oil, extra-heavy crude oil, and tar sand bitumen typically have a relatively high molecular weight fraction that comprises an assortment of different complex compounds deficient in hydrogen and with high carbon and heteroatoms (metal, sulfur, and nitrogen) content that significantly contribute to the poor fluid properties of the oil and providing low mobility. These properties hinder the extraction, upgrading, transport, and refining of these resources to produce high quality and economic value fractions, such as naphtha, kerosene, diesel, and liquefied petroleum gas (LPG).

1.3 Physical Properties

The classification of crude oil involves an assessment of various properties as well as knowledge of diverse methods of recovery from reservoirs. Conventional petroleum is often referred to as the crude material accumulated in reservoirs that can be recovered utilizing only naturally occurring forces or resulting from the injection of water or an immiscible gas at moderate pressure (primary and secondary recovery methods). In contrast to conventional crude oils, unconventional oils present low mobility through the reservoir due to their high viscosity. Thus, the recovery of these resources may require the application of thermal stimulation techniques for more efficient recovery of the oil, along with heating and dilution methods for their transportation to refineries by pipelines. Recent studies estimate that unconventional oil reserves, including heavy crude oil, extra-heavy crude oil, and tar sand bitumen, exceed 6 trillion barrels that is equivalent to about 70% of all energy resources derived from fossil fuels in the world. Some organizations have reported various classifications for crude oil based on certain properties such as sulfur content, API gravity, or viscosity. For instance, those crude oils with a sulfur content higher than 0.5 wt% are classified as sour oil, while crude oils with less than 0.5 wt% are termed as sweet oils (Gray 2015). Generally, the most used classification for crude oil is based on the API gravity, which is a measurement related to oil density (Section 1.3.1). The American Petroleum Institute considers light oil that type of petroleum that has higher than 31.1∘ API, while oil with API gravity between 22.3∘ and 31.1∘ API is classified as medium crude oils. Otherwise, a crude oil that has an API gravity less than 22.3∘ API is termed as heavy crude oil. Nevertheless, the criterion for oil classification based on the API degree range has not been standardized. Diverse oil organizations have adapted their own API scales considering financial aspects and the degree of technological improvement. The World Petroleum Congress and the US Geological Survey state that heavy crude oil has an API in the range of 10 and 20 and a gas-free viscosity between 100 and 10 000 mPa s at original reservoir temperature. On the other hand, Petrobras defines heavy crude oils as those in the range between 10∘ and 19∘ API (Huc 2010). Basically, heavy crude oil has a much higher viscosity (i.e. resistance to flow) and density than conventional crude oil and generally has sulfur content higher than 2 wt%. Consequently, recovery of heavy crude oil usually requires thermal stimulation of the reservoir content (Speight 2019). The term “extra-heavy crude oil” is used to define the subcategory of crude oil with less than 10∘ API and is more viscous than heavy crude oil. Nonetheless, this crude oil type has mobility in the reservoir under at reservoir temperature. On the other hand, tar sand bitumen is typically incapable of mobility (free flow) under the conditions in the tar sand deposit, presenting high viscosities (>10 000 cP). Therefore, tar sand bitumen requires more advanced methods than heavy crude oils. The mobility of extra-heavy crude oil is due to a high reservoir temperature (higher than the pour point of the extra-heavy crude oil) or other factors subject to variations and specific conditions in the reservoir. These classifications have been suitable for achieving a general understanding of the properties of crude oils despite their inaccuracy because they do not always reflect the true nature of heavy and extra-heavy crudes (Speight 2016).

1.3

Physical Properties

The importance of certain physical properties has been highlighted for the classification of crude oil and complications of recovery from reservoirs. Therefore, a clear understanding of the significance of these properties is required to provide vital insights into the behavior, characteristics, and quality of crude oil. Physical properties such as density, viscosity, and pour point may vary widely in crude oils from different locations. Certain methods that measure the physical properties of heavy crude oil and extra-heavy crude oil can identify the quality of crude oil (Ancheyta and Speight 2007).

3

4

1 Properties of Heavy and Extra-Heavy Crude Oils

1.3.1

Density, Specific Gravity, and API Gravity

Density is defined as the mass of a unit volume of a substance at a specified temperature (usually expressed for oils in dimensions of grams per cubic centimeter). On the other hand, relative density or specific gravity (sg) is the ratio of the weight of a unit volume of oil to the weight of the same volume of water at specified temperatures, which are usually determined at the standard temperature of 60 ∘ F (15.6 ∘ C). Different standardized procedures for measuring the density or specific gravity apply to heavy crude oil and may be measured utilizing a hydrometer (ASTM D287 and ASTM D1298) or using a pycnometer (ASTM D941 and ASTM D1217). Density and specific gravity are widely used in the industry for preliminary assessment of the character of the oil. The values for density (and specific gravity) cover a narrow range considering the differences in the feedstock appearance and behavior. To introduce a more meaningful relationship between the physical properties and processability of the various crude oils, the American Petroleum Institute devised a measurement of specific gravity to determine the quality of crude oils and refinery streams by means of the following equation: API =

141.5 − 131.5 sg

(1.1)

The API gravity scale helps emphasize differences in specific gravity values between crude oil samples. For instance, crude oil samples might have specific gravities of 0.88 and 0.84, which would appear to be a small difference. However, the API gravities of these liquids are 29 and 37, respectively. The specific gravity usually ranges from about 0.8 (45.3∘ API) for the lighter crude oils to over 1.0 (10∘ API) for heavy crude oils and bitumen, which is consistent with the general trend that increased aromaticity leads to a decrease in API gravity (i.e. an increase in specific gravity). The extra-heavy crude oil and bitumen have an API gravity less than 10, which means a specific gravity higher than 1, being heavier than pure water. The variation of density with temperature is a property of great technical importance since most petroleum products are processed, treated, and marketed based on their volume, which has fundamental application in both petroleum production and processing as well as its transportation and storage. These properties are used in the calculations related to the sizing of pipes, valves, and storage tanks, the power required by pumps and compressors, and flow-measuring devices. In addition, density oil is used in reservoir simulation to estimate the amount of oil and gas in a reservoir, as well as the amount of their production at various reservoir conditions (Ancheyta and Speight 2007; Riazi 2005).

1.3.2

Viscosity

Viscosity is the measure of the internal resistance to fluid motion because of the forces of cohesion between molecules or molecular groups. It characterizes the motion of crude oils and is typically expressed in g/cm s or Poise (1 P = 100 cP). Since viscosity varies with temperature, measured values of viscosity must be reported at specified temperatures. Generally, the kinematic viscosity of petroleum fractions is measured at standard temperatures of 37.8 ∘ C (100 ∘ F) and 98.9 ∘ C (210 ∘ F). The viscosity values of crude oils vary markedly over a wide range from less than 10 cP to many thousands of centipoises at room temperature. Moreover, this parameter depends on soil composition, pressure, and gas solubility, and it is a critical property in predicting oil recovery since viscosity reduction and thermal expansion are the key properties for increasing the productivity of heavy crude oil. As stated previously, heavy crude oils present high viscosity values (in comparison with conventional oils), being the main drawback for its production from reserves and transportation

1.4 Chemical Properties

through the pipelines turning difficult the production process. Some documents have set the maximum viscosity at 250 cSt (100 ∘ F) and the minimum API gravity at 16∘ for transportation purposes. Several standardized methods are available for the experimental determination of viscosity for different types of crude oils and petroleum products. The most used are ASTM D88, D341, D445, D2161, D2170, D2270, D7483, and E102. These methods differ concerning the type and required amount of sample, the experimental setup, the time for analysis, the operating conditions of apparatus, and the viscosity ranges in which the device can be used. Therefore, the experimental determination of the viscosity was made based on the characteristics of the sample. In practice, it has been observed that measuring the viscosity of crude oils with low API gravity is complicated due to their nature and difficulty in handling. This drawback makes the analysis time-consuming and more amount of sample to obtain reliable results is necessary (Sánchez-Minero et al. 2014; Speight 2013). Some terms have been derived from viscosity to determine the behavior and handling of fluids. The kinematic viscosity is the ratio of viscosity and specific gravity. The unit of kinematic viscosity is the stoke (cm2 /s), although centistokes (cSt) are commonly used units. On the other hand, fluidity is defined as the reciprocal of viscosity.

1.3.3

Pour Point

The pour point is referred to as the lowest temperature at which the oil flows under standard test conditions (ASTM D97). The pour point of crude oils is a critical parameter to the proper functioning of the production line and generally varies from 52 to −59 ∘ C (125 to −75 ∘ F). This parameter is influenced by the content of waxes, viscosity, and even the thermal history of the sample, that is the degree and duration of heating and cooling to which the sample has been exposed. Moreover, it is emphasized that the tendency of the oil to flow is influenced by the size and shape of the container, the head of the oil, and the physical structure of the solidified oil. The pour point of the oils is, therefore, a predictor of the temperature at which flow ceases in the reservoir. This term has significant consideration since, for efficient production of unconventional oils (heavy crude oil, extra heavy crude oil, among others), additional energy must be supplied to the reservoir by a thermal process to increase the reservoir temperature beyond the pour point and improving the ability of the heavy crude oil to flow (Riazi 2005; Speight 2014).

1.4

Chemical Properties

Heavy and extra-heavy crude oils are complex mixtures of different hydrocarbon types, and their composition depends on their geological origin. These crude oils present larger amount of heavy molecular weight hydrocarbons, hindering the refining process due to the impurities content such as heteroatoms and metals. Although the number of heavy fractions is small compared with the number of countless compounds in this type of crude oils, the adverse effect that these impurities cause is high (Caumette et al. 2009; Chacón-Patiño et al. 2022; Santos et al. 2014; Tirado et al. 2022; Yadykova and Ilyin 2022). Chemical properties are indispensable to know the quality of heavy or extra-heavy crude oils in order to evaluate the refining capacity likewise the type of products generated. Therefore, a good manner to know the complexity of a petroleum to refine is by analyzing the chemical composition since this conformation of elements will determine the nature of reactions performed in the upgrading process. In this way, the products generated during the reaction depend on the composition of the heavy or extra-heavy crude oil fed (Ancheyta and Speight 2007; Kayukova et al. 2017; Rodrigues et al. 2018; Silva et al. 2011).

5

6

1 Properties of Heavy and Extra-Heavy Crude Oils

Most of the heteroatoms and metals are presented in highly condensed and large molecular weight compounds such as asphaltenes and resins. A feedstock with low content of heavy fraction such as asphaltenes, and thus small number of heteroatoms and large number of light fractions, is more easily processed in a refinery than a feed with greater content of heavy molecules. The reason is because in these types of crude oils a series of complex reactions to remove the impurities takes place (such metals and heteroatoms). Moreover, these impurities tend to deactivate the catalyst employed to upgrade the feedstock into more valuable products hindering the refining process (Muraza and Galadima 2015; Yadykova and Ilyin 2022; Zhao et al. 2005).

1.4.1

Elemental Analysis (CHONS)

One of the first methods to evaluate the general quality of petroleum is the elemental analysis, which provides the percentages of carbon (C), hydrogen (H), oxygen (O), nitrogen (N), and sulfur (S) as the basic constituents of the heavy and extra heavy crude oils whose contents range for different types of petroleum, depending on the origin, in the following intervals: C, 83.0–87.0%; H, 10.0–14.0%; N, 0.1–2.0%; O, 0.05–1.5%; S, 0.05–6.0%. However, the quality of heavy and extra heavy crude oils depends on the heteroatoms content (O, N, and S) as well as hydrogen content since a difference in these values provokes significant changes in their properties (Ancheyta and Speight 2007; Dehkissia et al. 2004; Riazi 2005). Most of the elemental analyzers use a similar methodology, where the sample is combusted in an oxygen atmosphere, while product gases are separated and analyzed with different detectors based on different methods (ASTM D5291, Chinese industry standard SY/T 5122, classical Pregl-Dumas method, etc.) (Li et al. 2018; Riazi 2005; Zhang et al. 2021). Generally, the oxygen content is calculated by the difference of 100 and the sum of C, H, N, and S content (Ganeeva et al. 2021; Leyva et al. 2013; Medina et al. 2022). In addition, there are some methods to analyze specific elements separately as depicted in Table 1.1 (Ancheyta and Speight 2007; Kumar et al. 2018; Quitian and Ancheyta 2016; Rana et al. 2008; Riazi 2005). Among all the heteroatoms, which are other atoms apart of carbon and hydrogen, sulfur presents the highest quantity in heavy and extra-heavy crude oils as thiols, sulfides, thiophene, benzothiophene, dibenzothiophene (DBT), and its derivatives that are considered as undesirable compounds. Sulfur aromatic compounds (such as alkyl-substituted DBT) are difficult to convert during the refining of heavy and extra-heavy crude oils because of their steric hindrance as they attempt to adsorb on the catalyst surface (Ancheyta et al. 2009; Santos et al. 2014). Whereas nitrogen in heavy and extra-heavy crude oils is generally present in the form of basic (pyridine) and nonbasic (pyrroles, Table 1.1

Standard methods for elemental analysis.

Elemental analysis

Method

C

ASTM D3178, D5291, E777

H

ASTM D1018, D3178, D3343, D3701, D4808, D5291, D7171, E777

O

ASTM E385

N

ASTM D3179, D3228, D3431, D4629, D5291, D5762, E148, E258, E778

S

ASTM D129, D1266, D1552, D1757, D2622, D3177, D4045, D4294, D4952, D5453, IP 30, IP 61, IP 107, IP 243, IP 336, IP 447

Source: Ancheyta and Speight (2007)/Informa UK Limited; Garcia-Montoto et al. (2020)/Elsevier; Kumar et al. (2018)/Elsevier; Quitian et al. (2016)/American Chemical Society; Rana et al. (2008)/Elsevier; Riazi (2005)/Google LLC; Speight (2002)/Google LLC.

1.4 Chemical Properties

indoles, and carbazoles) compounds, being more stable compounds than sulfur species (Santos et al. 2014; Wei et al. 2015). Oxygen is present in low amounts as carboxylic and phenolic groups mainly, although the presence of ketones, ethers, and anhydrides has also been reported by influencing the acidity of petroleum crude oil, which is especially important during refining processes and hence affects its market price (Santos et al. 2014). The content of different elements, such as C, H, and O, can be related to the rearrangements of the different compounds containing these elements in the petroleum during the refining process. Furthermore, the ratios of these elements provide knowledge about the quality of the heavy fractions (Ancheyta and Speight 2007; Riazi 2005). Another indicator of the quality of the petroleum is the carbon-to-hydrogen mass percent ratio (C/H) because if the C/H relationship increases, the molecular weight and API gravity of petroleum increases and decreases, respectively. Moreover, a crude oil with a low C/H ratio is a good feedstock for the refining process because of the low hydrogen requirement. This behavior is also observed with the content of sulfur and nitrogen, increasing the quality of a heavy or extra-heavy crude oil by decreasing the content of these elements (Dehkissia et al. 2004; Riazi 2005). A preliminary hint of the aromaticity in crude oils can be obtained by the hydrogen-to-oil (H/C) atomic ratio which is calculated with Eq. (1.2) since the aromaticity decreases when the H/C ratio increases. This atomic ratio supplies a quantitative estimation of the nature in petroleum, considering that the aromatic molecules such as toluene (H/C ratio of 1.14) have low values, while the aliphatic constituents like heptane (H/C ratio of 2.29) have higher values (Leyva et al. 2013; Li et al. 2022; Riazi 2005; Sullivan et al. 2007): H∕C(atomic ratio) =

11.9147 C∕H(weight ratio)

(1.2)

The nitrogen and sulfur atoms are usually found in aromatic compounds where sulfur species are present in a large amount. The amount of sulfur in crude oil fractions increases as the molecular weight of the compound does, where asphaltene is the fraction containing the higher amount of sulfur. There are more catalysts focused on sulfur removal than nitrogen deletion since the latter is more difficult to eliminate while only a few catalysts have been developed to remove nitrogen. Moreover, the sulfur conversion during petroleum upgrading can reach up to 99% using novel catalysts in refineries (Ancheyta and Speight 2007; Leyva et al. 2013; Shafiq et al. 2020). The content of these two atoms in the heavy crude oil enhances the production of pollutant gases such as nitrogen oxides (NOx ) and sulfur oxides (SOx ). In addition, the sulfur compounds in crude oil products like gasoline or diesel contribute to the pollution through the exhaust gases and cause rusting and corrosion of the engine. Gas emission is penalized if sulfur concentration overcomes the upper limits allowed by local laws (Ancheyta and Speight 2007; Shafiq et al. 2020).

1.4.2

Metal Content

Heavy and extra-heavy crude oils display different concentrations of metals depending on their origin. Nickel (Ni), vanadium (V), and iron (Fe) are the most abundant metals in heavy crude oils ranging between 10 and 1000 ppm. However, other elements such as lead (Pb), barium (Ba), tin (Sn), silver (Ag), cobalt (Co), copper (Cu), molybdenum (Mo), titanium (Ti), and zinc (Zn) are present as well in fewer concentrations (1–50 ppm range) in heavy crude oils depending on their quality or composition because when the API gravity of crude oil decreases, the content of metals increases (Caumette et al. 2009; Dehkissia et al. 2004; Riazi 2005; Zhao et al. 2013). Metals are commonly present as porphyrins, nonporphyrin, and naphthenic acid salts, and different methods are applied to measure their content. Hence, the following individual analyses are employed to quantify the content of metals: ASTM D1548, D2788, D3605, D4628, D4927, D5708, D5863, D6443, IP 265,

7

8

1 Properties of Heavy and Extra-Heavy Crude Oils

IP 285, IP 288, IP 433, and IP 465. In all these methods, the sample is burnt to ash and diluted in acid to measure the absorbance of the metal by atomic absorption spectroscopy or inductively coupled argon plasma spectrometry (Ancheyta and Speight 2007; Rana et al. 2008; Riazi 2005; Speight 2002). The metalloporphyrins are constituted by several porphyrins bonded to a central metal (Ni, V, Fe, or Cu), and more than 50 types of porphyrins have been identified in crude oils containing Ni or V, mainly. The nonporphyrin metal compounds are less polar than metalloporphyrins and account for 50–80% of the metal species. Nonetheless, these compounds are currently hypothesized since there is not enough characterization of them. The naphthenic acid salts, which are the less abundant metal species, are linked to Ca, Mg, Zn, and Ti, but there is no evidence of the presence of V or Ni-based naphthenic acid complexes in petroleum (Caumette et al. 2009; Mironov et al. 2018). Metal compounds tend to concentrate in heavier fractions like residue, asphaltenes, or resins, and are connected by strong bonds or surrounded in macromolecular networks. Metalloporphyrins can be directly related to the type of rocks where the heavy or extra heavy crude oil is formed. Furthermore, the V-to-Ni concentration ratio provides a hint of the geological conditions of sedimentation to which the crude oil was exposed (Caumette et al. 2009; Gao et al. 2013; Mironov et al. 2018; Riazi 2005). The content of V and Ni impacts negatively in the refining process because their dehydrogenation activity increases the amount of coke and gases that are generated, decreasing the yield of liquid. Moreover, these metals also decrease the catalyst activity due to the poisoning of the active sites (Ancheyta and Speight 2007; Caumette et al. 2009; Riazi 2005; Santos et al. 2014; Shang et al. 2016). Vanadium-based species present in diesel fuels produce corrosive compounds, which at elevated temperatures damage some engine parts; hence, magnesium-based additives are employed to avoid corrosion. Nevertheless, the presence of lead suppresses the effect of these additives enhancing corrosion. Calcium compounds do not cause corrosion and even help to inhibit the corrosive action of vanadium compounds in refineries, yet they generate deposits that cannot be easily removed (Shang et al. 2016).

1.4.3

Carbon Residue

Carbon residue is a good predictor of crude oil quality, as well as the amount of carbonaceous deposits (asphalt or coke) that can be produced by the influence of heat (Duarte et al. 2016; Rodrigues et al. 2018). There are three types of carbon residue analyses that are applied to crude oils: Conradson carbon residue (ASTM D189, IP 13, JIS K2270-1, ISO 6615:1983, GB/T268-1987), Ramsbottom carbon residue (ASTM D524, IP 14), and the Micro carbon residue (ASTM D4530, IP 398, JIS K2270-2, ISO 10370). The application of these analyses must be carried out on crude oils free of volatile compounds as they are subjected to atmospheric distillations. Care must be taken during the analyses if samples easily produce ash, owing to measurements that may be erroneously obtained. These tests can be also applied to evaluate the deposits of carbon in engines generated by fuels (Ancheyta and Speight 2007; Kumar et al. 2018; Palacio Lozano et al. 2017; Riazi 2005; Speight 2002). Although these methods are based on the distillation of a hydrocarbon sample, differences among them arise. For the Conradson carbon residue method, the sample is burnt in a crucible for a determined period weighing the residue at the end; for the Ramsbottom test, the sample is weighed and placed into a glass bulb that has a capillary opening. Later, it is heated in a furnace at 550 ∘ C for a fixed period, and at the end, the carbonaceous deposits are weighed. On the other hand, for the Micro carbon residue method, the sample is burnt (at 500 ∘ C) in an inert atmosphere for a specific time. Finally, the carbon residue remaining is weighed. All these methods are expressed as a percentage (wt/wt) of carbon residue in the initial sample. Among these techniques,

1.5 Composition

the Conradson and Micro carbon residue tests can be correlated directly, whereas the Micro carbon residue is the preferred technique for analyzing carbon residue since several runs are carried out simultaneously using small amounts of sample, keeping the distillation well controlled (Duarte et al. 2016; Palacio Lozano et al. 2017; Speight 2002). A high content of metallic compounds, as observed in heavy and extra-heavy crude oils, can interfere in the measure of carbon deposits since the metals remain in residues. Therefore, these compounds need to be eliminated from the sample or taken as ash by the complete combustion of carbons deposits after the analysis (Riazi 2005).

1.4.4

Molecular Weight

The molecular weight of a heavy or extra-heavy crude oil is the average number of the molecular weight of the complex mixture of components present (at least several thousand) or the weight average molecular weight of all constituents. There is a wide variety of methods to determine the molecular weight that is divided into those that do not require any standard and those that require calibration with a material of known molecular weight (Azinfar et al. 2018; Speight 2014). These analyses can be classified into methods that determine an average molecular weight value and those providing a complete distribution. Among them, viscosity vapor pressure osmometry (VPO, ASTM D2503, D2878, and UOP 676-84) and gel permeation (size exclusion) chromatography (GPC, ASTM D5296, and D6579) have been widely used because methods requiring calibration are generally easier and faster to be done. The main difference between these two analyses is that in GPC, the separation depends on the size of molecules, taking more time to pass through the chromatography column for the small molecules than for bigger compounds, providing a molecular weight distribution. Whereas VPO relies on the difference in vapor pressure when a drop of solute is added to a drop of pure solvent due to the molecular weight is related to the change of the vapor pressure in the solvent. The result is reported as molecular weight average (Álvarez et al. 2019; Ancheyta and Speight 2007; Azinfar et al. 2018; Castro and Vazquez 2009; Peramanu et al. 1999; Satya et al. 2007; Zhao et al. 2005). In addition, new methods, such as Matrix-assisted laser desorption/ionization–time of flight (MALDI TOF), have been applied to heavy crude oils to measure the molecular weight distribution. The need of a small amount of sample (liquid or solid) is a huge advantage for this analysis (Kim et al. 2016; Zhang et al. 2015; Zheng et al. 2018). The combination of methods to measure the distribution of molecular weight has been carried out for heavy crude oil, such as the use of GPC together with VPO, high performance liquid chromatography (HPLC), and several spectroscopy methods (Azinfar et al. 2018; Dettman et al. 2005; Leontaritis and Mansoori 1989). The molecular weight is a suitable property to know the quality of heavy and extra-heavy crude oils since it provides the average molecular mass or the molecular mass distribution of the vast mixture of compounds. Furthermore, the molecular weight distribution can be beneficial for the characterization of these complex mixtures because simulations (computational thermodynamics, phase equilibria, etc.) of the hydrocarbon systems are carried out taking this distribution (Azinfar et al. 2018; Speight 2002). The molecular weight in the upgrading of heavy and extra-heavy crude oils illustrates the conversion from higher-to lower-molecular weight fractions, providing evidence of the reactions nature of these feeds (Ancheyta and Speight 2007).

1.5

Composition

Crude oil has been commonly lumped into four fractions according to the literature, and different techniques have been used for its fractionation into saturates, aromatics, resins, and asphaltenes,

9

10

1 Properties of Heavy and Extra-Heavy Crude Oils

which are defined in terms of their solubility in different solvents. It is considered that asphaltenes impact on deposition during production and processing of petroleum. The classical colloidal model establishes that asphaltenes are in the core of solid particles surrounded by resins and aromatic molecules. Resins and asphaltenes are formed by polar heteroatoms, but both fractions differ in their solubility in alkanes (pentane or heptane). Asphaltenes are insoluble in alkanes, while resins are miscible. Thus, asphaltenes and resins constitute the disperse phase considering the crude oil as a colloidal system, whereas saturates and aromatics correspond to the continuous phase (Ashoori et al. 2017). The composition of crude oils based on SARA fractionation, true boiling point (TBP) curves, and characterization through elemental analysis, metals content, and carbon residue, among other properties, allows knowing the crude oil behavior during refining, particularly when obtaining different cuts through atmospheric and residue distillation. Depending on the content of heavy fractions, more or less yield of distillates is obtained. During distillation, naphtha, kerosene, diesel, gasoil, atmospheric, and vacuum residue are obtained as temperature increases. For each fraction, different characterization tests are required. For example, for naphtha fraction, it is commonly reported its density, aromatics and naphthenes content, octane number, and sulfur content. For kerosene fraction used as aviation fuel, the density, sulfur content, pour point, freezing point, and aromatic content is required, while if kerosene is used as diesel fuel precursor, then the cetane index, pour point, density, and cold filter plugging point are reported. For diesel fraction, it is required to know its density, sulfur and aromatic contents, cetane index, pour point, and cold filter plugging point. For gasoil fraction, the UOP K factor, nitrogen, and Conradson carbon content are reported since this fraction is upgraded in the Fluid Catalytic Cracking (FCC) units. For atmospheric residue and vacuum gasoil, the density, pour point, sulfur and metals content, viscosity, UOP K factor, and Conradson carbon content are determined since atmospheric residue is upgraded through FCC, while hydrocracking is used for vacuum gasoil. The vacuum residue cut is characterized by density, pour point, sulfur and metals content, UOP K factor, and Conradson carbon content (Stratiev et al. 2010). The crude oil base is characterized according to the K factor by which the crude oil is paraffinic base (K factor: 12.9–12.2), intermediate base (K factor: 12.2–11.5), or naphthene base (K factor: 11.5–10.5) (Behrenbruch and Dedigama 2007). Taking into consideration the aforementioned fractions and properties, the API gravity for cuts at different temperature intervals is plotted in Figure 1.1 for two reference crude oils (West Texas Intermediate from the USA and Brent blend from the North Sea) and four heavy crude oils (Maya from Mexico, Lloydminster from Canada, Emeraude from Congo, and Tia Juana from Venezuela). Naphtha is mainly composed of light compounds by which its API gravity is quite high for all types of plotted crude oils. This value decreases as the temperature interval increases. However, it is observed that for the heaviest crude oil (Tia Juana), the API gravity has the lowest value in the atmospheric residue. Another important property commonly reported for each fraction is the sulfur content, as observed in Figure 1.1 as well. In this case, the opposite behavior regarding API gravity is attained as expected, i.e. the heavier the fraction, the larger the sulfur amount since its heteroatoms are more difficult to be removed when present in heavy fractions. If the same crude oils plotted before are divided according to their atmospheric and vacuum residues and characterization tests are carried out, it is observed in Figure 1.2 that Conradson carbon content is higher in the vacuum residue because more refractory compounds are deposited in cuts having the highest temperature interval. The same trend is observed for the metals (Ni + V) content. Low content of metals is observed for lighter crude oils; however, heavy crude oils display high concentration of metals in vacuum residue where Maya crude oil showed the largest (Ni + V) content. Asphaltene content also increases in the vacuum residue, being the Maya crude oil with

1.5 Composition

100 90 80

WTI

Brent

Maya

Tia Juana

Lloydminster

Emeraude

API gravity (°)

70 60 50 40 30 20 10 0

IBP−70 °C 70−100 °C 100−190 °C 190−235 °C 235−280 °C 280−343 °C 343−565 °C Light naphtha

Medium naphtha

Heavy naphtha

Kerosene

Diesel

Gasoil

Atmospheric residue

3.5

Sulfur content (wt%)

3.0

WTI

Brent

Maya

Tia Juana

Lloydminster

Emeraude

2.5 2.0 1.5 1.0 0.5 0.0 IBP−70 °C 70−100 °C 100−190 °C 190−235 °C 235−280 °C 280−343 °C 343−565 °C Light naphtha

Medium naphtha

Heavy naphtha

Kerosene

Diesel

Gasoil

Atmospheric residue

Figure 1.1 API gravity and sulfur content for different fractions based on temperature intervals. Source: Adapted from Stratiev et al. (2010).

11

1 Properties of Heavy and Extra-Heavy Crude Oils

Conradson carbon (wt%)

30

Atmospheric residue Vacuum residue

25 20 15 10 5 0

Ni+V content (ppm)

WTI 900 800 700 600 500 400 300 200 100 0

Brent

Tia Juana Lloydminster

Maya

Emeraude

Tia Juana Lloydminster

Maya

Emeraude

Tia Juana Lloydminster

Maya

Emeraude

Atmospheric residue Vacuum residue

WTI 18

Brent

Atmospheric residue Vacuum residue

16 Asphaltenes (wt%)

12

14 12 10 8 6 4 2 0 WTI

Figure 1.2

Brent

Conradson carbon, metals, and asphaltenes deposited in atmospheric and vacuum residue.

the highest content. Despite being considered as heavy, in Emeraude crude oil (12.1∘ API), the metal, asphaltene, and Conradson carbon contents are quite low compared with other heavy crude oils plotted in the figure.

1.5.1

SARA Analysis

Heavy and extra-heavy crude oils contain large amounts of the so-called resin and asphaltene fractions with high aromaticity and polar atoms such as nitrogen, sulfur, and oxygen. Most of the techniques used on lighter crude oils are not applicable to these crude oils by which separation based on polarity needs to be carried out as preparative method (Merdrignac and Espinat 2007).

1.5 Composition

Separation of fractions from heavy and extra-heavy crude oil involves precipitating firstly asphaltenes and then, the deasphalted oil (DAO) is passed through an open-column chromatograph by which four fractions are obtained based on the standard method ASTM D4124. Asphaltene separation includes the use of linear alkane such as pentane or heptane. The heavy crude oil and heptane are placed into a flask to be heated and stirred. Further, the sample is cooled to room temperature, and filtration is carried out to retain the solids that correspond to asphaltene fraction, while the liquid is commonly called as maltene fraction or DAO, and it is composed of saturates, aromatics, and resins. The solvent used to precipitate asphaltenes is evaporated to recover the maltene fraction, and the open-column chromatography is used to recuperate the remaining fractions in a column that is packed with activated alumina. The maltene fraction is dissolved in dichloromethane and poured into the column. Once maltenes were adsorbed in the activated alumina, different solvents are used to elute saturate, aromatic, or resin fraction. Thus, heptane is added to the column to extract the saturate fraction, while toluene separates aromatics. Furthermore, a toluene/methanol mixture is used to recover resins. Each fraction is sequentially drained from the bottom of the column and solvent used to separate the fraction is evaporated. Finally, each fraction is weighed and reported as the percentage of the whole sample. Examples of separations based on SARA analysis have been reported in the literature (Park et al. 2022). Aromaticity factor, which is a parameter related to the amount of aromatic carbon in regard to total carbon content, influences the tendency to form coke when upgrading heavy crude oil. This factor is quite low in heavy fractions such as resins and asphaltenes because aromatics in heavy and extra-heavy crude oils ranges widely (Alonso-Ramírez et al. 2021, 2020; Félix and Ancheyta 2019; Ortiz Moreno et al. 2014). Aromatic fraction is mainly composed of mono-, di-, tri-, and polyaromatics having alkyl chains. Major changes in structure and compositions are observed in asphaltene fraction. When processing heavy or extra-heavy crude oils, resins are converted into lower molecular weight aromatics. Even distillates are formed as a consequence of cracking heavier molecules. However, resins may form condensed structures such as asphaltenes. If more asphaltenes are formed, then the resin fraction is not enough to peptize asphaltenes and colloidal stability is decreased. During crude oil upgrading, the higher reaction temperatures enhance the dealkylation reaction and formation of free radicals that condense to form coke (Wang et al. 2012). Resins tend to associate with asphaltenes and disperse them into the crude oil. It is considered that asphaltene molecules are in the core of micelles surrounded by resins. The knowledge of SARA fractions in heavy and extra-heavy crude oils allows for predicting the relative stability of crude oils based on the nature of asphaltenes and dispersion medium. However, the stability of the crude oil depends on the interaction among all fractions. Aromatics also contribute to keep asphaltenes soluble by solvating the aromatic part of asphaltenes, while the polar section is solvated by resins. On the contrary, saturate molecules make asphaltene flocculate and precipitate. Thus, higher amounts of aromatics and resins will keep asphaltene fraction soluble in the crude oil (Ashoori et al. 2017). Considering the SARA fractions, the colloidal instability index (CII) is used as a preliminary test analysis for stability/instability of heavy or extra heavy crude oils. This value is calculated as follows: Saturates + Asphaltenes (1.3) CII = Aromatics + Resins Below 0.7, the crude oil is considered as stable, while instability occurs when CII > 0.9. More stable crude oils keep asphaltene soluble and avoid its precipitation. The crude oils with values between 0.7 and 0.9 correspond to metastable with minor problems of deposition. Thus, the CII criterion may recognize potential troubles of asphaltenes deposition through SARA fractionation,

13

14

1 Properties of Heavy and Extra-Heavy Crude Oils

which may turn it into a screening test for preliminary examination on stability/instability of crude oils. Reports in the literature concluded that CII criterion predicts better the instability than stability of crude oils (Guzmán et al. 2017). Other indicators based on Chamkalani Stability Classifier (CSC), Stankiewicz Plot (SP), and Modified Jamal plot predict well instability of crude oils, while Stability Index (SI) and Jamaluddin plot (Jamal) are conveniently applied as stability predictors of crude oils (Ali et al. 2021). Different SARA compositions have been reported in the literature as well as API gravity. Table 1.2 summarizes the values of saturate, aromatic, resin, and asphaltene fractions along with the API gravity of heavy and extra-heavy crude oils. Additionally, the colloidal instability index for each sample is reported and further is plotted against API gravity as shown in Figure 1.3 with data obtained from Table 1.2. For heavy and extra-heavy crude oils, the API gravity ranged from 4∘ API up to 21∘ API. Taking into consideration that crude oil stability is achieved when the CII value is below 0.7, it is observed that most of crude oils are stable according to this parameter by which stability is attained despite having a wide range of API gravity. However, the SARA composition needs to be also considered to predict the crude oil behavior during transportation and/or upgrading. When plotting each SARA fraction and the colloidal instability index for the parent crude oil (Figure 1.4), it is observed that aromatics, resins, and asphaltenes may vary in wide interval, and stability is performed in most of cases. However, when saturate fraction and CII value are plotted, it may be stated that crude oil is stable as fewer amount of saturates is present in the crude oil. For most of plotted crude oils, stability is attained if saturate content is lower than 30 wt%. Higher concentration of saturates enhances the crude oil to be instable since more paraffinic compounds cause destabilization of asphaltenes in the micelle and precipitation occurs. Heavy crude oils may be instable not necessarily if having high content of asphaltenes as obtained from SARA fractionation. Most commonly, the higher the saturate fraction, the higher the crude oil instability. On the contrary, aromatic fraction is a good solvent for asphaltene by bridging the micelle containing asphaltene molecules and saturates, by which dispersion of all components occurs in the crude oil. Resins are associated to asphaltene in the micelle forming layers avoiding the precipitation of asphaltenes. However, this protection is broken when changes in the medium take place or when crude oil is blended with lighter crude oils (Hongfu et al. 2002). When upgrading heavy crude oil in presence of catalysts, it is to be expected that resins do not form more asphaltene; instead, resins could be obtained from asphaltene decomposition, as well as more light ends. On the contrary, if noncatalytic processes are taking place, conversion of resins into asphaltenes is attained by elimination of aliphatic chains and aromatization reactions (Ortiz Moreno et al. 2014). Other reports have stated that the increase in lighter cuts is due mainly to transformation of residue fraction of heavy crude oils, such as vacuum and atmospheric residues (Alonso-Ramírez et al. 2020).

1.5.2

TBP Distillation

The true boiling point test consists in distilling the crude oil or hydrocarbon mixture in a fractionation column. Initially, distillation is carried out from the initial boiling point to around 210 ∘ C under atmospheric pressure. Then, partial vacuum is applied to distillation to avoid cracking of heavier compounds in crude oil or fractions at higher temperatures. Cuts are collected at specified temperature intervals and mass and density of each fraction are able to be measured. Conversion to volumetric yield is made using the mass and density data. The vapor temperature measured at reduced pressure is converted to atmospheric equivalent temperature (AET) and distillation still continues up to 400 ∘ C AET. TBP curve is plotted as mass or volume yield versus boiling temperature in terms of AET and its shape is dependent on composition of crude oil or fractions.

1.5 Composition

Table 1.2 SARA fractionation and colloidal instability index for different heavy and extra-heavy crude oils (fractions of saturates, aromatics, resins, and asphaltenes are expressed in wt%). Heavy/extra-heavy crude oil

Bachaquero Lagunillas Boscan Sur Mediano

API, ∘ Saturates Aromatics Resins Asphaltenes CII

References

9.00 25.00

33.00

29.00

13.00

0.61 Ocanto et al. (2009)

15.00 30.00

26.00

32.00

12.00

0.72 Ocanto et al. (2009)

8.00 10.00

23.00

48.00

19.00

0.41 Ocanto et al. (2009)

15.00 25.00

28.00

35.00

11.00

0.57 Ocanto et al. (2009)

Hamaca

8.00 11.00

19.00

57.00

13.00

0.32 Ocanto et al. (2009)

Merey 16

20.00 25.00

24.00

36.00

15.00

0.67 Ocanto et al. (2009)

CNS

21.00 21.00

27.00

37.00

15.00

0.56 Ocanto et al. (2009)

Mesa 30

20.00 44.00

25.00

21.00

10.00

1.17 Ocanto et al. (2009)

UD 672

21.00 22.00

30.00

44.00

4.00

0.35 Ocanto et al. (2009)

Furrial

21.00 35.00

24.00

32.00

9.00

0.79 Ocanto et al. (2009)

8.00 19.00

28.00

42.00

11.00

0.43 Marcano et al. (2011)

Hamaca

9.00 19.00

25.00

43.00

13.00

0.47 Marcano et al. (2011)

Boscan

10.30 12.00

36.00

38.00

14.00

0.35 Marcano et al. (2011)

Furrial

23.70 55.00

1.44 Marcano et al. (2011)

Carabobo

28.00

13.00

4.00

Korean VR

4.10

9.53

30.71

40.37

19.39

0.41 Park et al. (2022)

Athabascaa)

4.04

7.22

33.12

44.69

14.97

0.29 Danial-Fortain et al. (2010)

Urala)

9.58 11.88

46.80

36.65

4.67

0.20 Danial-Fortain et al. (2010)

Duria)

15.30 22.80

31.05

39.96

6.19

0.41 Danial-Fortain et al. (2010)

6.95 11.69

48.99

31.55

7.76

0.24 Danial-Fortain et al. (2010)

12.50 16.30

25.20

39.50

19.00

0.55 Alonso-Ramírez et al. (2020)

Cold Lake

10.71 20.74

39.20

24.81

15.25

0.56 Peramanu et al. (1999)

Wolf Lake

10.50 25.18

37.40

27.33

10.09

0.54 Greaves and Xia (2004)

Heavy crude oil A

8.67 17.24

38.6

32.66

11.50

0.40 Arciniegas and Babadagli (2014)

Heavy crude oil B

10.28 19.45

45.60

25.34

9.60

0.41 Arciniegas and Babadagli (2014)

Arabian Lighta) Ku-Maloob-Zaap

Athabasca

8.05 17.27

39.70

25.75

17.28

0.53 Peramanu et al. (1999)

13.95 22.41

50.65

26.95

7.67

0.39 Khansari et al. (2014)

Liaohe

9.54 20.43

22.05

54.52

3.00

0.31 Li et al. (2007)

Liaohe

9.87 20.43

22.05

48.22

9.30

0.42 Wu et al. (2006)

Liaohe

15.20 40.98

28.79

21.25

5.50

0.93 Shang et al. (2018)

Shengli

18.00 31.90

19.30

43.80

5.00

0.58 Dolbear et al. (1987)

Xinjiang

16.00 28.09

31.45

32.86

7.60

0.55 Zhao et al. (2018)

Xinjiang

19.69 50.69

30.58

14.81

3.92

1.20 Zhao et al. (2022)

Xinjiang

16.75 35.64

20.78

28.3

15.28

1.04 Zhao et al. (2022)

6.40 13.00

16.90

49.90

20.20

0.50 Taborda et al. (2017b)

13.00 14.79

32.39

40.51

12.31

0.37 Taborda et al. (2017a)

Lloydminster

Extra-heavy crude oil Heavy crude oil

(Continued)

15

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.2

(Continued)

Heavy/extra-heavy crude oil

API, ∘ Saturates Aromatics Resins Asphaltenes CII

Boca de Jaruco

14.07 17.10

40.40

22.90

19.60

0.58 Novikov et al. (2019)

Arta 4

16.20 13.71

56.83

19.19

10.28

0.32 Mohammad et al. (2012)

Azadegan

17.19

8.13

58.44

20.68

12.75

0.26 Taheri-Shakib et al. (2018)

West Paydar

18.53 14.10

58.10

18.30

9.50

0.31 Mozafari and Nasri (2017)

Heavy crude oil

15.82 10.49

9.00

64.12

16.39

0.37 Castro and Vazquez (2009)

Extra-heavy crude oil Maya

References

9.17 15.00

19.11

46.78

19.11

0.52 Castro and Vazquez (2009)

21.00 18.17

28.97

41.52

11.31

0.42 Rana et al. (2008)

36.25

29.44

21.29

0.52 Félix and Ancheyta (2019)

5.28

70.93

15.85

0.31 Murillo-Hernández et al. (2009)

Ku-Maloob-Zaap

11.97 13.02

Heavy crude oil

11.60

Aguacate

12.80 10.7

3.40

62.60

23.30

0.52 Duran Armas (2021)

Aguacate

12.00 26.16

21.27

28.28

23.30

0.99 Coronel-García et al. (2021)

Gulf of Mexico

16.40 32.50

21.80

31.30

14.30

0.88 Martínez-Palou et al. (2013)

Tatar

15.00 25.40

44.70

23.70

6.4

0.46 Yeletsky et al. (2020)

Ashalchinskoye

15.10 23.10

45.60

23.80

7.50

0.44 Yadykova and Ilyin (2022)

Ashalcha

15.50 26.33

39.55

27.37

6.75

0.48 Mukhamatdinov et al. (2021) 0.53 Félix et al. (2022)

7.94

Ashalcha

13.35 28.79

44.32

20.98

5.91

Heavy crude oil

12.60 20.31

38.81

29.72

10.94

0.46 Zou (2017)

a) Normalized values.

1.4 1.2 Colloidal instability index (CII)

16

Instability

1.0 0.8

Stability

0.6 0.4 0.2 0.0 0

5

10 15 API gravity (°)

20

25

Figure 1.3 API gravity for different heavy and extra-heavy crude oils and their corresponding values of colloidal instability index.

1.5 Composition 1.4

Colloidal instability index (CII)

Colloidal instability index (CII)

1.4 1.2 1.0 0.8 0.6 0.4

0.2

1.0 0.8 0.6 0.4 0.2 0.0

0.0

0

10

20 30 40 Saturates (wt%)

50

0

60

10

20 30 40 Aromatics (wt%)

50

60

1.4 Colloidal instability index (CII)

1.4 Colloidal instability index (CII)

1.2

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

Figure 1.4

20

40 Resins (wt%)

60

80

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

5

10 15 Asphaltenes (wt%)

20

25

CII values for the parent crude oils and their relationship with SARA fractions.

Standardized methods to carry out TBP distillation are disclosed in ASTM D2892 and ASTM D5236 being appropriate to crude oils, petroleum fractions, and condensates; however, the methods are not able to be used in light naphtha or fractions having initial boiling points greater than 400 ∘ C. Each TBP curve for crude oils is unique, and it is used for refiners to marketing purposes and crude oil characterization (Behrenbruch and Dedigama 2007). In the case of heavy crude oils, the TBP curves have steeper slopes (Dhankar et al. 2019). Based on reported data, TBP curves are depicted in Figure 1.5 for selected crude oils having different composition (Equinor 2021; Stratiev et al. 2010). As a reference, the Algerian condensate considered as a low sulfur condensate (68.3∘ API, Algeria) is shown at the top of the figure. Then, the TBP curves of West Texas Intermediate (40.8∘ API, USA) and Brent (38.3∘ API, North Sea) crude oils commonly used as reference in the marketing of crude oil are plotted. Cabinda (31.7∘ API, Angola) crude oil and its TBP curve is also plotted. Depending on the refiner, the definition of light, medium, heavy, or extra-heavy crude oil is attained; however, the following intervals may be accepted to define crude oils according to their API values: light crude oil (API higher than 31.1∘ ); medium or intermediate crude oils (API values among 22.3∘ to 31.1∘ ); heavy crude oils (API values between 22.3∘ and 10∘ ); extra-heavy crudes (API lower than 10∘ ). TBP plots for heavy crude oils such as Maya (21∘ to 22∘ API, Mexico), Hebron (20.4∘ API, Canada), Peregrino (13.5∘ API, Brazil), and Tia Juana (12.1∘ API, Venezuela) heavy crude oils are shown in Figure 1.5. It is observed that heavier crude oils have a lower slope in the mid-region of the curve compared with lighter crude oils, and the distilled yield is low having larger yields for atmospheric and vacuum residue. The initial boiling point (IBP) for heavy crude oils starts at higher temperature since refractory compounds are found in these crude oils. In summary, a description of some standardized methods to carry out the distillation of crude oils or fractions is disclosed. For example, the standard method ASTM D2892 is applied to stabilized crude oil with an initial boiling point of 150 ∘ C, while the final cut temperature is 400 ∘ C atmospheric equivalent temperature (AET). The fractionating column is considered to behave like

17

1 Properties of Heavy and Extra-Heavy Crude Oils

100

Algerian condensate WTI Tia Juana Maya Cabinda Brent Peregrino Hebron

90 Cumulative mass percent distillied

18

80 70 60 50 40 30 20 10 0 0

Figure 1.5

100

200

300 400 Boiling temperature (°C)

500

600

TBP curves for different crude oils.

14–18 theoretical plates with a reflux ratio of 5 : 1. During distillation, a plot of temperature against mass or volume distilled is obtained. The standard method ASTM D5236 discloses the procedure to distillate heavy crude oils, petroleum distillates, residues, etc., in a potstill with a low pressure drop entrainment separator. The initial boiling point for hydrocarbons is greater than 150 ∘ C, while the final boiling point commonly is 565 ∘ C depending on the heat sensitive samples. The recommended distillation method for crude oil having ending cuts of 400 ∘ C AET is the ASTM D2892 method; however, distillation curves obtained by these methods are not comparable among them. The standard method ASTM D86 is applied to carry out the atmospheric distillation of petroleum and petroleum products relating its composition with energy content and boiling range distribution. It is possible to know the tendency to form deposits that cause obstruction of pipelines with this curve. For this reason, the distillation yield at different temperatures is requested, especially the temperatures at 10 vol% (T10), 50 vol% (T50), and 90 vol% (T90) of distilled volume as well as the final boiling point. The standard method ASTM D1160 includes the determination of the range of boiling points for petroleum products at reduced pressures, which are vaporized partially or fully at a maximum liquid temperature of 400 ∘ C. A conversion of heavy crude oil distillation data from ASTM D1160 to ASTM D5236 has been reported elsewhere. For heavy crude oils, methods based on Daubert, Edmister-Okamoto with modified coefficients gave the best correlations to convert distillation data from ASTM D1160 to ASTM D5236. The method proposed by authors also yielded good correlation (Nikolaychuk et al. 2015). The standard method ASTM D2887 is applicable to petroleum, petroleum products and fractions having a distillation range between 55.5 and 538 ∘ C at atmospheric pressure. The analysis time is reduced to about eight minutes, and it is available to samples having vapor pressures sufficiently low to be handled at ambient temperature. The standard method ASTM D7169 is complementary of the ASTM D2887 because it is applied to samples that do not elute completely during simulated distillation by which it is used to determine the boiling point distribution up to 720 ∘ C. The maximum temperature corresponds to elution

1.6 Typical Heavy Crude Oils

Table 1.3

Typical methods to analyze the distillation behavior of crude oils.

Analysis type

Standard method

Boiling temperature interval (∘ C)

Atmospheric distillation of crude oil

ASTM D2892

150–400

Distillation of heavy hydrocarbon mixtures

ASTM D5236

150–400

Simulated distillation

ASTM D2887 ASTM D7169

55–538 Up to 720

Distillation

ASTM D86 ASTM D1160

150–350 Up to 400

of C100 and atmospheric residues, vacuum residues, among other heavy crude oils or cuts may be analyzed by this technique. Since capillary columns with thin films are present, incomplete separation of C4 –C8 is attained. The aforementioned methods are depicted in Table 1.3.

1.6

Typical Heavy Crude Oils

Heavy crude oils from different origins present similar ranges of properties because they are characterized as viscous liquids with high content of heteroatoms providing low API gravity values. Different reservoirs in North and South America, as well as Middle East countries, have similar properties, such as high sulfur content (Guo et al. 2016). However, although these heavy crude oils have similar properties, there are some exceptions, where the value of some properties is outside the established range. Hence, some heavy crude oils may be atypical based on the property of interest as discussed below.

1.6.1

Properties

Mexico produces different heavy and extra-heavy crude oils with diverse physical and chemical properties as summarized in Table 1.4. One of the most known heavy crude oils from Mexico is Maya, which is characterized by the high content of heteroatoms, such as sulfur and metals, as well as heavy fractions (asphaltenes). Conversely, this oil has low viscosity and density, provoking an elevated API gravity almost similar to intermediate crude oils. Another heavy crude oil with relatively low viscosity is the Gulf of Mexico heavy crude oil, which presents a small amount of asphaltenes and high content of saturates and sulfur. The Aguacate field heavy crude oil accounts for larger concentration of sulfur and metals due to the high content of Conradson carbon along with the resins content, causing the low amount of light fractions (aromatics and saturates) and high viscosity. The Ku-Maloob-Zaap is an extra-heavy crude oil characterized by the high viscosity, amount of metals, heteroatoms content (S, O, and N), molecular weight, and Conradson carbon together with high content of heavy fractions (asphaltenes and resins) and low H/C atomic ratio. Heavy and extra-heavy crude oils from China are characterized by the low content of sulfur and metals (Table 1.5). Xinjiang heavy crude oil has high values of H/C atomic ratio as well as O and N content. However, the low content of metals (especially V), Conradson carbon, sulfur, and light fractions (aromatics and saturates), as well as low pour point and viscosity values increase its API gravity. Shengli heavy crude oil has similar properties to Xinjiang oil since the larger amount of

19

20

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.4

Physical and chemical properties of Mexican heavy and extra-heavy crude oils.

Property

Maya

Gulf of Mexico

Aguacate

Ku-Maloob-Zaap

API gravity

19.43–21.97

12.50–16.40

12.00–15.82

9.17–11.97

83.01

Elemental analysis (wt%) C

84.30–86.90

79.96–84.28

H

8.30–10.40

10.28–10.64

9.66

O

0.50

0.01

1.52

N

0.30–0.52

0.29–0.75

0.35–0.51

0.54

S

3.51–4.70

4.40–5.56

5.02–5.74

5.27

1.45–1.56

1.39

268.80

415.00–506.68

H/C atomic ratio

1.47–1.54 Metals content (ppm)

V

204.00–413.00

Ni

36.00–83.00

63.35–97.20

81.00–97.82

4.98–5.67

4.24

5.12–5.18

378.50–486.00

486.00–507.80

17.15

17.75

Ni/V ratio MW (g/mol) Conradson carbon (wt%)

15.30 Dynamic viscosity (Pa s)

At 30 ∘ C At 50 ∘ C

0.70

13.31

0.23

0.51 Kinematic viscosity (cSt)

At 25 ∘ C

2984.97

At 50 ∘ C Pour point (∘ C)

43 233.00 2082.00–15 854.80

−30.00

−12.00

13.50

SARA fractions (wt%) As

11.31–25.20

14.30

15.85–23.3

19.11–21.29

Re

25.90–41.52

31.30

28.28–62.6

29.44–46.78

Ar

26.72–28.97

21.80

3.40–21.27

19.11–36.25

Sa

18.17–29.56

32.50

10.7–26.16

13.02–15.00

heteroatoms increases the resin fraction. Additionally, lower values of V/Ni ratios can be caused by the small asphaltenes content. Lunpola heavy crude oil is another crude oil having low sulfur and asphaltene contents as well as Conradson carbon. Nevertheless, this crude oil is defined for its high viscosity and resins amount. Depending on the extraction well, the Liaohe oil can be categorized as heavy or extra-heavy crude oil. This petroleum has elevated values of viscosity and molecular weight but low asphaltene contents, which provide larger amounts of resins and saturates. Moreover, the content of V is poor compared with the Ni content, which gives higher values of V/Ni ratio as the Chinese heavy crude oils aforementioned. Russia has a large reserve of heavy crude oils, and the properties of these oils can be observed in Table 1.6. Ashalcha heavy crude oil is characterized by a small amount of asphaltenes and

1.6 Typical Heavy Crude Oils

Table 1.5

Physical and chemical properties of Chinese heavy and extra-heavy crude oils.

Property

Xinjiang

Shengli

Lunpola

Liaohe

API gravity

16.00–20.45

13.72–20.71

16.98

9.54–15.20

84.58

83.96–86.15

Elemental analysis (wt%) C

80.70–84.75

H

11.90–13.20

9.84–13.70

10.95–13.25

O

1.51–3.47

1.26

1.27–2.22

N

0.35–1.15

0.44–1.71

0.69

0.38–0.96

S

0.15–0.46

0.28–4.37

0.24

0.34–0.45

1.71–1.95

1.39–2.01

H/C atomic ratio

81.20–85.50

1.54–1.88

Metals content (ppm) V

0.20

3.40

Ni

13.90

42.30–47.60

46.80–125.00

0.01

0.08

0.02–0.03

5.40

7.50–9.70

Ni/V ratio

1.91–2.10

MW (g/mol) Conradson carbon (wt%)

482.00–624.00 2.70

9.00

Dynamic viscosity (Pa s) At 40 ∘ C At 50 ∘ C

17.42

At 60 ∘ C At 80 ∘ C

0.29–5.04

0.27

1.11 0.27–175.00

139.80–158.00 88.50–124.30

Kinematic viscosity (cSt) At 50 ∘ C

23 455.70

At 80 ∘ C Pour point (∘ C)

3661.00 −22.00

4.00–14.00

SARA fractions (wt%) As

3.92–15.28

0.70–5.00

0.48

3.00–9.30

Re

14.81–32.86

43.80–44.63

41.42

21.25–54.52

Ar

20.78–31.45

19.30–29.06

22.05–38.50

Sa

28.09–50.69

23.75–31.90

20.43–40.98

low molecular weight. The wide range of V/Ni ratio is due to some samples having insignificant content of V. In addition, this heavy crude oil has a high amount of aromatics hydrocarbons and dynamic viscosity values, despite the low content of high molecular weight fractions (asphaltenes). Another heavy crude oil is Usinsk having low amount of asphaltenes as well as low molecular weight. Moreover, this oil is characterized by its decreased value of pour point. Yarega heavy crude oil displays low pour point and sulfur content, high H/C atomic ratio but large content of metals (V and Ni). Whereas the Mordovo–Karmalskoye oil exhibits larger content of heteroatoms

21

22

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.6

Physical and chemical properties of Russian heavy and extra-heavy crude oils.

Property

Yarega

Mordovo–Karmalskoye

Ashalcha

Usinsk

API gravity

18.17–19.00

15.90

13.35–15.10

14.87

Elemental analysis (wt%) C

86.29

81.50

82.64–83.88

84.94

H

12.72

11.60

11.21–12.10

11.98

O

0.16

2.10

0.12–1.96

N

0.04

1.10

0.29–0.70

0.63

S

0.79–1.24

3.65–3.70

3.20–4.52

1.98

1.76

1.70–1.71

1.61–1.72

1.68

H/C atomic ratio

0.47

Metals content (ppm) V

160.00

1.26–200.00

Ni

47.00

10.00–60.00

Ni/V ratio

3.40

0.13–3.33

MW (g/mol)

452.00

385.12

Conradson carbon (wt%)

365.00

9.70 Dynamic viscosity (Pa s)

At 20 ∘ C

3.31 Kinematic viscosity (cSt)

At 20 ∘ C At 25 ∘ C Pour point (∘ C)

1609.00

3952.39 282 500.00

−18.00

−22.50 SARA fractions (wt%)

As

3.00–17.00

5.20

5.91–7.50

Re

20.00–32.00

24.50

20.98–27.37

Ar

35.00

45.40

39.55–45.60

Sa

16.00

24.90

23.10–28.79

8.10 18.00

(S, N, and O). However, this heavy crude oil presents low values of viscosity and asphaltenes content, together with high aromatics content and H/C atomic ratio. Canadian heavy and extra-heavy crude oils are similar to Mexican oils, having high content of sulfur, metals, and an elevated viscosity value, as observed in Table 1.7. Lloydminster heavy crude oil commonly presents low asphaltenes content and high aromatics fractions besides the high values of the viscosity, H/C atomic ratio, and low sulfur content. The Cold Lake heavy crude oil displays a high molecular weight value as well as high aromatics and sulfur content, whereas the Wolf Lake heavy crude oil exhibits properties similar to Lloydminster oil (asphaltenes, aromatics, sulfur content, and H/C atomic ratio) also with high content of metals. Finally, Athabasca bitumen is an extra-heavy crude oil with remarkably high viscosity and elevated content of heteroatoms, metals, Conradson carbon, and aromatics. Additionally, the high molecular weight value and the aforementioned properties give decreased value of API gravity.

1.6 Typical Heavy Crude Oils

Table 1.7

Physical and chemical properties of Canadian heavy and extra-heavy crude oils.

Property

Lloydminster

Cold Lake

API gravity

10.90–13.95

10.71

Wolf Lake

Athabasca bitumen

10.50

8.05–11.00

83.62–84.00

83.70

83.20–83.34

10.00–10.50

10.62

Elemental analysis (wt%) C

82.30–83.70

H

10.60–11.80

O

0.20–0.86

9.70–10.26 1.08–1.70

N

0.20–0.40

0.40–0.45

0.25

0.40–0.53

S

3.40–4.40

4.56–5.10

4.50

4.64–5.30

1.53–1.68

1.42–1.50

1.51

1.39–1.52

H/C atomic ratio

Metals content (ppm) V

192.00

247.00

Ni

75.00

93.00

2.56

2.66

Ni/V ratio MW (g/mol)

440.00

550.00

557.00

Conradson carbon (wt%)

12.00–12.32 Dynamic viscosity (Pa s)

at 20 ∘ C at 30 ∘ C

13.44–14.60

at 40 ∘ C at 50 ∘ C

0.81–2.27

5.26

581.00 32.40 SARA fractions (wt%)

As

7.12

15.25

10.09

11.67–18.60

Re

25.03

24.81

27.33

16.80–25.75

Ar

47.04

39.20

37.40

39.70–48.50

Sa

20.81

20.74

25.18

16.10–17.27

The low asphaltene content and high pour point value are properties commonly found in heavy and extra-heavy crude oils from the USA, as summarized in Table 1.8. Hondo heavy crude oil is characterized by its high sulfur, metals, and resins content and present an increased H/C atomic ratio. For the Alba and Thums heavy crude oils, the low asphaltenes content, high H/C atomic ratio, and small value of viscosity are similar properties in these oils, whereas Kern River and Coalinga heavy crude oils display low values in the S, V, and asphaltenes contents. For the pour point and Ni content, the values are high, causing a low V/Ni ratio. The Cymric and Midway Sunset heavy crude oils show alike physicochemical properties: small amount of asphaltenes, saturates, and sulfur, elevated value of pour point and similar content of V and Ni giving the V/Ni ratio close to 1. Venezuela is a country with large reserves of heavy crude oil that have different properties, as shown in Table 1.9. Tia Juana is a heavy crude oil with high Conradson carbon and V contents, but low viscosity, pour point, and asphaltenes fraction values. Boscan is a heavy crude oil rich in

23

24

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.8

Physical and chemical properties of heavy and extra-heavy crude oils from USA.

Property

Hondo

Alba

Thums

API gravity

13.40–19.35 19.03 8.60–17.13

Kern River

San Joaquin Valley Cymric

Midway Sunset

9.70–14.50

13.04

10.40– 10.50

8.70–11.70

Coalinga

9.70–10.30

Elemental analysis (wt%) C H O N

0.70–0.73

0.72–0.88

0.84–0.91

0.89–0.91 0.73–0.79

S

5.10

1.00

1.40–1.49

1.72–1.75 0.87–0.93

H/C atomic ratio

1.68

1.65 1.69–1.70

1.52

1.60–1.72

Metals content (ppm) V

280.00

25.00–75.00

65.00–73.00 105.00– 110.00

10.00

Ni

92.00

65.00–110.00

65.00–69.00 95.00

29.00–33.00

Ni/V ratio

3.04

0.38–0.68

1.00–1.06

1.11–1.16 0.30–0.34

Conradson carbon (wt%)

10.80

330.00–3260.00

800.00– 6420.00

3200.00– 3670.00

20.00–50.00

20.00–60.00 60.00

60.00

6.00–8.00

5.00

Dynamic viscosity (Pa s) At 40 ∘ C

0.36

0.14 0.15–0.66

1.39

Kinematic viscosity (cSt) At 50 ∘ C Pour point (∘ C)

2830.00– 2950.00

SARA fractions (wt%) As

13.90–14.80

Re

20.50–40.20 10.10 12.50–18.70

1.64 3.31–5.09

4.00–7.00

4.56

9.00

19.40

Ar

25.00–26.00

27.00–28.00 24.00

32.00–33.00

Sa

19.00–21.00

16.00–19.00 16.00

21.00–22.00

metals and sulfur content in which the great amount of V provides a high V/Ni ratio, while Hamaca and Cerro Negro heavy crude oils present similar properties where their sulfur and metals content, Conradson carbon, and viscosity are quite high. Heavy crude oils from Middle East countries are characterized by low asphaltenes fraction and high metals content, as seen in Table 1.10. Arta-4 heavy crude oil from Egypt displays high content of sulfur, metals, and aromatics, as well as high values of viscosity. The Iranian heavy crude oils (Azadegan, Gach Saran, and West-Paydar) have not only similar properties such as low viscosity, asphaltenes, and saturates contents, but also large aromatic fraction. Qayarah heavy crude oil from Iraq has extremely high sulfur content besides high amount of asphaltenes and Conradson carbon. Heavy crude oil from Kuwait presents a low H/C atomic ratio, small amounts of metals,

1.6 Typical Heavy Crude Oils

Table 1.9

Physical and chemical properties of Venezuelan heavy and extra-heavy crude oils.

Property

Tia Juana

Boscan

Hamaca

Cerro Negro

API gravity

12.58

10.10

8.50–9.10

8.88–8.90

Elemental analysis (wt%) C

83.93

H

9.58

O

1.42

N

0.30

0.44

0.75–0.89

0.75

S

2.50

5.66–5.70

3.75–3.78

3.99–4.00

H/C atomic ratio

1.36–1.37 Metals content (ppm)

V

397.00

Ni Ni/V ratio

1220.00

412.00–488.00

430.00

147.00

91.90–105.00

108.60

8.30

4.41–4.65

3.96

15.00

14.20–15.80

15.20

MW (g/mol) Conradson carbon (wt%)

12.30

Dynamic viscosity (Pa s) At 30 ∘ C At 40 ∘ C

500.00 0.89

At 60 ∘ C At 80 ∘ C Pour point (∘ C)

1.79–7.70 1.55

−1.00

10.00

6.10

4.91–15.50

1.81

1.97

27.00

SARA fractions (wt%) As

7.50

15.20–24.20

10.20–23.50

10.10–19.90

Re Ar Sa

and asphaltenes as high molecular weight and sulfur content values. Two heavy crude oils from Saudi Arabia present diverse properties since the Heavy Arabian oil has low viscosity, asphaltenes, and resins values, while the other heavy crude oil presents high content of metals, sulfur, and Conradson carbon. Latin American countries also commercialize a variety of heavy and extra-heavy crude oils (Table 1.11). Marlim heavy crude oil from Brazil is characterized by low impurities content (metals, sulfur, and asphaltenes) and low viscosity and pour point values which is suitable for refining. The Castilla and one extra-heavy crude oils from Colombia have a low H/C atomic ratio, larger amount of resins, and sulfur content as well as high values of viscosity but the extra-heavy crude oil has higher values, especially for sulfur. Whereas Boca de Jaruco heavy crude oil from Cuba presents high impurities (sulfur and metals) and aromatics content together with a high H/C atomic ratio. Ecuador has the Napo heavy crude oil, which displays a low pour point and small sulfur amount. In Table 1.12, other heavy crude oils from different countries around the world are

25

26

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.10 countries.

Physical and chemical properties of heavy and extra-heavy crude oils from Middle East

Country

Property

API gravity

Egypt

Iran

Iran

Iran

Iraq

Arta-4

Azadegan

Gach Saran

WestPaydar

Qayarah

18.53

15.28

15.60

Kuwait Heavy crude oil

16.20

17.19

12.20

C

80.55

85.15

86.70

10.11

8.20

Saudi Arabia

Saudi Arabia

Heavy crude oil

Heavy Arabian

12.60

18.08

Elemental analysis (wt%) H

10.63

O

3.04

N

0.34

1.53

0.41

S

4.50

3.21

2.60

1.57

1.41

H/C atomic ratio

8.40

0.50

0.26

4.60

4.23

1.13 Metals content (ppm)

V

183.97

108.00

Ni

113.05

36.00

1.63

3.00

Ni/V ratio

20.00

28.00 3.11

MW (g/mol) Conradson carbon (wt%)

87.00

411.00 11.34

8.80

15.60

12.60

Dynamic viscosity (Pa s) At 20 ∘ C At 30 ∘ C

0.88

3.90 1.50

At 40 ∘ C

0.67

0.03

Kinematic viscosity (cSt) At 25 ∘ C At 50 ∘ C At 80 ∘ C Pour point (∘ C)

76.87

2500.00 343.00

196.39 12.00 SARA fractions (wt%)

As

10.28

12.75

6.80

9.50

Re

19.19

20.68

28.50

18.30

Ar

56.83

58.44

58.10

Sa

13.71

8.13

14.10

20.40

8.00

12.60

6.68

27.50

7.46

shown. Van Gogh and Doba heavy crude oils from Australia and Chad, respectively, show similar properties since both have low contents of impurities (S, N, Ni, and Conradson carbon content) and decreased values of viscosity and pour point, while the Germany heavy crude oil presents a high H/C atomic ratio and low content of asphaltenes.

1.6 Typical Heavy Crude Oils

Table 1.11 countries.

Physical and chemical properties of heavy and extra-heavy crude oils from Latin American

Country Brazil Property

API gravity

Marlim

19.20

Colombia

Colombia

Cuba

Ecuador

Castilla

Extra-heavy crude oil

Boca de Jaruco

Napo

13.00–13.40

6.40

14.07

17.40

Elemental analysis (wt%) C

80.40

84.80

75.47

H

9.46

7.50

10.12

O

8.40

N

0.49

0.47–2.30

0.88

0.41

S

0.78

2.16–4.84

6.82

5.29–5.60

1.40

1.05

1.60

H/C atomic ratio

2.18

Metals content (ppm) V

25.00

311.70

76.00

Ni

20.00

78.20

26.00

1.25

3.99

2.92

Ni/V ratio Conradson carbon (wt%)

15.17 Dynamic viscosity (Pa s)

At 20 ∘ C

271.00 Kinematic viscosity (cSt)

At 20 ∘ C At 50 ∘ C

544.60 971.90

At 80 ∘ C Pour point (∘ C)

370.63

144.60 −39.00

−7.22 SARA fractions (wt%)

As

2.60

Re Ar Sa

1.6.2

42.00

12.31–15.50

20.20

19.60

40.51

49.90

22.90

32.39

16.90

40.40

14.79

13.00

17.10

12.06

Relationship Between Properties

The physical and chemical properties of all these heavy and extra-heavy crude oils present some relationships between them. The API gravity has different correlations with physical (viscosity and pour point) and chemical (H/C atomic ratio, molecular weight, and Conradson carbon) properties as observed in Figure 1.6. The higher the H/C ratio, the higher the API gravity. Higher value of H/C relationship is particularly important when upgrading heavy crude oils due to less amount of hydrogen that is required if hydrogenolysis reactions are carried out such as hydrodesulfurization.

27

28

1 Properties of Heavy and Extra-Heavy Crude Oils

Table 1.12 Physical and chemical properties of heavy and extra-heavy crude oils from different countries. Country

Property

API gravity

Australia

Chad

Germany

Van Gogh

Doba

Heavy crude oil

18.80

13.03

17.10

Elemental analysis (wt%) C

86.25

86.00

H

12.10

12.50

O

1.16

N

0.184

0.25

S

0.38

0.14

1.25

1.67

1.73

H/C atomic ratio Metals content (ppm) Ni

1.70

Conradson carbon (wt%)

5.66

Dynamic viscosity (Pa s) At 30 ∘ C

0.56

At 40 ∘ C At 50 ∘ C

0.28

At 60 ∘ C

0.11

Pour point (∘ C)

0.239

0.16 −17.50 SARA fractions (wt%)

As

1.50

Re

18.90

Ar

34.20

Sa

22.50

The API gravity is inversely proportional to density, which decreases as the light fraction content like saturates increases. However, some heavy crude oils (Shengli, Thums, and Azadegan) do not follow this trend. The saturates content is in agreement with the H/C atomic ratio due to it increases as API gravity increments for the same reason, except for the Azadegan and West-Paydar heavy crude oils, which have low amount of saturates at relatively elevated API gravity. On the contrary, the dynamic viscosity and pour point diminished as the API gravity increased as expected because lighter crude oils have improved mobility; however, Thums, Xinjiang, and Boscan heavy crude oils do not follow this trend. Other properties vary depending on the API gravity, i.e. the higher the API gravity value, the lower the Conradson carbon and the molecular weight values as a consequence of heavier fractions such as asphaltenes and resins are less abundant in crude oils. The asphaltenes content is an important chemical property for heavy and extra-heavy crude oils since this fraction represents the main issue during crude oil upgrading. Thus, understanding the relationship of this fraction with other properties (Figure 1.7) is essential to screen the quality of

1.6 Typical Heavy Crude Oils 700

Molecular weight (g/mol)

H/C atomic ratio

2.2 1.9 1.6 1.3 1.0 8

10

12 14 API gravity

16

18

400 300

20

6

Dynamic viscosity (Pa s) at 40 °C

20 Conradson carbon (wt%)

500

200 6

16 12 8 4 0 6

8

10

12 14 API gravity

16

18

16

18

20

0.10 0.01 6

60

50

50

−10

12 14 API gravity

1.00

70

10

10

10.00

20

30

8

100.00

Saturates (wt%)

Pour point (°C)

600

8

10

12 14 API gravity

16

18

20

18

20

40 30 20 10

−30

0 6

8

10

12 14 API gravity

16

18

20

6

8

10

12 14 API gravity

16

Figure 1.6 Correlations between the API gravity and other properties of different heavy and extra-heavy crude oils.

the oil preliminary. Like the API gravity does, the asphaltene fraction behaves as a function of the H/C atomic ratio and saturate fraction whose values decrease as the content of asphaltenes increases. The H/C atomic ratio varies inversely with the asphaltene content because larger values of H/C ratio imply higher content of aromatic molecules commonly contained in asphaltenes as aromatic ring clusters. Therefore, increasing the amount of asphaltene fraction heightens the aromatic nature of the oil, reaching values of H/C atomic ratio similar to toluene. In addition, values of Conradson carbon and dynamic viscosity growth as asphaltenes content increased owing to asphaltenes are carbon residue producer. Its content also influences the increase of dynamic viscosity. Other values such as V/Ni ratio also depend on the amount of asphaltene. The higher the amount of asphaltenes, the higher the amount of metals because most of them are contained in asphaltene molecules as porphyrins. Vanadium is commonly more abundant than nickel with exception of Chinese and Ashalcha heavy crude oils whose Ni content is higher than V. Sulfur content follows the same trend like V/Ni ratio because most of sulfur species are contained in resins and asphaltene fractions.

29

1 Properties of Heavy and Extra-Heavy Crude Oils 10

2.2 Sulfur content (wt%)

H/C atomic ratio

2.0 1.8 1.6 1.4 1.2 1.0 5

10 15 Asphaltenes (wt%)

20

6 4 2

25

0

5

10 15 20 Asphaltenes (wt%)

25

30

0

5

10 15 20 Asphaltenes (wt%)

25

30

20 Conradson carbon (wt%)

9

V/Ni ratio

8

0 0

6

3

0

16 12 8 4 0

0

5

10 15 20 Asphaltenes (wt%)

25

30

100.00

60 50

10.00

Saturates (wt%)

Dynamic viscosity (Pa.s) at 40 °C

30

1.00 0.10

40 30 20 10

0.01

0 0

4

8 12 Asphaltenes (wt%)

16

20

0

5

10 15 Asphaltenes (wt%)

20

25

Figure 1.7 Correlations between the asphaltenes content (wt%) and other properties of different heavy and extra-heavy crude oils.

The values of Conradson carbon behave similarly to asphaltene content, as depicted in Figure 1.8. The Conradson carbon content increases as the H/C atomic ratio and saturate fraction decrease. On the contrary, values of dynamic viscosity, sulfur content, and V/Ni ratio increased. The aforementioned behavior is due to the fact that Conradson carbon depends directly on asphaltene fraction that contains most of the impurities (sulfur, metals). Moreover, the relations between SARA fractions of some heavy and extra heavy crude oils are correlated with the Conradson carbon owing to the asphaltenes-to-resins (As/Re) ratio, the asphaltenes-to-aromatics (As/Ar) ratio, and the asphaltenes-to-saturates (As/Sa) ratio also increased as the Conradson carbon content is larger as observed. The relationships between the H/C atomic ratios with other properties are observed in Figure 1.9. The lower the sulfur content, the higher the H/C atomic ratio, which implies more aromatic crude oils as H/C ratio diminishes because sulfur species are mainly aromatic-based compounds. The saturate fraction increases as the H/C atomic ratio is larger; however, derived from SARA analysis,

10

1.8 1.6 1.4 1.2

6 4

4

8 12 16 Conradson carbon (wt%)

20

Saturates (wt%)

100.0

10.0

1.0

4

8 12 16 Conradson carbon (wt%)

1.0

40

0.8

30 20 10

8 12 16 Conradson carbon (wt%)

20

8 12 16 Conradson carbon (wt%)

20

8 12 16 Conradson carbon (wt%)

20

0.6 0.4 0.2 0.0

4

8 12 16 Conradson carbon (wt%)

0.8

2.0

0.6

1.5

20

4

As/Sa ratio

As/Ar ratio

4

4

50

0.4

1.0

0.2

0.5

0.0

0.0 4

Figure 1.8

20

0

0.1

3

0 0

As/Re ratio

0

6

2 0

1.0

Dynamic viscosity (Pa.s) at 40 °C

9

8 V/Ni ratio

2.0

Sulfur content (wt%)

H/C atomic ratio

2.2

8 12 16 Conradson carbon (wt%)

20

4

8 12 16 Conradson carbon (wt%)

20

Correlations between the Conradson carbon (wt%) and other properties of different heavy and extra-heavy crude oils.

8

40

6

30

Saturates (wt%)

Sulfur content (wt%)

1 Properties of Heavy and Extra-Heavy Crude Oils

4 2 0

20 10 0

1.0

1.2

1.4 1.6 1.8 H/C atomic ratio

2.0

2.2

1.2

1.5

3.0

1.2

2.5 Re/Sa ratio

Coloidal instability index (CII)

32

0.9 0.6 0.3

1.4

1.6 1.8 H/C atomic ratio

2.0

2.2

1.4

1.6 1.8 H/C atomic ratio

2.0

2.2

2.0 1.5 1.0 0.5

0.0

0.0 1.2

1.4

1.6 1.8 H/C atomic ratio

2.0

2.2

1.2

Figure 1.9 Correlations between the H/C atomic ratio and other properties of different heavy and extra-heavy crude oils.

the Re/Sa ratio decreases as the H/C atomic ratio increments. This behavior is owing to fact that the saturate fraction turns the crude oil into more aliphatic by increasing the H/C atomic ratio. Furthermore, the CII values heighten when the H/C atomic ratio increases because the aromatic fractions (resins and aromatics) have less content. The sulfur content is another relevant property to determine the quality of heavy and extra-heavy crude oils, and different properties are dependent on its content, as observed in Figure 1.10. The V/Ni ratio increases as the sulfur content does, indicating that the sulfur and V compounds are present in heavy fractions like asphaltenes. It was observed from Figure 1.7 that the higher the amount of asphaltenes, the higher the V/Ni ratio and sulfur content since it is expected that larger amounts of sulfur and metals are contained in the heaviest fraction. However, the Ashalcha and Arta-4 heavy crude oils are the exceptions to this behavior owing to the similar amount of Ni and V in these crude oils besides having high sulfur content. On the basis of this behavior, the As/Re and As/Sa ratios heighten as the sulfur content increases. Some heavy crude oils (Shengli and Ashalcha) do not follow this trend because of their low content of asphaltenes and high amount of sulfur. Another interesting value related to molecular weight of the crude oil is the colloidal instability index, which tends to increase as molecular weight does (Figure 1.11). This behavior is due to when the relatively low molecular weight aromatics (resins and aromatics) diminish their content, the asphaltenes fraction having the highest molecular weight compounds tend to increase its content, while the CII heightens and so does the molecular weight. Nonetheless, there are some exceptions to this behavior, such as the Ku-Maloob-Zaap heavy crude oil, which has high molecular weight and high content of aromatics and resins with low saturates content by which its CII value is low.

1.6 Typical Heavy Crude Oils

Figure 1.10 Correlations between the sulfur content and As/Re ratio, As/Sa ratio, and V/Ni ratio of different heavy and extra-heavy crude oils.

6

V/Ni ratio

5 4 3 2 1 0 0

1

2 3 4 Sulfur content (wt%)

5

6

1.2

As/Re ratio

0.9 0.6 0.3 0.0 0

1

2 3 4 Sulfur content (wt%)

5

6

0

1

2 3 4 Sulfur content (wt%)

5

6

2.4

As/Sa ratio

2.0 1.6 1.2 0.8 0.4

Figure 1.11 Correlations between the molecular weight and CII of different heavy and extra-heavy crude oils.

Coloidal instability index (CII)

0.0

0.7 0.6 0.5 0.4 0.3 0.2 300

400

500 600 Molecular weight

700

33

34

1 Properties of Heavy and Extra-Heavy Crude Oils

1.7

Concluding Remarks

The main properties and composition of heavy and extra-heavy crude oils were disclosed in this chapter. The characterization of petroleum provides a hint of its quality and knowledge about the difficulty during its transportation and upgrading in refineries. Most of heavy and extra-heavy crude oils analyzed have similar properties since they have similar geological origins; however, differences arise in some crude oils. Diverse properties summarized in this chapter correlate well among them, such as API gravity, asphaltene content, sulfur content, Conradson carbon, metals content, viscosity, H/C atomic ratio, and SARA fractions. All these relationships are justified on the basis of the current knowledge in the literature related to heavy and extra-heavy crude oils. Nevertheless, some exceptions were observed since a few number of heavy and extra-heavy crude oils presented some uncommon properties as expected according to their origin.

References Abdel-Aal, H.K., Aggour, M.A., and Fahim, M.A. (2003). Petroleum and Gas Field Processing, 1e. Boca Raton: CRC Press. Ali, S.I., Lalji, S.M., Haneef, J. et al. (2021). Comprehensive analysis of asphaltene stability predictors under different conditions. Petroleum Chemistry 61: 446–454. Alonso-Ramírez, G., Cuevas-García, R., Sanchez-Minero, F. et al. (2020). Catalytic hydrocracking of a Mexican heavy oil on a MoS2 /Al2 O3 catalyst: I. Study of the transformation of isolated saturates fraction obtained from SARA analysis. Catalysis Today 353: 153–162. Alonso-Ramírez, G., Cuevas-García, R., Sánchez-Minero, F. et al. (2021). Catalytic hydrocracking of a Mexican heavy oil on a MoS2 /Al2 O3 catalyst: II. Study of the transformation of isolated aromatics fraction obtained from SARA analysis. Fuel 288: 119541. Álvarez, P., Browning, B., Jansen, T. et al. (2019). Modeling of atmospheric and vacuum petroleum residue hydroconversion in a slurry semi-batch reactor: study of hydrogen consumption. Fuel Processing Technology 185: 68–78. Ancheyta, J. and Speight, J.G. (2007). Hydroprocessing of Heavy Oils and Residua. CRC Press. Ancheyta, J., Trejo, F., and Rana, M.S. (2009). Asphaltenes: Chemical Transformation During Hydroprocessing of Heavy Oils, 1e. CRC Press/Taylor & Francis. Arciniegas, L.M. and Babadagli, T. (2014). Asphaltene precipitation, flocculation and deposition during solvent injection at elevated temperatures for heavy oil recovery. Fuel 124: 202–211. Ashoori, S., Sharifi, M., Masoumi, M., and Mohammad Salehi, M. (2017). The relationship between SARA fractions and crude oil stability. Egyptian Journal of Petroleum 26: 209–213. Azinfar, B., Zirrahi, M., Hassanzadeh, H., and Abedi, J. (2018). Characterization of heavy crude oils and residues using combined Gel Permeation Chromatography and simulated distillation. Fuel 233: 885–893. Behrenbruch, P. and Dedigama, T. (2007). Classification and characterisation of crude oils based on distillation properties. Journal of Petroleum Science and Engineering 57: 166–180. Castro, L.V. and Vazquez, F. (2009). Fractionation and characterization of mexican crude oils. Energy and Fuels 23: 1603–1609. Caumette, G., Lienemann, C.P., Merdrignac, I. et al. (2009). Element speciation analysis of petroleum and related materials. Journal of Analytical Atomic Spectrometry 24: 263–276. Chacón-Patiño, M.L., Nelson, J., Rogel, E. et al. (2022). Vanadium and nickel distributions in selective-separated n-heptane asphaltenes of heavy crude oils. Fuel 312: 122939.

References

Coronel-García, M.A., Reyes de la Torre, A.I., Domínguez-Esquivel, J.M. et al. (2021). Heavy oil hydrocracking kinetics with nano-nickel dispersed in PEG300 as slurry phase catalyst using batch reactor. Fuel 283: 118930. Danial-Fortain, P., Gauthier, T., Merdrignac, I., and Budzinski, H. (2010). Reactivity study of Athabasca vacuum residue in hydroconversion conditions. Catalysis Today 150: 255–263. Dehkissia, S., Larachi, F., Rodrigue, D., and Chornet, E. (2004). Characterization of Doba–Chad heavy crude oil in relation with the feasibility of pipeline transportation. Fuel 83: 2157–2168. Dettman, H., Inman, A., Salmon, S. et al. (2005). Chemical characterization of GPC fractions of Athabasca bitumen asphaltenes isolated before and after thermal treatment. Energy and Fuels 19: 1399–1404. Dhankar, S., Mohanta, H.K., Neogi, A., and Sande, P.C. (2019). Method to predict complete product fraction TBP distributions from that of the whole crude using regression techniques: applied to shale oil. Petroleum and Coal 61: 1089–1101. Dolbear, G.E., Tang, A., and Moorehead, E.L. (1987). Upgrading studies with Californian, Mexican and Middle Eastern heavy oils. Fuel 66: 267–270. Duarte, L.M., Filgueiras, P.R., Silva, S.R.C. et al. (2016). Determination of some physicochemical properties in Brazilian crude oil by 1H NMR spectroscopy associated to chemometric approach. Fuel 181: 660–669. Duran Armas, J.L. (2021). Application of in-situ upgrading in naturally fractured reservoirs. PhD Thesis. University of Calgary. Equinor (2021). Crude oil assays. https://www.equinor.com/energy/crude-oil-assays (accessed 8 August 2022). Félix, G. and Ancheyta, J. (2019). Comparison of hydrocracking kinetic models based on SARA fractions obtained in slurry-phase reactor. Fuel 241: 495–505. Félix, G., Tirado, A., Al-Muntaser, A. et al. (2022). SARA-based kinetic model for non-catalytic aquathermolysis of heavy crude oil. Journal of Petroleum Science and Engineering 216: 110845. Ganeeva, Y., Barskaya, E., Okhotnikova, E. et al. (2021). Asphaltenes of crude oils and bitumens: the similarities and differences. Petroleum Science and Technology 40: 734–750. Gao, Y.Y., Shen, B.X., and Liu, J.C. (2013). Distribution of nickel and vanadium in Venezuela crude oil. Petroleum Science and Technology 31: 509–515. Garcia-Montoto, V., Verdier, S., Maroun, Z. et al. (2020). Understanding the removal of V, Ni and S in crude oil atmospheric residue hydrodemetallization and hydrodesulfurization. Fuel Processing Technology 201: 106341. Gray, M.R. (2015). Upgrading Oilsands Bitumen and Heavy Oil, 1e. Alberta: The University of Alberta Press. Greaves, M. and Xia, T.X. (2004). Downhole catalytic process for upgrading heavy oil: produced oil properties and composition. Journal of Canadian Petroleum Technology 43: 25–30. Guo, K., Li, H., and Yu, Z. (2016). In-situ heavy and extra-heavy oil recovery: a review. Fuel 185: 886–902. Guzmán, R., Ancheyta, J., Trejo, F., and Rodríguez, S. (2017). Methods for determining asphaltene stability in crude oils. Fuel 188: 530–543. Hongfu, F., Yongjian, L., Liying, Z., and Xiaofei, Z. (2002). The study on composition changes of heavy oils during steam stimulation processes. Fuel 81: 1733–1738. Huc, A.-Y. (2010). Heavy Crude Oils: From Geology to Upgrading: An Overview, 1e. Paris: Editions Technip. Kayukova, G.P., Foss, L.E., Feoktistov, D.A. et al. (2017). Transformations of hydrocarbons of Ashal’hinskoe heavy oil under catalytic aquathermolysis conditions. Petroleum Chemistry 57: 657–665.

35

36

1 Properties of Heavy and Extra-Heavy Crude Oils

Khansari, Z., Gates, I.D., and Mahinpey, N. (2014). Low-temperature oxidation of Lloydminster heavy oil: kinetic study and product sequence estimation. Fuel 115: 534–538. Kim, J.G., Kim, J.H., Song, B.J. et al. (2016). Empirical approach to determine molecular weight distribution using MALDI-TOF analysis of petroleum-based heavy oil. Fuel 186: 20–23. Kumar, R., Voolapalli, R.K., and Upadhyayula, S. (2018). Prediction of crude oil blends compatibility and blend optimization for increasing heavy oil processing. Fuel Processing Technology 177: 309–327. Leontaritis, K.J. and Mansoori, G.A. (1989). Fast crude-oil heavy-component characterization using combination of ASTM, HPLC, and GPC methods. Journal of Petroleum Science and Engineering 2: 1–12. Leyva, C., Ancheyta, J., Berrueco, C., and Millán, M. (2013). Chemical characterization of asphaltenes from various crude oils. Fuel Processing Technology 106: 734–738. Li, W., Zhu, J. hua, and Qi J. hua (2007). Application of nano-nickel catalyst in the viscosity reduction of Liaohe extra-heavy oil by aqua-thermolysis. Journal of Fuel Chemistry and Technology 35: 176–180. Li, T., Xu, J., Zou, R. et al. (2018). Resin from Liaohe heavy oil: molecular structure, aggregation behavior, and effect on oil viscosity. Energy and Fuels 32: 306–313. Li, H., Gao, H., Zhao, X. et al. (2022). Experimental study on viscosity reduction of heavy oil with water content by synergistic effect of microwave and nano-catalyst. Journal of Petroleum Science and Engineering 208: 109271. Marcano, F., Flores, R., Chirinos, J., and Ranaudo, M.A. (2011). Distribution of Ni and V in A1 and A2 asphaltene fractions in stable and unstable Venezuelan crude oils. Energy and Fuels 25: 2137–2141. Martínez-Palou, R., Cerón-Camacho, R., Chávez, B. et al. (2013). Demulsification of heavy crude oil-in-water emulsions: a comparative study between microwave and thermal heating. Fuel 113: 407–414. Medina, O.E., Gallego, J., Cespedes, S. et al. (2022). Effect of pressure on thermo-oxidative reactions of saturates, aromatics, and resins (S-Ar-R) from extra-heavy crude oil. Fuel 311: 122596. Merdrignac, I. and Espinat, D. (2007). Physicochemical characterization of petroleum fractions: the state of the art. Oil & Gas Science and Technology – Rev IFP 62: 7–32. Mironov, N.A., Abilova, G.R., Borisova, Y.Y. et al. (2018). Comparative study of resins and asphaltenes of heavy oils as sources for obtaining pure vanadyl porphyrins by the sulfocationite-based chromatographic method. Energy & Fuels 32: 12435–12446. Mohammad, I.A., Abdallah, R.I., El-Naggar, A.Y. et al. (2012). Characterization of four Egyptian crude oils. College of Natural Science 10: 72–79. Mozafari, M. and Nasri, Z. (2017). Operational conditions effects on Iranian heavy oil upgrading using microwave irradiation. Journal of Petroleum Science and Engineering 151: 40–48. Mukhamatdinov, I.I., Salih, I.S.S., Khelkhal, M.A. et al. (2021). Application of aromatic and industrial solvents for enhancing heavy oil recovery from the Ashalcha field. Energy and Fuels 35: 374–385. Muraza, O. and Galadima, A. (2015). Aquathermolysis of heavy oil: a review and perspective on catalyst development. Fuel 157: 219–231. ´ Murillo-Hernández, J.A., García-Cruz, I., Lopez-Ramírez, S. et al. (2009). Aggregation behavior of heavy crude oil-ionic liquids solutions by fluorescence spectroscopy. Energy and Fuels 23: 4584–4592. Nikolaychuk, E., Stratiev, D., Velkov, I. et al. (2015). Conversion of heavy oil distillation data from ASTM D-1160 to ASTM D-5236. Petroleum and Coal 57: 266–279. Novikov, E.A., Sergeev, Y.A., Sanzharov, V.v. et al. (2019). Application of multidimensional analysis methods to dead oil characterization on the basis of data on thermal field-flow fractionation of native asphaltene nanoparticles. Petroleum Chemistry 59: 34–47. Ocanto, O., Marcano, F., Castillo, J. et al. (2009). Influence of experimental parameters on the determination of asphaltenes flocculation onset by the titration method. Energy and Fuels 23: 3039–3044.

References

Ortiz Moreno, H., Ramírez, J., Sanchez-Minero, F. et al. (2014). Hydrocracking of Maya crude oil in a slurry-phase batch reactor. II. Effect of catalyst load. Fuel 130: 263–272. Palacio Lozano, D.C., Orrego-Ruiz, J.A., Cabanzo Hernández, R. et al. (2017). APPI(+)-FTICR mass spectrometry coupled to partial least squares with genetic algorithm variable selection for prediction of API gravity and CCR of crude oil and vacuum residues. Fuel 193: 39–44. Park, J.W., Kim, M.Y., Im, S.I. et al. (2022). Development of correlations between deasphalted oil yield and Hansen solubility parameters of heavy oil SARA fractions for solvent deasphalting extraction. Journal of Industrial and Engineering Chemistry 107: 456–465. Peramanu, S., Pruden, B.B., and Rahimi, P. (1999). Molecular weight and specific gravity distributions for Athabasca and Cold Lake bitumens and their saturate, aromatic, resin, and asphaltene fractions. Industrial and Engineering Chemistry Research 38: 3121–3130. Quitian, A. and Ancheyta, J. (2016). Experimental methods for developing kinetic models for hydrocracking reactions with slurry-phase catalyst using batch reactors. Energy and Fuels 30: 4419–4437. Rana, M.S., Ancheyta, J., Maity, S.K., and Rayo, P. (2008). Heavy crude oil hydroprocessing: a zeolite-based CoMo catalyst and its spent catalyst characterization. Catalysis Today 130: 411–420. Riazi, M.R. (2005). Characterization and Properties of Petroleum, 1e. Philadelphia, PA: ASTM International. Rodrigues, É.V.A., Silva, S.R.C., Romão, W. et al. (2018). Determination of crude oil physicochemical properties by high-temperature gas chromatography associated with multivariate calibration. Fuel 220: 389–395. Sánchez-Minero, F., Sánchez-Reyna, G., Ancheyta, J., and Marroquin, G. (2014). Comparison of correlations based on API gravity for predicting viscosity of crude oils. Fuel 138: 193–199. Santos, R.G., Loh, W., Bannwart, A.C., and Trevisan, O. v. (2014). An overview of heavy oil properties and its recovery and transportation methods. Brazilian Journal of Chemical Engineering 31: 571–590. Satya, S., Roehner, R.M., Deo, M.D. et al. (2007). Estimation of properties of crude oil residual fractions using chemometrics. Energy and Fuels 21: 998–1005. Shafiq, I., Shafique, S., Akhter, P. et al. (2020). Recent developments in alumina supported hydrodesulfurization catalysts for the production of sulfur-free refinery products: a technical review. Catalysis Reviews 64: 1–86. Shang, H., Liu, Y., Shi, J.C. et al. (2016). Microwave-assisted nickel and vanadium removal from crude oil. Fuel Processing Technology 142: 250–257. Shang, H., Yue, Y., Zhang, J. et al. (2018). Effect of microwave irradiation on the viscosity of crude oil: a view at the molecular level. Fuel Processing Technology 170: 44–52. Silva, S.L., Silva, A.M.S., Ribeiro, J.C. et al. (2011). Chromatographic and spectroscopic analysis of heavy crude oil mixtures with emphasis in nuclear magnetic resonance spectroscopy: a review. Analytica Chimica Acta 707: 18–37. Speight, J.G. (2002). Handbook of Petroleum Product Analysis, 1e. Wiley. Speight, J.G. (2013). Heavy and Extra-heavy Oil Upgrading Technologies. Elsevier Science. Speight, J.G. (2014). The Chemistry and Technology of Petroleum, 5e. CRC Press. Speight, J.G. (2016). Introduction to Enhanced Recovery Methods for Heavy Oil and Tar Sands, 2e. Cambridge, MA: Elsevier. Speight, J.G. (2019). Handbook of Petrochemical Processes, 1e. Bica Raton, FL: CRC Press/Taylor & Francis Group. Stratiev, D., Dinkov, R., Petkov, K., and Stanulov, K. (2010). Evaluation of crude oil quality. Petroleum and Coal 52: 35–43. Sullivan, A.P., Zaki, N.N., Sjöblom, J., and Kilpatrick, P.K. (2007). The stability of water-in-crude and model oil emulsions. Canadian Journal of Chemical Engineering 85: 793–807.

37

38

1 Properties of Heavy and Extra-Heavy Crude Oils

Taborda, E.A., Alvarado, V., and Cortés, F.B. (2017a). Effect of SiO2 -based nanofluids in the reduction of naphtha consumption for heavy and extra-heavy oils transport: economic impacts on the Colombian market. Energy Conversion and Management 148: 30–42. Taborda, E.A., Franco, C.A., Ruiz, M.A. et al. (2017b). Experimental and theoretical study of viscosity reduction in heavy crude oils by addition of nanoparticles. Energy & Fuels 31: 1329–1338. Taheri-Shakib, J., Shekarifard, A., and Naderi, H. (2018). Heavy crude oil upgrading using nanoparticles by applying electromagnetic technique. Fuel 232: 704–711. Tirado, A., Yuan, C., Varfolomeev, M.A., and Ancheyta, J. (2022). Kinetic modeling of aquathermolysis for upgrading of heavy oils. Fuel 310: 122286. Wang, Q., Guo, L., Wang, Z.X. et al. (2012). Hydrogen donor visbreaking of Venezuelan vacuum residue. Journal of Fuel Chemistry and Technology 40: 1317–1322. Wei, Q., Wen, S., Tao, X. et al. (2015). Hydrodenitrogenation of basic and non-basic nitrogen-containing compounds in coker gas oil. Fuel Processing Technology 129: 76–84. Wu, B., Zhu, J., Wang, J., and Jiang, C. (2006). Technique for high-viscosity crude oil demetallization in the Liaohe oil field. Energy & Fuels 20: 1345–1349. Yadykova, A.Y. and Ilyin, S.O. (2022). Compatibility and rheology of bio-oil blends with light and heavy crude oils. Fuel 314: 122761. Yeletsky, P.M., Zaikina, O.O., Sosnin, G.A. et al. (2020). Heavy oil cracking in the presence of steam and nanodispersed catalysts based on different metals. Fuel Processing Technology 199: 106239. Zhang, W., Andersson, J.T., Räder, H.J., and Müllen, K. (2015). Molecular characterization of large polycyclic aromatic hydrocarbons in solid petroleum pitch and coal tar pitch by high resolution MALDI ToF MS and insights from ion mobility separation. Carbon 95: 672–680. Zhang, S., Huo, J., Sun, X. et al. (2021). Molecular composition reveals unique rheological property of Karamay heavy crude oil. Energy and Fuels 35: 473–478. Zhao, S., Xu, Z., Xu, C. et al. (2005). Systematic characterization of petroleum residua based on SFEF. Fuel 84: 635–645. Zhao, X., Liu, Y., Xu, C. et al. (2013). Separation and characterization of vanadyl porphyrins in Venezuela Orinoco heavy crude oil. Energy and Fuels 27: 2874–2882. Zhao, S., Pu, W., Varfolomeev, M.A. et al. (2018). Comprehensive investigations into low temperature oxidation of heavy crude oil. Journal of Petroleum Science and Engineering 171: 835–842. Zhao, S., Pu, W., Pan, J. et al. (2022). Thermo-oxidative characteristics and kinetics of light, heavy, and extra-heavy crude oils using accelerating rate calorimetry. Fuel 312: 123001. Zheng, C., Zhu, M., Zareie, R., and Zhang, D. (2018). Characterisation of subfractions of asphaltenes extracted from an oil sand using NMR, DEPT and MALDI-TOF. Journal of Petroleum Science and Engineering 168: 148–155. Zou, C. (2017). Heavy oil and bitumen. In: Unconventional Petroleum Geology, 345–370. Elsevier.

39

2 Advanced Characterization of Heavy Crude Oils and their Fractions

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

40

2.1 EPR Spectroscopy Marat R. Gafurov, Alexander Rodionov, Fadis Murzakhanov, and Georgy Mamin Institute of Physics, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia

Almost every type of crude oil and hydrocarbon-bearing reservoir is paramagnetic (Yen and Chilingarian 1994, 2000) to be analyzed with electron paramagnetic resonance (EPR). Paramagnetic species/centers (PCs) in crude oils, like various complexes of d-metals (V, Ni, Fe, Mn) and unpaired electrons delocalized over many conjugated or aromatic chemical bonds – carbon-centered stable “free” radicals (FRs) – are mainly concentrated in high-molecular components, such as asphaltenes, resins, and polycyclic aromatic hydrocarbons. The mass content of these components can reach 45% for petroleum and 75% for bitumen (Yen and Chilingarian 1994, 2000). The study of the complex behavior of high molecular weight components, which determine the high values of density and viscosity of petroleum systems, is an urgent and intractable task. Studies of both in situ and extracted petroleum components can be useful in determining the structure and structural transformations in such complex systems (Evdokimov et al. 2004). Understanding the key factors that determine the aggregative stability of petroleum systems, chemical and structural transformations of their components in various external conditions, and the presence of specially introduced chemical compounds is fundamental for creating technologies for the production and processing of heavy oil (Martyanov et al. 2017). Concentration of PCs in asphaltenes, for example, can exceed the values of 1022 spins per gram (Yen and Chilingarian 1994, 2000). This means that PCs can serve as effective indicators (probes, signaling molecules) to follow the processes of various external treatments, such as aging (Caputo et al. 2022), influence of the electromagnetic field (EMF) (Vakhin et al. 2021), or supercritical water (SCW) (Djimasbe et al. 2021) and other types of improved-oil-recovery (IOR) and enhanced-oil-recovery (EOR) methods (Martyanov et al. 2017). EPR has long been known as one of the most informative methods for detecting and identifying PCs in the carbon-containing systems (Khasanova et al. 2017; Gafurov et al. 2020). The method does not require additional sample preparation (like dilution), is nondestructive, and allows one to study solid and liquid species. The availability of pulsed and high-field EPR spectrometers opens new opportunities for the analysis, structural investigation of asphaltenes, and investigation of their treatment processes, using intrinsic paramagnetic centers as sensitive probes (Gafurov et al. 2020). We will describe basic EPR techniques in Section 2.1.1. Section 2.1.2 deals with pulsed EPR techniques such as electron spin echo (ESE), electron spin echo envelope modulation (ESEEM) and electron-nuclear double resonance (ENDOR). HYperfine Sublevel CORrelation (HYSCORE) spectroscopy as applied to the study of oil asphaltenes is described in Section 2.1.3. Examples of petroleum research using ENDOR are shown in Section 2.1.4.

2.1.1 Basic Principles of EPR Spectroscopy for Petroleum Investigation

2.1.1 Basic Principles of EPR Spectroscopy for Petroleum Investigation The most EPR spectrometers operate in continuous wave (CW) mode. In CW mode, low-power microwave (MW) radiation is constantly applied to the sample. Typically, petroleum species are placed in a quartz or a glass tube, which is then inserted into an EPR resonator located in an external magnetic field B0 . The simplest microwave resonator is a rectangular or cylindrical metal waveguide closed on both sides. Purpose of the resonator is to accumulate energy and separate the electric and magnetic components of the microwave field, since the EPR effect is caused by interaction with the magnetic component (Eaton et al. 2010). The standard size of the tube inserted into the resonator samples is 3–5 mm in diameter and 5–30 mm in height in the X-band (MW frequency, 𝜈 MW , of 9–10 GHz). Measurements in the high magnetic fields, like W-band (𝜈 MW of 94–96 GHz), require a much smaller volume of the substance (tubes with a diameter of 0.4–0.8 mm, sample length 2–10 mm). It can be considered as an advantage for measuring specimens of limited amount, new catalytic systems, etc. To increase the sensitivity of EPR spectrometers, in CW mode, B0 is modulated by a sinusoidal signal with a frequency of 10–100 kHz and a small amplitude (0.01–2 mT). In this case, the signal measured by the detector is amplitude-modulated at the modulation frequency, and the amplitude of its main harmonic is usually proportional to the first derivative of the absorption line (Figures 2.1.1 and 2.1.2). Consequently, the concentration of paramagnetic species can be determined from the EPR spectrum (Table 2.1.1) by its double integration (Eaton et al. 2010). The study of changes in the EPR line intensities (amplitudes) allows tracking radical combination–recombination processes (Dolomatov et al. 2016) of EOR/IOR treatments (Djimasbe et al. 2021). EPR revealed, for example, that an increasing temperature of pyrolysis promotes the formation of new free organic radicals in rock samples: in Domanic rocks of Semiluki–Mendym deposits at 350 ∘ S and in carbonate rocks of Dankov–Lebedyan horizon at 600 ∘ S (Kayukova et al. 2018). EPR spectra observed in petroleum components are due to the presence of magnetic moment for unpaired electrons and are affected by several types of interactions of the unpaired electron with its environment (Gafurov et al. 2020): (i) Zeeman interaction between the ensemble of unpaired electrons and the external magnetic field B0 , (ii) the spin–orbital interaction, (iii) electron-nuclear hyperfine interaction (HFI), (iv) interaction with other unpaired electrons (spin–spin interaction), (v) interaction with the environmental bath (spin–lattice interaction). The most significant contribution is the Zeeman interaction (interaction of the electron magnetic moment S with B0 ), which determines the position of the resonance lines (Gafurov et al. 2020): h𝜈MW = g𝛽B0 ,

(2.1.1)

where h is the Planck constant, g is the spectroscopic splitting factor, and 𝛽 is the Bohr magneton. The spin–orbital and spin–spin interactions are the reasons for the shifting of the spectroscopic g-factor from the known value for the free electron (ge = 2.0023). The typical range of g-factors in petroleum systems varies from 1.96 (for the vanadium-containing petroporphyrins with the skeleton VO2+ , see Figure 2.1.3 and description below) to 2.008 for oxygen- and sulfur-containing FR (Yen and Chilingarian 1994, 2000; Khasanova et al. 2017; Gizatullin et al. 2022). A number of oil-bearing rocks contain complexes of iron group Fe3+ ions with a characteristic signal in the region of “half” magnetic fields with g ≈ 4.3 (Figure 2.1.4), Mn2+ , SO3− , and SO2− ions. Exact determination of the EPR line positions, therefore, allows to spot the nature of the paramagnetic species in studied systems. Changes in the observed g-factors and EPR linewidths (usually measured as a value between the maximum and the minimum of CW EPR line, ΔBpp ) testify the structural changes during the EOR/IOR processes (Trukhan et al. 2022; Khasanova et al. 2017).

41

42

2.1 EPR Spectroscopy

Salym oil CW, X-band

Salym oil CW, W-band

g = 2.0025

ΔBpp = 0.31 mT

Fitting Experiment 339

340

341

342

343

344

345

346

347

3354

3356

3358

B0 (mT)

B0 (mT)

(a)

(b)

3360

3362

Salym oil ESE W-band g⊥ = 2.00145

Experiment g‖ = 2.00235

Fitting 3350

3355

B0

3360

(mT) (c)

Figure 2.1.1 EPR spectra (black) of oil sample from the Bazhenov formation (Salym area, West Siberian basin Russia) and their fittings (red) with the parameters presented in Table 2.1.1. (a) CW X-band EPR; (b) CW W-band EPR; (c) pulsed (ESE) W-band EPR.

High-field measurements give spectrally better-resolved information to be applied for the more accurate calculation of g-factors. From the other side, the resonance condition of Eq. (2.1.1) for g ≈ 2.0 can be fulfilled for B0 of about 0.35 T in the X-band and B0 of 3.5 T for the W-band. Thus, high-frequency (high-field) studies require the use of superconducting magnets, which significantly increase the cost of measurements. Figure 2.1.1 demonstrates conventional EPR spectra in the X- and W-band ranges for oil sample from the Bazhenov formation (Salym Area, West Siberian basin, Russia). Only one line from FR can be registered in this sample. In the X-band, it can be adequately fitted as a line with an isotropic g-factor close to that for “free” electron and a width of ΔBpp = 0.31 mT by combining of Gaussian (G) and Lorentzian (L) lineshapes (Table 2.1.1). Approximation of the FR EPR line by the Voigt function (convolution of the Lorentz and Gaussian lineshapes) is depicted in Figure 2.1.1a.

2.1.1 Basic Principles of EPR Spectroscopy for Petroleum Investigation

Asphaltenes subfractions from Athabasca oil sands ESE W-band A1 A2 A1: toluene D8 (1 : 20)

3344

3346

3348

3350

3352

3354

B0 (mT) Figure 2.1.2 Normalized ESE (pulsed) EPR in the W-band for the asphaltenes subfractions A1 (red line) and A2 (blue line) obtained from the Zuzeevsk oxidized bitumen (Russia) registered at T = 300 K and 𝜏 = 240 ns. Mixture of A1 in toluene D8 (m : m = 1 : 20) was registered at T = 200 K (green curve). Table 2.1.1 Concentrations of FR in the petroleum samples (1); parameters of approximation of EPR spectra for FR in the X-band (2) as convolutions of Gaussian (G) and Lorentz (L) lineshapes, and as the center of axial symmetry in the W-band (3).

Sample

Concentration of FR (spin/g)

Parameters of describing of the FR line in the X-band

Parameters of describing of the FR line in the W-band

Salym oil

(8.2 ± 0.8)×1018

g = 2.0025; G:L = 53 : 47; ΔH PP = 0.31 mT

g|| = 2.00235, g⟂ = 2.00145

g = 2.0032; G:L = 19 : 81; ΔH PP = 0.64 mT

g|| = 2.0045, g⟂ = 2.0029

Athabasca asphaltenes (A1)

20

(1.2 ± 0.2)×10

B2 c

c

a

B1 b

b

a

ΔH PP (L) = 0.38 mT

ΔH PP (L) = 1.30 mT

-V -N -O -C -H

Y Z

X

Figure 2.1.3 Schematic structure of the petroleum vanadylporphyrin molecule. The values of the external magnetic field parallel to the plane of the porphyrin (B1 ) or perpendicular to it (B2 ) are determined from the EPR spectra. X and Y axes are chosen to be in plane, and the Z axis is chosen to be out of plane.

43

44

2.1 EPR Spectroscopy

Figure 2.1.4 X-band EPR spectra for oil (red) and asphaltenes (black) samples from the Tahe oilfield (Tarim basin, China). The intensities of the spectra are normalized to the samples masses.

Tahe oil

Tahe asphaltenes

0

100

200

300 400 B0 (mT)

500

600

As illustrated in Figure 2.1.1b, FR spectrum in the W-band has an asymmetric line shape due to the better spectral resolution comparing to the low-field X-band measurements. Thus, the typically accepted description of EPR spectra for FR in the X-band does not allow describing the EPR spectrum at higher frequencies (Di Mauro et al. 2005). The single line can be fitted as a PC with axial symmetry, i.e. with two components of g-factor (Figure 2.1.1c and Table 2.1.1). EPR spectra in pulsed (ESE) mode at high frequency (W-band) for asphaltenes from the Zuzeevsk bitumen (Tatarstan, Russia) are shown in Figure 2.1.2 with the EPR parameters depicted in Table 2.1.1. As seen from Table 2.1.1, intensity of EPR line (concentration of FR), value of g-factors or their components, value of the EPR linewidth (ΔBpp ), relation between the Lorentz and Gaussian lineshapes are very specific for every type of petroleum system. Therefore, the extracted EPR parameters for FR can serve as fingerprints of oil systems. Although the presence of FR in petroleum is unquestionable, EPR signal does not readily reveal information about their structures. Three prototypical FR species are expected in the petroleum (Zhang et al. 2020). The first type of radicals consists of a large delocalized 𝜋 system with phenalenyl as a prototype. It was found in bitumen, petroleum asphaltens, jet fuel, coal tar, marine diesel, and soot samples. Large 𝜋-radicals are supposed to be even more stable than the phenalenyl radical as the unpaired electron can be delocalized over many carbons with the multiple possible resonance structures (Zhang et al. 2020). The second type with fluorenyl radical as a prototype is doubly benzylic radicals (𝜎-radicals) stabilized by delocalization into two aromatic systems confined by a five- or six-membered ring. This radical can be present in crude oil and in thermally processed hydrocarbons. The third type benzylic radicals are 𝜎-radicals attached to only one aromatic system. Benzylic-type FRs are more reactive than the other types of radicals and are less likely to be present in high concentrations in petroleum (Zhang et al. 2020). To probe the abilities of EPR to get information about the origin of FR, we have investigated asphaltenes subfractions from various oil systems (Mamin et al. 2016; Gafurov et al. 2019a). Asphaltenes can be fractionated, i.e. divided into fractions, using different ratios of solvents. We have used a duo of toluene and heptane. The fraction that precipitates from the asphaltene solution at a minimum content of n-alkane in binary solvents is characterized as heavy, insoluble, and polar (Ganeeva et al. 2011). This fraction is designated as A1, and the molecules that make up this fraction are molecules of the A1 type. The asphaltene fraction that precipitates at the maximum n-alkane content was described as light, soluble, and nonpolar. This fraction is called A2, and the molecules of this fraction are molecules of the A2 type. According to the results of

2.1.1 Basic Principles of EPR Spectroscopy for Petroleum Investigation

various investigations reviewed in the study by Ganeeva et al. (2011), subfraction A1 has a high molecular weight and aromaticity factor, a low H:C atomic ratio, and contain highly condensed aromatic compounds. Subfraction A2 has a high H:C ratio and a relatively low aromaticity factor. As seen from Figure 2.1.2, for some cases, even high-resolute EPR spectrum does not allow to get information about the FR structural changes during the asphaltene fractioning or dissolution. This may lead to the incorrect conclusion that the possibilities of EPR are extremely limited. More elaborated EPR techniques, as it will be shown below, should be applied for petroleum systems study. Though the traditional EPR lacks in certain instances to follow FR structural transformations, it can be still used to investigate other PCs, such as petroleum vanadylporphyrins (VPs, VO2+ ) complexes (Figure 2.1.3). The interaction of an unpaired electron with the nuclear magnetic moment of the nucleus (nuclei) leads to the splitting of electronic energy levels, forming the so-called “hyperfine structure” (HFS) of the EPR spectrum (Gizatullin et al. 2019). Each mS electron level is split into a group of (2I + 1) closely spaced levels, where I is a nuclear spin. Allowed transitions that contribute to the HFS of the EPR spectrum are only those transitions for which the conditions Δms = ±1 and ΔmI = 0 are fulfilled simultaneously. The HFI splitting between the adjacent lines of the HFS spectrum is called the HFI constant and denoted as A. Due to the axial symmetry of VO2+ (Figure 2.1.3), the presence of a nuclear spin 51 V (I = 7/2, natural abundance of 99.75%), and the powder (disordered) state of the petroleum samples, 16 overlapping resonant transitions are observed in the EPR spectrum (Figures 2.1.4 and 2.1.5) where the first eight lines (2I + 1 = 2 × 7/2+1 = 8) are for parallel (with g|| ), and the remaining eight for perpendicular orientation (with g⟂ ). The FR line overlaps several transitions from VO2+ . The spectroscopic parameters calculated from the simulation of the EPR spectrum for Ashalchinskoe oil (Ashalchinskoe oilfield, Russia) for each PC is listed in Table 2.1.2. The VP EPR can be described by the spin Hamiltonian of axial symmetry in the form of ̂ = g|| 𝛽Bz Ŝ z + g⟂ 𝛽(Bx Ŝ x + By Ŝ y ) + A|| Ŝ z Î z + A⟂ (Ŝ x Î x + Ŝ y Î y ), H

(2.1.2)

where Sk , I k , Bk with k = {x, y, z} are the projections of S, I, and B0 on the XYZ axes of VP (Figure 2.1.3), correspondingly. It follows that by fixing an appropriate value of B0 in EPR spectrum of VP, one can choose the preferable direction of VP in relation to the external magnetic field (in plane or out of plane). Therefore, EPR gives an opportunity for pulsed techniques described in Sections 2.1.2 and 2.1.3 to get orientation-dependent information not only for crystals but also even for the disordered (powder) systems, such as petroleum components (Gizatullin et al. 2019). It was shown by Dickson et al. (1972) that the values of the parameters g and A can be used to clarify the nature of paramagnetic centers and determine the origin of the petroleum (VP complexes, Figure 2.1.3). Therefore, for the ligand environment of vanadium VOS4 , the g-factor values are higher, and the A value for the HFI between electron and 51 V nuclei is lower compared to VON4 , VON2 O2 , or VOO4 . This fact continues to be used in a number of studies to control the processes of desulfurization (Espinosa et al. 2001; Cui et al. 2017), for example, although the changes in the parameters of EPR spectra observed in the X-band experiments are tiny due to the high stability of vanadylporphyrins. For Ashalchinskoe oil (Figure 2.1.5a), literature suggests that the total amount of vanadium is 0.023 wt.% concentrated mainly in asphaltenes and then in resins (Yakubov et al. 2017). Table 2.1.3 presents the results of measurements of concentrations of FR and VP, which show that paramagnetic species in this sample are mainly accumulated in asphaltenes, and their ratio is practically the same for initial oil and extracted components. It gives a sign that VP complexes and FR are “bounded” in asphaltenes and, probably, participate in the processes of asphaltene aggregation (Yen and Chilingarian 2000). Heating and dissolution are the most accessible and frequently used methods of studying aggregation.

45

46

2.1 EPR Spectroscopy

FR CW EPR X-band A⊥(for 51V in VO2+)

A (for 51V in VO2+)

280

300

320

340

360

380

400

420

B0 (mT) (a) FR

ESE X-band T = 250 K

VO2+

280

300

320

340

360

380

400

420

B0 (mT) (b) ESE W-band VO2+ FR

Experiment Fitting FR + VO2+ 3350

3400

B0 (mT) (c)

3450

3500

Figure 2.1.5 EPR spectra for Ashalchinskoe oil. (a) CW EPR at X-band; (b) ESE EPR at X-band; (c) ESE EPR at W-band. Positions of FR line and HFS from 51 V for VO2+ are shown. For (c) the repetition time was chosen to be of 5 μs to suppress the signal from FR.

2.1.1 Basic Principles of EPR Spectroscopy for Petroleum Investigation

Table 2.1.2

EPR fitting parameters for Ashalchinskoe oil samples.

VP

g⟂ = 1.9860; g|| = 1.9661 A⟂ = 160 MHz; A|| = 470(3) MHz g = 2.0041(4); G:L = 61 : 39; ΔBpp = 0.64(2) mT

FR

Table 2.1.3 Absolute and relative concentrations of FR and VO2+ in components of Ashalchinskoe oil from the X-band EPR measurements.

Oil

C(VO2+ ) (spin/g)

C(FR) (spin/g)

C(VO2+ )/C(FR)

1.1 × 1018

1.2 × 1017

9.2

18

Asphaltenes

7.5 × 10

8.1(9) × 10

Resins

0.8 × 1018

0.9 × 1017

17

9.1 9.0

With temperature increase, EPR parameters for FR such as g-factor, lineshape, or linewidth often alter insignificantly (Dolomatov et al. 2016). Tracking the temperature dependence of EPR parameters for FR gives usually information about the processes of radical combination–recombination (Dolomatov et al. 2016; Dappe et al. 2020). In contrast to that, HFS of VP in initial and diluted petroleum systems transforms drastically (Gafurov et al. 2018a) as Figure 2.1.6 illustrates. One can

T (K) 310 326 337 350 363 377 390 406 428 446 465 475 491 506 528 555 565 595 615 642 661 250

275

300

325

350 B0 (mT)

375

400

425

Figure 2.1.6 EPR spectra for Ashalchinskoe oil with temperature. In the vicinity of T = 490–530 K a transformation into the isotropic spectrum is to observe.

47

48

2.1 EPR Spectroscopy

observe a transition from the rigid regime (powder spectrum with 16 EPR lines) into the fast motion regime (the eight-line spectrum can be described by isotropic hyperfine constants). From that with help of established theoretical approaches, the rotational correlation times and size of asphaltene aggregates can be extracted (Trukhan et al. 2014). The temperature of the obtained motional transition varies with the studied system that allows comparing stability of different petroleum complexes in situ conditions (Gafurov et al. 2018a) and for the diluted species (Trukhan et al. 2014). It should be noted that EPR parameters for FR (except intensity) are not affected by heating of Ashalchinskoe oil sample (Gafurov et al. 2018a). Significant changes of line shape of FR in Ashalchinskoe oil was observed in the study by Djimasbe et al. (2021) under the influence of SCW. It was suggested that in the beginning of the processes of SCW upgrading a part of initial FR in the heavy oil can react easily by radical recombination reaction. However, some of FR can participate in the reaction only when their environment is destroyed first. The narrowing of FR was ascribed to the removal (destroying) of sulfur-containing radicals. Sharp decrease in the concentration of VO2+ complexes was interpreted in two ways: being directly transferred into coke without being destroyed during the upgrading process and being transformed to other valence states of vanadyl and removed by a series of chemical reactions.

2.1.2

Pulsed EPR Techniques

In contrast to CW EPR, in which continuous MW radiation reveals a picture of splitting the energy levels, pulsed EPR techniques can shed light on the dynamic behavior of the system, measure relaxation times, allowing untwisting complex spectra of interacting PCs, etc. By using pulse techniques, it became possible to identify the nature and determine the values of electron–nuclear interactions, which due to their small values compared to the width of the EPR lines cannot be resolved in the CW EPR spectra (Gafurov et al. 2020). For example, for petroleum systems, HFS with 1 H and 14 N nuclei can be determined (see below). Pulsed EPR spectra presented in this manuscript were recorded by the sweeping of the external magnetic field B0 and detecting an ESE with rectangular pulses: 𝜋 − 𝜏 − 𝜋 − 𝜏 − ESE (2.1.3) 2 where the minimal duration of 𝜋2 pulse was equal to 16 ns (𝜋 = 32 ns), the minimal delay between pulses 𝜏 = 200 ns. Examples of the ESE-detected EPR spectra for various petroleum systems are shown in Figures 2.1.1c, 2.1.2, 2.1.5b, and 2.1.5c. Pulse experiments are usually performed at the low-quality factor mode of EPR resonator, which reduces the sensitivity of EPR measurements in comparison with the CW regime. The shape of the spectra recorded in this way (the amplitude or integral values of ESE signal) usually coincides with the absorption spectra (cf. spectra presented in Figures 2.1.1 and 2.1.5). In some cases, for example, by studying light oils, it may be necessary to slow down the electron relaxation processes by lowering the sample temperature in order to detect ESE spectra. To determine the time of transverse (spin–spin, phase-memory) relaxation of T 2 , the same pulse sequence as shown in Eq. (2.1.3) was applied by increasing the time interval 𝜏 between the pulses with a step of 4 ns from 200 ns to the desired value. The dependence of ESE as a function of time 2𝜏 in a fixed magnetic field can be written as follows: ( ) 2𝜏 I(2𝜏) = I0 exp (2.1.4) T2

2.1.2 Pulsed EPR Techniques

A three-pulse sequence was also used to register electron–nuclear modulations in ESEEM (Murzakhanov et al. 2020): 𝜋∕2 − (𝜏1 + Δ𝜏) − 𝜋∕2 − (𝜏2 + Δ𝜏) − 𝜋∕2

(2.1.5)

To measure the longitudinal (spin–lattice) relaxation time T 1 , the inversion–recovery three-pulse sequences was used 𝜋 (2.1.6) 𝜋 − Ts − − 𝜏 − 𝜋 − 𝜏 − ESE 2 where 𝜏 is fixed, while the period Ts is changed. The dependence of the integral intensity of the ESE can be written as ) ( Time (2.1.7) I(Ts) = I1 − I0 exp − T1 Similar to nuclear magnetic resonance (NMR) applications, microwave pulses can be combined into special sequences where the delay time and pulse duration are configurable parameters. However, one should keep in mind that the values of these times in EPR are three to six orders of magnitude shorter than in NMR, which is due to the corresponding difference in the relaxation times of electrons and nuclei. In the research by Djimasbe et al. (2021), it was shown from the X-band measurements that the values of the electronic relaxation times of intrinsic paramagnetic centers are sensitive to the SCW conversion demonstrating an opportunity to follow the effectiveness of SCW treatment not only by tracking the intensity of the VO2+ radicals but also by measuring T 1e and T 2e for FR and VO2+ . Therefore, it was suggested that EPR relaxometry may serve as a fast (comparing to the NMR relaxometry) technique for analyzing oil treatment processes in both ex situ and in situ EOR/IOR of hard-to-recover hydrocarbons. Figure 2.1.7 demonstrates that FR relaxation in the W-band is sensitive to the influence of EMF. Increase of the relaxation times (as measured in the magnetic field corresponding to the maximum of ESE signal, Figure 2.1.1c) means structural and magnetic “ordering” in the nearest to the paramagnetic center environment with EMF exposure. It can be ascribed to the destruction of a significant T1 T2

26

Relaxation time (μs)

25 24 23 22 21 0.70 0.65 0.60 0

10

20 30 40 50 Time of EF impact (min)

60

Figure 2.1.7 Changes for the electronic relaxation times T 1 (upper panel, squares) and T 2 (lower panel, circles) for oil sample from the Bazhenov formation (Salym area, West Siberian basin Russia, see Figure 2.1.1c) with the time of impact of electromagnetic field of 50 Hz. The straight lines are given for eyes.

49

50

2.1 EPR Spectroscopy

Table 2.1.4 Values of electronic relaxation times T 1 and T 2 for Ashalchinskoe oil measured at T = 250 K in X-band. VO2+ (𝛍s)

FR (𝛍s)

T1

1.7

23.6

T2

0.3

0.3

part of hydrocarbons with the formation of isoalkanes of lower molecular weight, alkanes of normal structure, and elimination of alkyl substituents leading to the more uniform environments for the PCs inside the petroleum. It should be noted that no observable changes in EPR spectra were observed in Salym oil samples (Figure 2.1.1c) with EMF. Values of electronic relaxation times for Ashalchinskoe oil measured at T = 250 K are presented in Table 2.1.4. As seen, spin–lattice relaxation time of vanadyl complexes is in the order of magnitude shorter than this for FR. Due to the spin–spin (FR-VO2+ ) interaction, it may lead to the additional shortening of FR relaxation times. Comprehensive analysis of mutual impact of the mentioned paramagnetic species for various asphaltene (sub)fractions from four different oil sources was done in the study by Mamin et al. (2016). It was derived that FR and VO2+ are situated in the distances of 0.8–1.0 nm and form complexes which can be destroyed by dissolution (Figure 2.1.8). It should be noted again that no changes in EPR spectra for various Zuzeevsk bitumen (sub)fractions and their solutions even at higher toluene amount were observed (cf. Figure 2.1.2). The decays of the transverse magnetization of the VP with modulations (ESEEM) for Athabasca asphaltenes are shown in Figure 2.1.9a. The frequency (𝜈) and amplitude (depth) of the modulation for each center may differ markedly. This may be caused by a different ionic environment, the degree of anisotropy of the HFI, and the orientation of the center relative to the magnetic field. The oscillation frequency is determined by nuclear Larmor frequency, 𝜈 Larmor , which depends on the type of nucleus (the value of the gyromagnetic ratio 𝛾) and the presence of various interactions (hyperfine or quadrupole). The Fourier spectrum of the ESEEM signal from the VP is shown in Figure 2.1.9b. The line at 14.8 MHz belongs to the 1 H nuclei (𝛾/2𝜋 = 42.58 MHz/T) framing the VP. Intense low-frequency resonance lines correspond to vanadium 51 V (𝛾/2𝜋 = 7.05 MHz/T) and nitrogen 14 N (𝛾/2𝜋 = 3.077 MHz/T) nuclei. As seen, the ESEEM spectrum allows determining the type of core surrounded by a paramagnetic center even not detectable in EPR spectra.

100

Figure 2.1.8 T 2 decay curves for FR in the W-band for asphaltene fraction A1 from the Zuzeevsk bitumen (green dataset) and dissolved in deuterated toluene (m : m = 1 : 5, red dataset). The upper curve is a monoexponential fitting, and the lower curve is an approximation in the spin diffusion approach. Source: Adapted from Mamin et al. (2016).

No. 4

10–1

A1 in toluene

10–2 A1 10–3 500

1000

1500

2000 2τ (ns)

2500

3000

3500

2.1.3 HYSCORE Spectroscopy

ESE amplitude (a.u.)

Athabasca asphaltenes X-band VO2+ T = 50 K 2p ESEEM curves Fit by exp Nuclear modulation at 50 K 297 K

50 K

0

1

2

3

4

5

2τ (μs) (a)

FFT intensity (a.u.)

14N

ESEEM VO2+ T = 50 K 1H

0

10

20

30 ν (MHz) (b)

40

50

60

Figure 2.1.9 (a) T 2 decay curve for VO2+ complexes in the X-band for Athabasca asphaltenes at T = 50 and 297 K with the corresponding fitting for T = 50 K as a monoexponential function with T 2 = 880 ns. (b) The ESEEM spectrum for T = 50 K in the frequency domain after the Fourier transform. Peaks belonging to interaction with 14 N and 1 H nuclei are marked.

2.1.3

HYSCORE Spectroscopy

HYSCORE spectroscopy in EPR can be regarded as an analog of two-dimensional (2D) experiments in NMR. The HYSCORE makes it possible to study electron–nuclear interactions for nuclei with small magnetic moments (having a small value of 𝛾), to obtain information about the type of surrounding ligand, corresponding information about the magnitude and mechanism (dipole–dipole or contact Fermi), and determine the parameters of the quadrupole interaction (Q and 𝜂) in the case for nuclei with the nuclear spin I > 1/2. The method can provide information about the presence of equivalent nuclei, as well as distinguish two types of HFI: with weak coupling |A| < 2|vLarmor | and with strong coupling |A| > 2|𝜈 Larmor |. It gives a possibility to register

51

2.1 EPR Spectroscopy

Preparation EC → NC, NP

Mixing NC → NC Evolution

Detection NP, NC → EC

Evolution π

π/2

π/2

π/2 t2

t1

𝜏

𝜏

(a) Strong coupling

Weak coupling a/2π

2ν1

νβ ν2

52

να (– , +)

(+ , +)

–νβ

–να –a/4π

ν1

να

ν1

νβ

(b) Figure 2.1.10 (a) Pulse sequence used to acquire two-dimensional signals from electron–nuclear interactions, where NP – nuclear polarization, NC – nuclear coherences, and EC – electron coherences; (b) model of the HYSCORE spectra for an S = 1/2 centers with an anisotropic hyperfine interaction value with two nuclei (I = 1/2).

signals from the first coordination sphere of the studied complex(es) as well as from the second nuclear sphere (Gourier et al. 2010; Dutoit et al. 2022; Sadovnikova et al. 2022). The pulse sequence for HYSCORE is schematically shown in Figure 2.1.10a. Registration of the 2D time domain spectra was carried out by detection of the echo amplitude as a function of pulse delays t1 and t2 . The obtained modulation curves are then processed by 2D Fourier transformation to extract the resonance lines on the spectra with two frequencies 𝜈 1 and 𝜈 2 directed along the axes. The processed spectra for the simple system with S = 1/2 and two nuclei with I = 1/2 are shown in Figure 2.1.10b. Projected points that are symmetrically located at the diagonals in the spectra for an S = 1/2 species correspond to nuclear spin transitions for the mS = + 1/2 and − 1/2 electronic levels, correspondingly, denoted as 𝛼 and 𝛽 of a nonequivalent nucleus (Lund et al. 2011). The HYSCORE spectra for the Athabasca asphaltenes in the X-band in perpendicular (out of plane) orientation of B0 relative to the VO2+ skeleton (Figure 2.1.3) are presented in Figure 2.1.11a. In the right quadrant (+,+) for the nuclei with a weak bond (A < 2𝜈), signals from nitrogen 14 N, vanadium 51 V, and hydrogen 1 H nuclei were recorded. There are no additional splittings for 1 H and 14 N nuclei; however, for 51 V, a HFS is to detect with A = 4.7 MHz. Considering that the EPR spectrum of VO2+ exceeds hundreds of MHz, the unresolved signal from nitrogen indicates that these nuclei belong to the far environment (for example, the second coordination sphere) of VO2+ . In the left quadrant, there is a signal from one-quantum transitions from 14 N nuclei with the strong HFI interaction (A = 8.3–8.6 MHz).

2.1.3 HYSCORE Spectroscopy

B2 = 336.5 mT, c⊥B2 (+ , –)

(+ , +)

15

14N νI ≈ 1 MHz

10

5

0

νI = 6 MHz

14

N νI = 0.99 MHz 2A⊥ = 8.6 MHz –15

–10

νI = 3.7 MHz 2A = 4.73 MHz

51V,

–5

0

5

10

15

ν (MHz) (a) B1 = 299 mT, c║B1 (+ , +)

(+ , –)

1H

νI = 12.54 MHz

14

15

N νI = 0.91 MHz 2Aǁ = 15.4 MHz

10

νI = 9.86 MHz 5 νI = 4.86 MHz 0

–15

–10

–5

0

5

10

15

ν (MHz) (b) Figure 2.1.11 HYSCORE spectra for the Athabasca asphaltenes in X-band in magnetic fields B0 corresponding to the (a) perpendicular orientation of the VP complex and (b) parallel orientation of VP.

For the parallel (in plane) orientation of B0 relative to the VO2+ skeleton, no obvious signs of 14 N and 51 V are to be observed, only an intense signal from 1 H (Figure 2.1.11b). For the left quadrant, the signals from the nitrogen nuclei are still present, but they are recorded exclusively from two-quantum transitions with the selection rule ΔmI = ±2, located at twice the Larmor frequency of 14 N and with a splitting of 2A due to the additional quadrupole interaction for 14 N. The HYSCORE, therefore, made it possible to resolve the contributions from various types of HFI and determine the values of the HFI of the electron magnetic moment of VO2+ complexes with the surrounding nuclei such as 1 H, 14 N, and 51 V. From the obtained data, for example, one can determine values of the isotropic (Aiso = 8.3 MHz) and dipole–dipole (Add = −0.3 MHz) contributions for 14 N. From Add , an interatomic distance between the vanadium and the nearest nitrogen nuclei (r ≈ 2.1 Å) can be extracted (theoretical calculations by Gracheva et al. (2016) give the value of r = 2.069 Å).

53

54

2.1 EPR Spectroscopy

2.1.4

Pulsed ENDOR

The ability to detect ESE gives an opportunity to obtain ENDOR spectra by using Mims pulse sequence with an additional radiofrequency (RF) pulse inserted between the second and third microwave pulses (Figure 2.1.12). This approach requires additional RF equipment and specially designed resonator to excite and detect both EPR and NMR transitions. Application of ENDOR for investigations of asphaltenes aggregation and precipitation on various surfaces by using VO2+ as a paramagnetic probe is comprehensively described in the studies of Gracheva et al. (2016) and Gafurov et al. (2018, 2019b, 2019c). The features of structures of VP in various petroleum systems as derived from ENDOR measurements are presented in the investigation by Biktagirov et al. (2017). Here, we show the results related to the structural study of FR. The pulsed 1 H ENDOR spectrum for Tahe oil (see Figure 2.1.4 for EPR spectra) in the W-band resembles the stationary ENDOR for the light oil in the X-band (Galtsev et al. 1995) and is due to “remote” protons, contributing to the central part of the spectrum, and “near” protons (Figure 2.1.13). The central part can be fitted by a Lorentzian curve with a linewidth of Δ𝜈 = 0.16 MHz. It corresponds to 1 H nuclei at a distance more than 0.8 nm in the dipole–dipole approximation. A pair of wide “wings” in the ENDOR spectrum with a splitting between them of more than 1.3 MHz is due to the “near” protons with a wide distribution in the region of order 0.3–0.4 nm for the dipole–dipole approximation. Assuming a model of PC localization in polycondensed aromatic rings 0.85–1.5 nm in diameter and 1.5–3 Å thick is located one above the other, Galtsev et al. (1995) attributed the line in the middle of the ENDOR spectrum to FR inside asphaltene aggregates, while the “broad” spectrum – to individual asphaltene molecules are surrounded by molecules of other fractions (resins). As seen from Figure 2.1.13, 1 H ENDOR is a quite sensitive tool to follow the processes of disaggregation of asphaltenes’ supramolecular structures. Under the solvation in toluene, the main changes for Tahe oil are to be observed in the central part of ENDOR spectrum (Figure 2.1.13), while for the other type of oil (light one), Galtsev et al. (1995) detected the changes on the ENDOR wings. By applying more accurate mathematical calculations, we have analyzed the redistribution of protons for the asphaltenes (sub)fractions from Zuzeevsk bitumen (cf. Figure 2.1.2) in ENDOR spectra (Figure 2.1.14). To simplify the calculations, we supposed that only dipole–dipole electron–nuclear interaction contribute to the ENDOR spectrum and determined the contributions to the spectrum of protons located at different distances from FR. As can be seen from Figure 2.1.14, the deconvolution results (black lines) describe the experimental spectra quite well. For all asphaltene (sub)fractions, a peak is observed in the region of 0.75 nm (Figure 2.1.14, right panel). It corresponds to the nearest protons in aromatic chains, where FR is localized. The initial tp

t=0 τ

tm td τ

T

MW π/2

π/2

π/2 π

ESE

RF Figure 2.1.12

Pulse sequence to obtain Mims ENDOR spectra.

2.1.4 Pulsed ENDOR

Diffrence spectrum

Tahe: toluene D8 (1:1)

Tahe oil

141

142

143 νRF (MHz)

144

Figure 2.1.13 1 H Mims ENDOR spectra in the W-band for Tahe oil samples on FR line (black lower curve), Tahe oil dissolved in toluene (v : v = 1 : 1, blue line) and the difference spectrum between them. Central part of the difference spectrum is fitted by the Lorentzian curve (red line).

Oil. A1 A2 + toluene D8 A1 1 : 5

FR ENDOR Zuzeevsk bitumen

1.2

1.0

0.6

ρ (a.u.)

Echo intensity (a.u.)

0.8

0.4

0.2

0.0 –2

–1

0 Δν (MHz)

1

2 0.25

0.50

0.75 1.00 r (nm)

1.25

1.50

Figure 2.1.14 (Left panel) ENDOR spectra for FR in the W-band for the initial oil (red points), asphaltenes subfractions A1 (blue points), A2 (green points) and mixture of A1 in deuterated toluene (m : m = 1 : 5) obtained from the Zuzeevsk oxidized bitumen with their corresponding approximations (solid black lines). (Right panel) distribution function of the contribution of protons from the distance to FR, obtained from the ENDOR spectra.

55

56

2.1 EPR Spectroscopy

oil is characterized by a wide density distribution of interacting protons around FR, up to distance of 1.5 nm. The distance distributions for subfractions A1 and A2 differ in the vicinity of 0.45 nm. For the A1, there are practically no protons in this region, which is most likely due to the prevalence of aromatic chains. In fraction A2, there are a number of protons at this distance due to either alkane chains or the boundaries of the aromatic rings. When dissolved, both fractions show fewer protons at a distance of 1.25 nm than the initial oil samples. This is apparently due to the “washing out” of alkanes from asphaltenes.

References Biktagirov, T., Gafurov, M., Mamin, G. et al. (2017). In situ identification of various structural features of vanadyl porphyrins in crude oil by high-field (3.4 T) electron–nuclear double resonance spectroscopy combined with density functional theory calculations. Energy and Fuels 31 (2): 1243–1249. Caputo, P., Ventruti, G., Calandra, P. et al. (2022). Searching effective indicators of microstructural changes in bitumens during aging: a multi-technique approach. Colloids and Surfaces A: Physicochemical and Engineering Aspects 641: 128529. Cui, Q., Nakabayashi, K., Ma, X. et al. (2017). Examining the molecular entanglement between V=O complexes and their matrices in atmospheric residues by ESR. RSC Advances 7 (60): 37908–37914. Dappe, V., Ben Tayeb, K., Vezin, H. et al. (2020). Effect of thermal treatment of different petroleum fractions: characterization by in situ EPR spectroscopy. Energy and Fuels 34 (10): 12026–12032. Di Mauro, E., Guedes, C.L.B., and Nascimento, O.R. (2005). Multifrequency (X-band to W-band) CW EPR of the organic free radical in petroleum asphaltene. Applied Magnetic Resonance 29 (4): 569–575. Dickson, F.E., Kunesh, C.J., McGinnis, E.L., and Petrakis, L. (1972). Use of electron spin resonance to characterize the vanadium(IV)-sulfur species in petroleum. Analytical Chemistry 44 (6): 978–981. Djimasbe, R., Varfolomeev, M.A., Al-Muntaser, A.A. et al. (2021). Deep insights into heavy oil upgrading using supercritical water by a comprehensive analysis of GC, GC–MS, NMR, and SEM–EDX with the aid of EPR as a complementary technical analysis. ACS Omega 6 (1): 135–147. Dolomatov, M.U., Rodionov, A.A., Gafurov, M.R. et al. (2016). Concentration of paramagnetic centres at low-temperature thermal destruction of asphaltenes of heavy petroleum distillates. Magnetic Resonance in Solids 18 (1): 16101. Dutoit, C.E., Binet, L., Vezin, H. et al. (2022). Insight into the structure of black coatings of ancient Egyptian mummies by advanced electron magnetic resonance of vanadyl complexes. Magnetic Resonance 3 (2): 111–124. Eaton, G., Eaton, S., Barr, D., and Weber, R. (2010). Quantitative EPR. Vienna: Springer-Verlag. Espinosa, P.M., Campero, A., and Salcedo, R. (2001). Electron spin resonance and electronic structure of vanadyl–porphyrin in heavy crude oils. Inorganic Chemistry 40 (18): 4543–4549. Evdokimov, I., Eliseev, N., and Eliseev, D. (2004). Effect of asphaltenes on the thermal properties of emulsions encountered in oil recovery operations. Fuel 83 (7, 8): 897–903. Gafurov, M.R., Gracheva, I.N., Mamin, G.V. et al. (2018). Study of organic self-assembled nanosystems by means of high-frequency ESR/ENDOR: the case of oil asphaltenes. Russian Journal of General Chemistry 88 (11): 2374–2380. Gafurov, M.R., Volodin, M.A., Rodionov, A.A. et al. (2018a). EPR study of spectra transformations of the intrinsic vanadyl-porphyrin complexes in heavy crude oils with temperature to probe the asphaltenes’ aggregation. Journal of Petroleum Science and Engineering 166: 363–368. Gafurov, M., Mamin, G., Ganeeva, Y. et al. (2019a). Multifrequency (9 and 95 GHz) EPR study of stable radicals in asphaltenes fractions of oils and bitumen. IOP Conference Series: Earth and Environmental Science 282 (1): 012016.

References

Gafurov, M., Galukhin, A., Osin, Y. et al. (2019b). Probing the surface of synthetic opals with the vanadyl containing crude oil by using EPR and ENDOR techniques. Magnetic Resonance in Solids 21 (1): 19101. Gafurov, M., Mamin, G., Gracheva, I. et al. (2019c). High-field (3.4 T) ENDOR investigation of asphaltenes in native oil and vanadyl complexes by asphaltene adsorption on alumina surface. Geofluids 2019: 3812875. Gafurov, M.R., Ponomarev, A.A., Mamin, G.V. et al. (2020). Application of pulsed and high-frequency electron paramagnetic resonance techniques to study petroleum disperse systems. Georesursy – Georesources 22 (4): 2–14. Galtsev, V.E., Ametov, I.M., and Grinberg, O.Y. (1995). Asphaltene association in crude oil as studied by ENDOR. Fuel 74 (5): 670–673. Ganeeva, Y.M., Yusupova, T.N., and Romanov, G.E. (2011). Asphaltene nano-aggregates: structure, phase transitions and effect on petroleum systems. Russian Chemical Reviews 80 (10): 993–1008. Gizatullin, B., Gafurov, M., Vakhin, A. et al. (2019). Native vanadyl complexes in crude oil as polarizing agents for in situ proton dynamic nuclear polarization. Energy and Fuels 33: 10923–10932. Gizatullin, B., Mattea, C., Shikhov, I. et al. (2022). Modeling molecular interactions with wetting and non-wetting rock surfaces by combining electron paramagnetic resonance and NMR relaxometry. Langmuir 38 (36): 11033–11053. Gourier, D., Delpoux, O., Bonduelle, A. et al. (2010). EPR, ENDOR, and HYSCORE study of the structure and the stability of vanadyl−porphyrin complexes encapsulated in silica: potential paramagnetic biomarkers for the origin of life. Journal of Physical Chemistry B 114 (10): 3714–3725. Gracheva, I.N., Gafurov, M.R., Mamin, G.V. et al. (2016). ENDOR study of nitrogen hyperfine and quadrupole tensors in vanadyl porphyrins of heavy crude oil. Magnetic Resonance in Solids 18 (1): 16102. Kayukova, G.P., Mikhailova, A.N., Khasanova, N.M. et al. (2018). Influence of hydrothermal and pyrolysis processes on the transformation of organic matter of dense low-permeability rocks from Domanic formations of the Romashkino oil field. Geofluids 2018: 9730642. Khasanova, N.M., Gabdrakhmanov, D.T., Kayukova, G.P. et al. (2017). EPR study of hydrocarbon generation potential of organic-rich domanik rocks. Magnetic Resonance in Solids 19 (1): 17102. Lund, A., Shiotani, M., and Shimada, S. (2011). Principles and Applications of ESR Spectroscopy. Dordrecht: Springer. Mamin, G.V., Gafurov, M.R., Yusupov, R.V. et al. (2016). Toward the asphaltene structure by electron paramagnetic resonance relaxation studies at high fields (3.4 T). Energy and Fuels 30 (9): 6942–6946. Martyanov, O.N., Larichev, Y.V., Morozov, E.V. et al. (2017). The stability and evolution of oil systems studied via advanced methods in situ. Russian Chemical Reviews 86 (11): 999–1023. Murzakhanov, F.F., Mamin, G.V., Goldberg, M.A. et al. (2020). EPR of radiation-induced nitrogen centers in hydroxyapatite: new approaches to the study of electron-nuclear interactions. Russian Journal of Coordination Chemistry 46 (11): 729–737. Sadovnikova, M.A., Murzakhanov, F.F., Mamin, G.V., and Gafurov, M.R. (2022). HYSCORE spectroscopy to resolve electron–nuclear structure of vanadyl porphyrins in asphaltenes from the Athabasca oil sands in situ conditions. Energies 15 (17): 6204. Trukhan, S.N., Yudanov, V.F., Gabrienko, A.A. et al. (2014). In situ electron spin resonance study of molecular dynamics of asphaltenes at elevated temperature and pressure. Energy and Fuels 28 (10): 6315–6321. Trukhan, S.N., Yakushkin, S.S., and Martyanov, O.N. (2022). Fine-tuning simulation of the ESR spectrum – sensitive tool to identify the local environment of asphaltenes in situ. Journal of Physical Chemistry C 126 (26): 10729–10741.

57

58

2.1 EPR Spectroscopy

Vakhin, A.V., Khelkhal, M.A., Tajik, A. et al. (2021). The role of nanodispersed catalysts in microwave application during the development of unconventional hydrocarbon reserves: a review of potential applications. Processes 9 (3): 420. Yakubov, M.R., Milordov, D.V., Yakubova, S.G. et al. (2017). Vanadium and paramagnetic vanadyl complexes content in asphaltenes of heavy oils of various productive sediments. Petroleum Science and Technology 35 (14): 1468–1472. Yen, T. and Chilingarian, G. (1994). Asphaltenes and Asphalts. 1, Developments in Petroleum Science. New York: Elsevier. Yen, T. and Chilingarian, G. (2000). Asphaltenes and Asphalts. 2, Developments in Petroleum Science. New York: Elsevier. Zhang, Y., Siskin, M., Gray, M.R. et al. (2020). Mechanisms of asphaltene aggregation: puzzles and a new hypothesis. Energy and Fuels 34 (8): 9094–9107.

59

2.2 NMR-Spectroscopy and NMR-Relaxometry Ilfat Z. Rakhmatullin 1,2 , Sergey V. Efimov 1 , Marsel G. Fazlyyyakhmatov 2 , Ranel I. Galeev 2 , and Vladimir V. Klochkov 1,2 1 2

Institute of Physics, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia

In general case, a spectrum of nuclear magnetic resonance (NMR) contains several resonance signals (also called lines or peaks), which can be described using a set of spectral parameters: frequency (chemical shift), intensity, and multiplet structure. These signals appear under special conditions when a sample, most often in the liquid form, is placed in a strong magnetic field and irradiated by a radiofrequency pulsed field. The signal exists for a while, typically tens and hundreds of milliseconds. This decay time is called relaxation time. Together with the spectral parameters mentioned above, it brings information about the molecules in the sample; in particular, NMR spectroscopy is used for proving structural formulas, spatial and electronic structure of artificial or natural organic compounds, and analysis of composition of mixtures. Use of NMR spectroscopy for studying various objects in petroleum chemistry becomes more and more popular (Gao et al. 2017; Mondal et al. 2017; Alcazar-Vara et al. 2016; Smirnov and Vanyukova 2014). The NMR method opens wide opportunities in studying the structure of oil disperse systems and determining their physical and chemical characteristics. In the case of oil and petroleum products containing typically hundreds of compounds, an important feature of NMR spectroscopy is a correlation of intensities of characteristic signal groups in proton and carbon (1 H and 13 C) NMR spectra with the content of the corresponding molecular fragment types (Holzgrabe 2017; Poveda et al. 2014; Da Silva Oliveira et al. 2014). Modern NMR spectroscopy also applies two-dimensional (2D) methods, which allow to easily analyze information on the composition of the samples. Taking into account the importance of oil to the economy, there is a very important and urgent task for the adaptation of modern one-dimensional (1D) and 2D NMR (COSY, HSQC experiments) spectroscopy for the determination of oil composition. NMR relaxometry has frequently been used for many decades in oil exploration and production to determine the petrophysical properties of rock samples such as porosity (Timur 1969), pore size distribution (Yao et al. 2010), water and oil saturation (Straley et al. 1997), permeability and fluid flow (Kenyon 1992), and NMR logging (Kenyon 1997). Due to the depletion of traditional reserves, the attention of the oil industry is now increasingly directed to the study of unconventional reserves such as bitumen and kerogen. Therefore, the development of new express methods for studying crude oil and bitumen as well as their saturates, aromatics, resins, and asphaltenes (SARAs) fractions and methods for evaluating enhanced oil recovery (EOR) effectiveness are topical tasks for NMR relaxometry today.

60

2.2 NMR-Spectroscopy and NMR-Relaxometry

While NMR spectroscopy gives accurate spectra with the content of the corresponding molecular fragment types, NMR relaxometry can offer fast, low-cost estimations of SARA fraction composition using portable benchtop equipment with reasonable accuracy. In the following sections, we will describe the theory underlying the phenomenon of NMR (Section 2.2.1), and technical aspects of NMR spectroscopy and relaxometry (Section 2.2.2 and 2.2.4), and show examples of petroleum research using this approach (Section 2.2.3 and 2.2.5).

2.2.1

Phenomenon of Nuclear Magnetic Resonance

As mentioned above, NMR is observed in “strong” magnetic fields. Exactly speaking, there is a linear correlation between the angular frequency 𝜔 of the signal and the field intensity B0 , expressed as 𝜔 = 𝛾B0 . The factor 𝛾 – the so-called “gyromagnetic ratio” – is specific for every nuclear isotope. Isotopes with even number of protons and neutrons have zero nuclear magnetic moment and do not give rise to NMR. Among others, the most convenient nuclei for observing NMR are nuclei with spin I = 1/2; these include protons 1 H and carbons 13 C, which are the most important in the studies of organic materials. Thus, in the field of 11.75 T, protons will absorb radiofrequency energy at nearly 500 MHz, and nuclei of carbon-13 will have the resonance frequency of 125.8 MHz. Generally, increase in the magnetic field is advantageous due to better sensitivity of the method and better spectral resolution. In the magnetic field, the nucleus can exist in one of (2I + 1) states differing by the projection of its own magnetic moment with respect to the external field, which is described by the quantum number m. For nuclei with spin I = 1/2, there are two energy levels, with m = +1/2 and m = −1/2; transitions occur between neighbor levels that change the number m by ±1 and lead to absorption (or emission) of a quantum of electromagnetic radiation with the energy ΔE = 𝛾ℏB0 . To stimulate the transitions between the energy levels, nuclei are irradiated with a radiofrequency field with the resonance frequency given by the relation: ℏ𝜔 = ΔE

(2.2.1)

Transitions from the lower level to the upper one are accompanied by the absorption of energy and inverse transitions (to the lower level) correspond to the emission of energy. Excitation of nuclei is carried out using a short pulse with a typical duration of several microseconds, which provides almost uniform excitation of nuclei with the resonance frequency range of tens of kilohertz. After the sample has been subject to a pulse (or a special sequence of radiofrequency (RF) pulses), the electromotive force induced by the precessing magnetization in the receiving coil is acquired and digitized. The Fourier transform of this signal allows moving from one representation to another and is a standard method for analyzing the results of pulsed experiments: +∞

f (𝜔) =

∫−∞

f (t)e−i𝜔t dt

(2.2.2)

where f (t) corresponds to the spectrum in the time domain (so-called “free induction decay signal”), and f (𝜔) is the spectrum in the frequency domain, as they are usually represented. The nuclei of any given isotope have the same resonance frequency (Larmor frequency) only to the first approximation. More exactly, it depends on the diamagnetic effect induced by the electrons in the electronic shells. Since the electronic density around any atom in a molecule is distorted compared to the lonely atom in vacuum, its resonance signal is slightly shifted. This modification, which usually makes a nucleus to resonate in a stronger field than a naked nucleus without electrons would do, is described by the shielding 𝜎: 𝜔 = 𝛾B0 (1 − 𝜎)

(2.2.3)

2.2.1 Phenomenon of Nuclear Magnetic Resonance

As a result, atomic nuclei in a molecule give rise to a number of signals at different frequencies. To make the spectrum independent of the resonance frequency 𝜔0 of a given spectrometer, the dimensionless concept of chemical shift is introduced as follows: 𝜔 − 𝜔ref × 106 ppm (2.2.4) 𝛿= 𝜔ref where 𝜔ref is the reference frequency of some standard compounds added in the sample. Without this addition, the NMR signal of the solvent can be chosen as the standard (reference) signal. At the next step, the influence of adjacent spins on each other should be considered. This is the scalar coupling, which makes the energy level of a spin depend on the current energy level of one or several neighbor spins coupled to it through chemical bonds. In the spectra, this leads to splitting of a signal to a certain pattern of close peaks, so-called “multiplet.” In the spectroscopy of carbons and other heteronuclei (other than protons), scalar coupling between 13 C and 1 H would increase the number of observed peaks and decrease their intensity. For this reason, scalar coupling is usually suppressed during the experiments by special decoupling radiofrequency pulse sequences. Figure 2.2.1 demonstrates the mentioned spectral parameters. The spectrum is recorded on a spectrometer with the operating frequency of 500 MHz; thus, the chemical shift difference of 1 ppm corresponds here to the frequency difference of 500 Hz. Chemical shift of methyl protons is the smallest, and their signal appears in the right part of the spectrum. Other characteristic groups have their own typical resonance ranges. The integral intensity of the lines corresponds to the molar fraction of corresponding protons in the sample (see numbers below the chemical shift axis). Remember that due to artifacts of mathematical processing of the spectra, imperfectness of RF pulses which rotate the macroscopic magnetization vector and different relaxation times of different protons, integration of the spectral lines usually has the accuracy level of several percent. At last, the splitting of the methyl group signal to three separate peaks arises due to the scalar coupling with neighbor CH2 group; since the protons in the methyl group and the methylene group 3 (1H,1H)

1 (1H,13C)

5.2

5.1

5.0

2.0

1.9

1.8

1.7

1.6

1.5

1.4 CH3

CH2

OH H2O

4.0

3.5

3.0

2.5 integrals:

2.0

δ, ppm

3.00

4.5

2.00

1.02

Figure 2.2.1

5.0 0.31

5.5

6.0

Example of the 1 H NMR spectrum of ethanol–water mix (500 MHz).

61

62

2.2 NMR-Spectroscopy and NMR-Relaxometry

are bound to each other via three consecutive bonds (HCCH), corresponding scalar coupling is denoted as 3 J. Here, 3 J HCCH = 7 Hz. Small triplets at the sides of the main central signal belong to the molecules which have the 13 C atom in their CH3 group; total fraction of these molecules at the natural isotope abundance is only 1.1%. The distance between the two small triplets (dashed green line) shows the scalar coupling with the protons with the magnetic nucleus of carbon-13, and in this case, it turns out to be 1 J HC = 125 Hz. The signal of residual water at 𝛿 = 5.1 ppm shows no splitting because both protons in the H2 O molecule are magnetically equivalent; the lineshape in this case is described by the Lorentzian function.

2.2.2

Technical Aspects of NMR Spectroscopy

Modern high-resolution NMR spectrometers usually require liquid samples with the volume of 0.5–1.0 ml. The less viscous is the fluid, the narrower are the spectral lines. Hence, in order to achieve better resolution, viscous oil samples can be dissolved in chloroform, CCl4 , or some other solvent. The signal of the solvent can be recognized by its chemical shift and typically large intensity. NMR spectroscopy suffers from low sensitivity compared to most other experimental methods (optical spectroscopy, fluorescence, electron paramagnetic resonance (EPR), etc.). Therefore, the signal needs to be accumulated for a long time in order to average the relative noise level. Recording a one-dimensional NMR spectrum on protons may require a few minutes, while a 13 C NMR measurement can last for tens of minutes and hours. What is the reason for this? Accumulation is a repeated summation of free induction decay signals, which are excited by RF pulses every several seconds. The period of the excitation should be long enough for magnetization of the sample to return to equilibrium. To achieve full magnetization recovery of the signal with the longitudinal relaxation time of T 1 , the period between the RF pulses should exceed 5T 1 . When this condition is not fulfilled, the intensity of the signals with long relaxation time becomes underestimated. This fact should be kept in mind always when recording spectra for quantitative analysis. Another factor influencing the sensitivity is the magnetic field intensity and operating frequency. They are directly related to each other and to the energy gap between the magnetic states of the nucleus according to Eq. (2.2.1). The larger is the energy gap ΔE, the more prominent is the difference of populations of the quantum states with the magnetization projection m = +1/2 and m = −1/2, and hence, the more intensive is the resonant absorption of the RF power. The value of ΔE depends also on the gyromagnetic ratio 𝛾, and that means that nuclei with low value of 𝛾 are difficult to observe (e.g. 13 C has fourfold lower gyromagnetic value than that of protons). Therefore, much struggle has been made to achieve strong and at the same time homogeneous magnetic field; nowadays, superconducting magnets are used, and operating frequency in the proton channel in most spectrometers lies from 400 to 800 MHz. The superconducting state can only be obtained at very low temperatures and kept for a long time when the magnet is immersed in liquid helium. High-temperature superconductivity is still the lot of purely fundamental research. The central part of the magnet is surrounded by a shell of liquid nitrogen through a vacuum interlayer. Finally, the nitrogen shell is isolated from room temperature by an external vacuum interlayer. Such a system is capable of maintaining the required temperature of a superconducting magnet for a very long time, although for this, it is necessary to regularly pour liquid nitrogen and helium into the magnet. The advantage of such magnets over electromagnets, in addition to the ability to obtain high magnetic fields, is also that they do not consume energy: after starting the magnet, the current runs along the superconducting wires with practically no losses for many years.

2.2.2 Technical Aspects of NMR Spectroscopy

Recently, new types of desktop NMR spectrometers appeared on the market. They provide the same field intensity as the electromagnets of 1970s so that the proton resonance frequency is from 60 to 100 MHz, but with the modern permanent magnets, electronics, and software, they can be very convenient to solve common analytical problems in chemical industry. Since the NMR spectra of oils contain typically hundreds of signals which merge together into broad resonance lines, the loss in resolution due to the weaker magnetic field should not be a severe problem. The same is true for the problem of sensitivity: while organic chemists usually deal with the concentrations in the order of 1–10 mM or smaller when recording NMR spectra, an oil sample diluted with a deuterated solvent for the sake of field stabilization and diminishing the viscosity still has much greater concentration of the main components. Furthermore, relaxation time also depends on the magnetic field. It depends both on the resonance frequency 𝜔 and on the characteristic time of molecular motion called correlation time 𝜏 c . In modern high-resolution spectrometers with strong magnetic field relaxation time of 13 C nuclei may reach several seconds, which slows down significantly quantitative spectral measurements. Use of desktop spectrometers may be beneficial in this case. The group analysis method can be used to determine the oil composition, wherein the composition is expressed as relative quantities of different types of compounds such as aromatic molecules, olefins, alkanes, and cycloalkanes, and their numerous isomeric analogues. Oil fractions can be described as the average ratio of structural groups (structural-group analysis). It means that some “average” molecule of mixture is defined by a certain set of numbers. 1 H NMR spectroscopy is a relatively fast method, which does not require complex sample preparation. Accessible spectral range of proton resonances is not wide, and hence, little information can be obtained from direct comparison of 1 H spectra. 1 H NMR spectra of crude oils contain a great number of signals reflecting their chemical complexity (Gupta et al. 1986; Kushnarev et al. 1989; Hasan et al. 1983). Therefore, to exploit the full information content of the NMR data acquired on complex systems, different multivariate data analysis methods were developed (Alam and Alam 2004). Statistical analysis may be successfully used to predict important properties of oil (API gravity, carbon residue, wax appearance temperature, and basic organic nitrogen) Table 2.2.1 Distribution of 1 H NMR chemical shifts of functional groups defining the composition of oil samples. 1

H NMR chemical shifts range (ppm)

Organic functional groups

0.5–1.0

γ-CH3 – groups and some CH– and CH2 – groups in naphthenic fragments

1.0–1.7

𝛽-CH2 – and some 𝛽-CH– groups in aromatic compounds

1.7–1.9

𝛽-CH–, CH2 – groups in hydroaromatic compounds

1.9–2.1

Methyl groups (CH3 ) in 𝛼-olefins

2.1–2.4

Methyl groups (CH3 ) in 𝛼-position to aromatic carbons

2.4–3.5

CH– and CH2 – groups in 𝛼-position to aromatic carbons

3.5–4.5

Bridging CH2 – group (diphenylmethane)

4.5–6.0

Protons of olefin groups

6.0–7.2

Protons of single-cycle arenes

7.2–8.3

Di-, three- and tetra-aromatic protons

8.3–8.9

Some three- and tetra-aromatic protons

8.9–9.3

Protons of some four-cycle arenes

63

64

2.2 NMR-Spectroscopy and NMR-Relaxometry

based on a large set of studied samples (Duarte et al. 2016). Use of 1 H NMR spectroscopy on neat crude oil samples combined with partial least squares (PLSs) modeling has high potential to predict crude oil properties (Masili et al. 2012). If 13 C data are added, much more information on chemical composition can be extracted; however, this requires more experimental time. NMR data can be combined with other experiments (mass-spectrometry, elemental analysis, etc.); an example of such a study using Monte Carlo simulation to reconstruct molecular composition of gasoils and heavy fractions is presented by Hudebine and Verstraete (2004) and Verstraete et al. (2010). Organic compounds contained in rocks can also be studied in solid phase by NMR using the magic-angle spinning method; as an example, studies of kerogen by Bushnev et al. (2010) and Bushnev and Burdel’naya (2012) may be mentioned. Both 1 H and 13 C NMR spectra of oil samples in solution already have been described in literature by Kalabin et al. (2000). Corresponding detailed information is presented in Tables 2.2.1 and 2.2.2.

Table 2.2.2 samples.

Distribution of 13 C NMR chemical shifts of functional groups defining the composition of oil

13 C NMR chemical shifts range (ppm)

Organic functional groups

11.0–12.5

γ-CH3 – groups and some CH– and CH2 – groups in aromatic fragments, CH3 – group in ethyl-substituted cyclohexane

12.5–15.0

𝛾-CH3 – (and more distant) methyl groups of aromatic cycle; CH3 – group, shielded by two neighboring aromatic rings

15.0–18.0

𝛽-CH3 – substituent in ethylene group

18.0–20.5

𝛼-CH3 – group, shielded by one aromatic; some 𝛼-CH3 – and CH2 – groups in hydroaromatic and naphthenic fragments

20.5–22.5

𝛼-CH3 – group, unshielded by aromatic; some 𝛼-CH3 – and CH2 – groups in hydroaromatic and naphthenic fragments

22.5–24.0

𝛾-CH2 – and CH3 – groups; 𝛽-CH2 – groups in unsubstituted tetralin structures

24.0–27.5

Methylene (CH2 ) groups in naphthenic fragments; 𝛼-CH– and 𝛽-CH2 – groups in a propyl and indan fragments; 𝛽-CH3 – group in isopropyl

27.5–37.0

Methylene (CH2 ) groups, do not neighboring with methine (CH) group in alkyl compounds; methylene (CH2 ) group in cycle

37.0–60.0

Methine (CH) group in alkyl fragments; CH and CH2 alkyl groups of naphthenic fragments, adjacent to CH group

108.0–118.0

Olefin fragments

118.0–129.5

Protonated arenes

129.5–133.0

Internal aromatic carbon atoms

133.0–135.0

Methyl substituted arenes

135.0–138.0

Arenes, substituted by naphthenes

138.0–160.0

Alkyl substituted (except for methyl substituted) arenes; heteroatomic (N, O, S) arenes

165.0–175.0

Ester or amide carboxy carbon atom

170.0–182.0

Acid carboxy carbon atom

182.0–192.0

Quinone carboxy carbon atom

195.0–205.0

Aldehyde carbonyl carbon atom

202.0–220.0

Ketone carbonyl carbon atom

2.2.3 NMR Spectroscopy in Study of Oil Samples and Their Individual SARA Fractions

Another possibility to enhance NMR signal in heavy oils is dynamic nuclear polarization (DNP) by applying methods of EPR to the indigenous (Gizatullin et al. 2018, 2019; Alexandrov et al. 2014; Poindexter 1958; Peksoz et al. 2010) or artificially introduced (Chen et al. 2019) paramagnetic centers. It was shown that 1 H NMR signal enhancement depends not only on the type, electronic properties, concentration of paramagnetic species, and viscosity of oils but also on the SARA fraction (Gizatullin et al. 2018, 2019) that can be selected in situ by choosing an appropriate repetition time of NMR experiment. To the best of the authors’ knowledge, all the oil DNP measurements known so far are done on 1 H nuclei. Obviously, as in the case of other NMR/DNP applications for biological investigations and material science (Corzilius 2020), one should expect the rapid development of 13 C DNP for oil studies. In Section 2.2.4, we present an example of NMR study of oil, including heavy oil, and SARA fractions carried out on a high-field NMR spectrometer (proton operating frequency of 700 MHz).

2.2.3 NMR Spectroscopy in Study of Oil Samples and Their Individual SARA Fractions Some examples of NMR spectra of crude oil and the way of processing are described below. Figures 2.2.2, 2.2.3, and 2.2.4 demonstrate one-dimensional spectra of oil measured on 1 H and 13 C. The sample was diluted with CCl4 . Proton spectra provide limited information because resonance lines of organic fragments with the similar structure merge in broad signals and become difficult to distinguish. Generally, the rightmost signals with the chemical shift of ∼1 ppm or less belong to methyl groups; the next peak to the left corresponds to methylene groups. The signals beyond ∼6.5 ppm belong to protons of aromatic cores; their signals are shifted so far downfield because of the ring current induced in the cyclic aromatic systems. The central part of the spectrum shown

CH2 CH3

Haliphatic

1.4

1.2

1.0

0.8 ppm

Haromatic

8.5

10

8.0

7.5

8

7.0 6.5 Haromatic 6

Haliphatic

ppm H2O*

4

2

ppm

Figure 2.2.2 1 H (700 MHz) NMR spectrum of an oil sample in CCl4 (insets show zoomed signals of aromatic and aliphatic protons). External standard (heavy water D2 O) in a glass capillary was used for magnetic field stabilization. Residual water signal in the capillary is marked by an asterisk (*).

65

2.2 NMR-Spectroscopy and NMR-Relaxometry

CH2

CH CH3

Aromatic cores (amplified)

160

150

140

130

110 ppm

120

40 35 30 25 20 15

ppm

CCl4 Aromatic cores

10 ppm 2.302

40 30 20 5.412

50 1.973

70 60

0.464

0.978

0.242

190 180 170 160 150 140 130 120 110 100 90 80 1.000

Figure 2.2.3 13 C NMR spectrum of an oil sample in CCl4 . Inset shows increased NMR signals of aromatic and aliphatic carbons. Integrated regions are marked in red.

β – CH3 CH3 CH CH3 CH aromatic

C aromatic

CH2 CCl4

66

140

120

100

80

60

40

20 ppm

Figure 2.2.4 13 C APT NMR spectrum of an oil sample in CCl4 . Magnetic field is the same as in Figures 2.2.2 and 2.2.3; the resonance frequency of carbon-13 is 176 MHz.

2.2.3 NMR Spectroscopy in Study of Oil Samples and Their Individual SARA Fractions

in Figure 2.2.2 contains little number of signals, but here we expect to see peaks of protons in unsaturated fragments (–CH=CH–) or the signal of admixture water. 13 C NMR spectra contain numerous distinguishable signals, which can be assigned to different typical regions and thus give information on the fractions of aromatic, primary (CH3 ), secondary (CH2 ), and other types of carbon atoms (Figures 2.2.3 and 2.2.4). The spectra shown here are obtained with proton decoupling; usually, they are denoted as 13 C-{1 H}. The APT experiment (attached proton test) in addition separates directly the signals of carbons with even and odd numbers of attached protons and shows them as peaks looking up or down. The combination of 1 H and 13 C NMR spectra allows estimating the average length and branching of aliphatic chains. 1 H and 13 C NMR signals of the oil samples are assigned following the literature data (Rakhmatullin et al. 2017). To identify the molar fractions of functional groups in the samples, integral intensities of 1 H NMR signals are normalized according to the number of protons in each group. Thus, signal intensities of methyl groups are divided by 3; methylene groups, by 2. With this recalculation, integral intensities of signals in 1 H NMR spectra can be represented as proportions of functional groups. Figure 2.2.5 shows an example of the so-called “two-dimensional” NMR spectrum 1 H, 13 C-HSQC. Each signal in this spectrum belongs to a CH group and shows chemical shifts of two n nuclei at the same time; quaternary carbons are not observed here. Two main signal groups are seen: aliphatic signals (1 H: [1–4 ppm], 13 C [10–60 ppm]) and aromatic ones (1 H: [6–8 ppm], 13 C [115–135 ppm]). These spectra suffer less from signal overlap, than one-dimensional spectra, and also can be used as “fingerprints.” The correctness and effectiveness of NMR techniques for determining SARA composition analysis without oil fractioning are still not clear. A comparative study using high-resolution 13 C NMR spectra in the field B0 = 16.4 T (13 C NMR frequency of 176 MHz), and the results of SARA fractioning for four various heavy oil samples with the viscosities ranging from 100 to 50 000 mPa⋅s were carried out (Rakhmatullin et al. 2020). The presence of all major hydrocarbon components H2O Haromatic

1H F1 [ppm]

13C

Haliphatic

50

Caliphatic

100

CCl4

150

Caromatic

10 Figure 2.2.5

1

8

6

4

2 F2 [ppm]

H,13 C-HSQC (700 MHz) NMR spectrum of an oil sample in CCl4 .

67

68

2.2 NMR-Spectroscopy and NMR-Relaxometry

both in crude oil and in each of its fractions was established quantitatively by NMR spectroscopy. Contribution of SARA fractions in the aliphatic (10–60 ppm) and aromatic (110–160 ppm) regions of the 13 C NMR spectra was identified. Quantitative fractions of aromatic molecules and oil functional groups were determined. Obtained results show the feasibility of 13 C NMR spectroscopy for the express analysis of oil to follow the oil treatment processes. Information given by SARA analysis can be useful to predict compatibility and stability of blends of crude oils and to anticipate problems of sedimentation during storage and transportation (Aske et al. 2002). Some studies have revealed that NMR can be used for predicting SARA composition of crude oils (Molina et al. 2010; Sanchez-Minero et al. 2013). NMR results were used to obtain the concentration of aromatic hydrogen and aromatic carbon, from which aromaticity factors were computed and correlated to SARA analysis. This allowed developing correlations for predicting SARA composition of crude oils with a wide range of API gravity (10–33∘ ) (Sanchez-Minero et al. 2013). The attempts to find a correlation between the SARA analyses and NMR measurements are known for a long time; however, these results cannot be generalized for all oil samples. It is mandatory to incorporate a wider range of experimental data of crude oils characterization and their respective fractions (Corzilius 2020). Four heavy (viscous) oil samples of various origin and their four SARA fractions (totally 20 samples) were selected for the study. A list of the studied samples along with the viscosities of the initial oil is presented in Table 2.2.3. The results of SARA analysis of studied oil samples are reported in Table 2.2.4. The 13 C NMR spectra of crude oils 1–4 were recorded in solutions in deuterated chloroform CDCl3 . As an example, the spectra of the SARA fractions of sample 1 are compared in Figure 2.2.6. The broadening of resonance lines in the aromatic region of the spectrum in the transition from the light oil sample 1 to the heavier oil samples 2–4 is observed, which agrees with the increase in the viscosity. Molar content of various carbon groups in samples 1–4 were determined by integration of the corresponding regions in the 13 C NMR spectra. Quantitative analysis of the composition of the oil samples (proportions of primary, secondary, tertiary, quaternary, and aromatic carbon atoms) studied by NMR showed that as the oil viscosity increases, a decrease in the parameters Cp and Csq Table 2.2.3

Studied oil samples and the viscosities of the initial oil species. Viscosity (mPa⋅s)

Sample numbera)

1 (fractions: 1s, 1ar, 1r, 1as) 2

106 1430

3

2420

4

49 700

Origin of oil

Iraq Ambar gatur (Turkmenistan) Ashal’cha (Republic of Tatarstan, Russia) Cuba

a) s – saturated compounds, ar – aromatics, r – resins, as – asphaltene fraction.

Table 2.2.4

SARA analysis (%) of studied oil samples.

Sample number

1

2

3

4

Saturates (%)

59.6

73.7

26.2

31.0

Aromatics (%)

26.7

14.3

40.6

39.2

Resins (%)

12.1

9.8

28.5

14.2

1.6

2.2

4.7

15.6

Asphaltenes (%)

2.2.3 NMR Spectroscopy in Study of Oil Samples and Their Individual SARA Fractions

1s

1ar

1r

1as CDC13 160

140

120

100

80

60

40

20 ppm

Figure 2.2.6 13 C NMR spectra of oil samples 1s–1as. Labels: s – saturated compounds, ar – aromatics, r – resins, as – asphaltene fraction.

and increase in Ct and Car are observed. Moreover, the concentration of aromatic groups (Car ) in heavy oil sample 4 increases almost twice as compared to the less viscous oil sample 1. At first glance, the spectra can be divided into two main regions: the low-field part of 120–150 ppm, containing the signals of aromatic compounds, and the high-field part. In the latter, the rightmost peaks correspond to methyl protons in acyl chains; methylene protons have chemical shifts from 20 to 40 ppm. Signals of methylene substituents at aromatic rings are shifted nearly to 20.5–22.5 ppm; olefins give rise to peaks at 108–118 ppm. Very sharp peaks (e.g. in the resins fraction in Figure 2.2.6) correspond to nuclei with relatively long transverse relaxation time T 2 ; these belong to low-molecular-weight components. On the contrary, very broad signals such as those of asphaltenes (the bottom spectrum in Figure 2.2.6) give the evidence that these molecules or molecular aggregates are extremely large (tens of kDa). Evidently, the quantitative parameters obtained by NMR spectroscopy change according to two criteria: the gravity of the fraction (growing from saturates to asphaltenes) and viscosity of the sample (an increase from sample 1 to sample 4; see Figure 2.2.7). Thus, for the Cp value, there is a general decrease in values nearly twofold as the gravity of the fraction increases. However, analysis of the dependence of Cp on the sample viscosity among individual fractions shows an ambiguous picture: for saturates and asphaltenes, there is a slight decrease in the values of Cp , while for the aromatic and resin fractions, the values of Cp slightly increase. A similar trend in changes in individual fractions is observed for the parameter Csq . However, the reduction in Csq for saturates and asphaltenes as the viscosity of the sample increases occurs more sharply than for Cp . Analysis of the parameter Ct showed that for each fraction, its value increases strongly as the viscosity of the samples increases, and when the gravity of the fraction increases, there is an increase in Ct values for samples 1 and 4 and a decrease for samples 2 and 3. The behavior of Car parameter as the gravity of the fraction increases is exactly opposite to Cp : its values increase sharply. However, its values within individual fractions as the sample viscosity increases have a general tendency to decrease.

69

70

2.2 NMR-Spectroscopy and NMR-Relaxometry 18

70

16 14 12 10 8 6 4

60 50

Cp(1) Cp(2)

Csq(1)

40

Cp(3)

30

Cp(4)

20

2

Csq(2) Csq(3) Csq(4)

10

0 1s,2s,3s,4s

1ar,2ar,3ar,4ar

1r,2r,3r,4r

0

1as,2as,3as,4as

1s,2s,3s,4s 1ar,2ar,3ar,4ar

35 30 25

Ct(1)

20 15

Ct(2) Ct(3)

10

Ct(4)

5 0 1s,2s,3s,4s 1ar,2ar,3ar,4ar

Figure 2.2.7

1r,2r,3r,4r

1as,2as,3as,4as

1r,2r,3r,4r 1as,2as,3as,4as

50 45 40 35 30 25 20 15 10 5 0

Car(1) Car(2) Car(3) Car(4)

1s,2s,3s,4s

1ar,2ar,3ar,4ar

1r,2r,3r,4r

1as,2as,3as,4as

Diagram of changes in the measured values of carbon group fractions Cp , Csq , Ct , and Car .

Though, as an exception to this rule, there is a sharp increase in the values of the Car parameter in sample 3 for saturates and asphaltene fractions. Also, there is a very sharp drop of Car value (down to 4.3%) in sample 2 for aromatics fraction.

2.2.4

Technical Aspects of NMR Relaxometry

The main direction of NMR spectroscopy methods development is a continuing increase in the magnetic field strength with an optimum of homogeneity that have been aspired together with progress in pulse sequence design and data processing. It is known that the signal amplitude is strongly dependent on magnetic field strength (Eccles 2017). Commercial spectroscopy systems are available with magnetic field up to about 28 T. These instruments are usually designed for use in laboratories, for example for chemical structure analysis in mainly liquid samples (Rudszuck et al. 2021). These spectroscopy systems use superconducting magnets requiring supply and containment of cryogens (liquid helium and nitrogen). Installing and technical service of such systems in industrial environments are also costly, which keeps interest in low-field technology. Low-field permanent magnets can provide a suitable compromise between magnetic field strength and experimental versatility for installation in industrial environments and the investigation of heterogeneous materials (Mitchell et al. 2014). Nowadays, the experimental bandwidth of low-field nuclear magnetic resonance (LF-NMR) ranges from diffusion, relaxation, or time-domain nuclear magnetic resonance (TD-NMR), and imaging (LF-MRI) up to spectroscopy (LF-NMR spectroscopy). Magnetic fields range from a few mT up to 2 T (Rudszuck et al. 2021). The first magnets employed to detect NMR in condensed matter were electromagnets (Blümich 2019), but currently this is the range of field strengths that are quite conveniently produced by permanent magnets. NdBFe and SmCo magnets are the two most common materials in use, and magnets in a variety of shapes and sizes are readily available (Eccles 2017). Current LF-NMR spectrometers are often based on Halbach magnets (Rudszuck et al. 2021). Halbach design uses a cylindrical array of magnets to produce a horizontal field inside a vertical bore which can then be conveniently converted into an NMR probe with the addition of a vertically oriented B1 solenoid and sample (Eccles 2017). For a long time and currently, relaxation and diffusion NMR signals are being analyzed in the time domain (TD-NMR) by fitting model functions to experimental signal-decay curves to extract

2.2.5 NMR Relaxometry in the Study of Oil Samples, Oil-Saturated Rock Samples, and Their Individual SARA Fractions

the relaxation times and self-diffusion coefficients. In 1972, Bruker started marketing the Minispec p20, a compact tabletop relaxometer, and not much afterward, Oxford Instruments followed with a related permanent magnet-based tabletop instrument, the Newport Analyzer, after acquiring Newport Instruments in 1975. This instrument was subsequently succeeded by the MQC series (Blümich 2019). Both series of instruments are still actively produced. In Russia, for a long time, similar tabletop relaxometers Proton-20 (Radiotechnological Equipment Plant, Belarus, Gomel) and later Chromates-Proton 20M, manufactured by SDO Chromatec JSC, were produced. At present, the Spin Track relaxometer manufactured by Resonance Systems LLC is presented on the market. In TD-NMR measurements of transverse relaxation, often the Carr Purcell Meiboom Gill (CPMG) pulse sequence is used for differentiation of substances with different molecular mobility, as well as aggregate state and morphology. In addition, different states of a substance can also be discriminated, for example, free and bound water (Rudszuck et al. 2021). Although simple liquids typically show single exponentially decaying relaxation behavior, more complex materials produce multiexponential data, which can be analyzed using inverse Laplace transform (ILT) algorithms to produce relaxation spectra (Eccles 2017). One major disadvantage of this approach is its numerical instability for signals with small signal-to-noise ratios, which are almost always present in TD-NMR relaxation data. A regularization parameter is often user-specific and therefore more or less subjective. Especially in the case of automated data processing, the use of the ILT is difficult with respect to repeatability and reliability. Comparability of data and physicochemical interpretation are sometimes limited as well (Rudszuck et al. 2019). In Section 2.2.5, we present an example of TD-NMR study of oil, oil-saturated rock samples, and SARA fractions carried out on a LF-NMR relaxometer Chromatec-Proton 20M.

2.2.5 NMR Relaxometry in the Study of Oil Samples, Oil-Saturated Rock Samples, and Their Individual SARA Fractions The LF-NMR relaxation allows one to estimate SARA composition of crude oils with wide range of API gravity (from 4.6 to 42 API gravity data have a high correlation with standard chemical analysis by using ASTM D4124-09) (Volkov et al. 2021). This method has many advantages as fast and low-cost technique that requires no more than 15 minutes to obtain a result and no more than 1 mg of oil sample for analysis. Such significant possibilities make this investigation reasonable for implementation to assess efficiency of EOR methods. The oil fraction composition studies by the NMR relaxation started with the work of Mirotchnik et al. (2001). Based on a study of many various hydrocarbons and complex mixtures, they showed that the spectrum of transverse relaxation times (T 2 ) of oil components can be divided into frames in the range from 5 × 10−2 to 3 seconds. It was noted that for hydrocarbons in conventional oil, the relaxation spectrum can be divided into frames (ranges), where the longest relaxation frame corresponds to C4 –C15 saturates, followed by aromatics, olefins, waxes-resins, and asphaltenes, in descending relaxation range order. The frames can be analyzed as follows for heavy oil and bitumen samples, and the shorter relaxation frame represents the asphaltenes, followed by the second shortest relaxation frame, which represents the resins, followed by the longer relaxation frame, which represents the saturated compounds. In addition, a preliminary review of the full NMR decays of the transverse magnetization of protons of crude oils in situ makes it easy to determine the type of samples analyzed by the shape and duration of these curves. Mathematical analysis of these curves makes it possible, without separating, changing the chemical composition, or destroying of oil samples, to evaluate the content of high molecular weight compounds (asphaltenes and

71

2.2 NMR-Spectroscopy and NMR-Relaxometry

100

Relative amplitude (a.u.)

72

10

1

0,1

0,01

0

Figure 2.2.8

1000

2000

3000

4000 5000 Time (ms)

6000

7000

8000

N°1 N°2 N°3 N°4 N°5 N°6 N°7 N°8 N°9 N°10

Typical CPMG decays of crude oil (from light (1) to extra heavy (10)).

resins) with short T 2 times, and low molecular weight (light fractions: saturates and aromatics) characterized by longer relaxation times T 2 and high mobility of their molecules (Mirotchnik et al. 2001). The longest frame represents the aromatics (Freed 2007; Mirotchnik et al. 2001). It should be noted that, for measuring short relaxation time, it is required to have NMR-relaxometer with a relatively high frequency, which can reliably measure the short relaxation times of the microsecond range corresponding to high-viscosity components such as resins and asphaltenes. Unfortunately, the addition of a solvent violates the information purity of the samples. The main advantage of the NMR method is its noninvasiveness, which is practically lost. Even though the extraction of components from the sample does not occur and all oil components remain in the mixture, the relationship between them under the influence of the solvent changes significantly. Our investigations correspond to a technique that is based on obtaining complete information from the free induction decay (FID) data with the determination of the signal amplitude at the initial time. Based on this idea, Volkov et al. (2018) developed a simple NMR relaxation technique for heavy oils, which uses all experimental points of FID and SPMG signal to numerically determine its heterogeneous components. This technique made it possible to quantitatively determine the content of not only asphaltenes but also resins, aromatics, and saturates in heavy oils. In addition, the technique was tested on the wide range from light to extra heavy oils and revealed the high correlation of the resins and asphaltenes content obtained by this technique (R2 = 0.91 and 0.98, respectively) with the data on SARA ASTM D4124-09 (Volkov et al. 2021). Figure 2.2.8 shows an example of CPMG decays of crude oil from light to extra heavy. Accurate quantitative estimates of the NMR relaxation decay can be obtained through mathematical processing, which is carried out using a multistage fitting based on model functions. The FID signals of core samples are best approximated by Gaussian and exponential functions for the solid-phase component and the sum of exponential and Voight functions for the liquid-phase components: ( )2 )⎤ ( ⎞ ⎡ ⎛ t t ⎟+f ⎥ ⋅ exp − AFID (t) = AS0 ⋅ ⎢(1 − fSam ) ⋅ exp ⎜− ⎢ ⎜ T2Sg ⎟ Sam T2Sam ⎥ ⎦ ⎣ ⎠ ⎝ ( ( ) ) )2 ) ( ( t t t + Ale0 ⋅ exp − + Alm0 ⋅ exp − ⋅ exp − (2.2.5) T2le T2lm T2ll

2.2.5 NMR Relaxometry in the Study of Oil Samples, Oil-Saturated Rock Samples, and Their Individual SARA Fractions

where AS0 is the amplitude of the solid phase (asphaltenes), Ale0 and Alm0 are the amplitudes of the liquid-phase (saturates, naphthenic aromatics, and resins) components; T 2Sg is the relaxation time of the crystalline part of the solid-phase component; T 2Sam is the relaxation time of the amorphous part of the solid-phase component; T 2le is the relaxation time of the short exponential function liquid-phase component; T 2lm is the relaxation time of the long Voight function liquid-phase component; T 2ll is the relaxation time taking into account the field inhomogeneity; f Sam is the proportion of the amorphous part of the solid-phase component. The influence of the magnetic field inhomogeneity on the FID rate of the liquid-phase component is considered using the Voight function. Voight function is the multiplication of exponential and Gaussian functions. The relaxation time T 2ll is directly determined by the magnetic field inhomogeneity in the sample. The most adequate model of the CPMG signal is the sum of exponential functions. Each exponential function describes a separate component. The number of exponential functions is determined by successive subtraction of long-time components: ) ( n ∑ t (2.2.6) Ai ⋅ exp − ACPMG (t) = T2i i=1 where ACPMG (t) is the measured CPMG signal, the indices i = a, b, c, d are arranged in decreasing of the relaxation time T 2i . Volkov et al. (2018) and (2021) showed that the exponential components of model (2) for oil and oil products are related to its SARA fractions. The component with the longest relaxation time T 2a represents the proton population of saturates fraction with amplitude Aa . The component with T 2b represents the proton population of the naphthenic aromatics with amplitude Ab ; T 2c and T 2d are groups of benzene and alcohol-benzene resins with amplitudes Ac and Ad , respectively. The fraction content is determined from the ratio of the fraction amplitude (Ai ) to the total amplitude of the NMR signal at the initial time (A0 ). The total amplitude is determined from the FID signal (Eq. (2.2.5)). The total amplitude of the NMR signal is equal to the sum and represent Relative Hydrogen Index (RHI) of the sample: A0 = AS0 + Al0

(2.2.7)

The total amplitude of the SPMG signal (ACPMG ) is equal to the amplitude of the liquid-phase component of the FID signal (Al0 ): ACPMG = Ad + Ac + Ab + Aa = Al0 = Ale0 + Alm0

(2.2.8)

Thus, both the joint registration of FID and CPMG signals and mathematical processing of solid-phase and liquid-phase components are necessary. The good agreement of data obtained by conventional SARA and LF-NMR relaxation technique for high molecular weight fractions (resins and asphaltenes) and for the sum of low molecular weight fractions (saturates + aromatics) shows that proposed method has a great potential for studying the composition of crude oils, including light, medium, heavy, and extra heavy oils. However, proposed method has some advantages comparing with ASTM D4124. It is less timeand material-consuming and gives opportunity to determine SARA-composition of crude oils in situ without any chemical or physical treatment of the sample. Recently, the NMR technique of two-dimensional T 1 –T 2 mapping has been widely used (Fleury and Romero-Sarmiento 2016; Mehana and El-monier 2016; Song and Kausik 2019). The 2D T 1 –T 2 maps technique is based on detecting the echo signal in the CPMG series with subsequent ILT. The studies of T 1 –T 2 signatures in combination with various analytical tools: extraction, drying, demineralization, separation of bitumen and kerogen, thermogravimetry, and pyrolysis are undertaken to divide 2D T 1 –T 2 maps into areas and quantitatively measure and type the proton population

73

74

2.2 NMR-Spectroscopy and NMR-Relaxometry

(Romero-Sarmiento et al. 2017; Khatibi et al. 2019; Li et al. 2020; Ma et al. 2020). Yang et al. (2020) used a combined Gaussian and exponential inversion method to fit the combined FID and CPMG data. The numerical results were then compared with the bitumen and moisture content of the respective samples. Using an approximate mathematical apparatus based on the ILT to describe the relaxation decay of the solid and liquid parts of the proton population gives only semiquantitative estimates and has several significant limitations discussed in the last paragraph of Section 2.2.4. In addition, solid-phase components with short T 2 are not detected due to the NMR relaxometry dead time. As a result, the contributions from solid-phase components with short echo signals are underestimated or absent in T 1 –T 2 maps. We upgraded the method for assessing SARA fractions for crude oils and implemented it for bitumen in rock formations. The main idea of experiments is for the estimation of SARA of oil samples; however, parameters of signal registration and technique of result interpretation are distinguished (Galeev et al. 2022). As an example, bitumen-saturated carbonate rocks, weakly consolidated, were chosen, where the amount of bitumen and its group composition could not be determined only by standard methods: extraction (Mueller and Philip 1998; Tikhonova et al. 2019), Rock-Eval pyrolysis (Espitalié et al. 1977; Behar et al. 2001; Carvajal-Ortiz and Gentzis 2015), thermogravimetric analysis (TGA) (Marshall et al. 2002; Kok et al. 2017; Ganeeva et al. 2020), and elemental analysis (CHNS analysis) (Zauer 2018) due to methods limitations. Any mechanical and chemical action led to the loss of light hydrocarbons or the destruction of the mineral matrix with asphaltenes blocking the pore space. The method was demonstrated on 20 samples collected from the productive carbonate formations (Galeev et al. 2022). The FID and CPMG data combination showed the possibility of estimating the SARA fractions directly in the rock, without resorting to destructive methods. NMR data on the bitumen content were compared with TGA and pyrolysis (R2 = 0.99 and 0.97, respectively). It was found by focused ion beam scanning electron microscope (FIB-SEM) that after extraction, the part of the bitumen remains in the pores, which distorts the results of SARA analysis according to ASTM D4124. The proposed NMR method can be used to calculate the rock saturation in situ on the oilfield, as well as to estimate the composition of the fractions when selecting the optimal EOR. Example for assessing efficiency was carried out for four types of EOR (Steam, Steam + solvent, Steam + solvent + catalyst, Hot water + solvent injections). Rock samples before and after injections were studied by LF-NMR relaxation (rocks after implementation of EOR technologies that were divided into four zones from an area that the first undergoes the treatment – 1 to the last one – 4 to have deep comprehension about processes of oil displacement). Typical decays of CPMG pulse sequence are presented in Figure 2.2.9. Short decays are corresponded to few amounts of liquid part (aromatic and saturated hydrocarbons) since movable molecules of oil are displaced by EOR. In this investigation of oil displacement under the influence of steam with a catalyst, the content of asphaltenes in the displaced oil decreases more than two times compared with the injection of steam and steam with a solvent, which indicates the thermal transformation of oil in the presence of a catalyst. When exposed to hot water, the effect of thermal transformation of oil is less pronounced, which is associated with a decrease in temperature. The SARA fractions distribution of displaced oils is illustrated in Figure 2.2.10. In a series of experiments at a certain temperature in the control experiment on steam injection, oil displacement was equal to 27.9%. The presence of a solvent increases oil displacement by 9.3–37.2%. In an experiment that includes a combination of steam injection and solvents + catalyst, oil displacement made up 45%, which indicates an increase in the displacement efficiency when using a catalyst solution. This fast and accurate method allows to determine the best type of oil displacement for certain rock samples by using LF-NMR relaxation technique.

References

Relative amplidude (a.u.)

100

1 zone 3 zone Initial rock sample

2 zone 4 zone Produced oil

10

1

0

50

Figure 2.2.9

100 Time (ms)

150

200

Typical CPMG decays of rock samples and produced oil.

SARA composition of extracted and produced oils 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

12,7

23,6

26,1

23,5

33,9

34,9

43,7

42,5

39

32,8

Steam injection

Steam + solvent injection

Extracted oil

43,2

Aromatic + saturated hydrocarbons

Figure 2.2.10

25,2 42,7

44,1

Resines

Steam + solvent + catalyst injection

32,1 Hot water + solvent injection

Asphaltenes

SARA composition of extracted and produced oils.

References Alcazar-Vara, L.A., Zamudio-Rivera, L.S., and Buenrostro-González, E. (2016). Effect of asphaltenes and resins on asphaltene aggregation inhibition, rheological behaviour and waterflood oil-recovery. Journal of Dispersion Science and Technology 37: 1544–1554. Alexandrov, A.S., Archipov, R.V., Ivanov, A.A. et al. (2014). The low-field pulsed mode dynamic nuclear polarization in the pentavalent chromium complex and crude oils. Applied Magnetic Resonance 45: 1275–1287. Alam, T.M. and Alam, M.K. (2004). Chemometric analysis of NMR spectroscopy data: a review. Annual Reports on NMR Spectroscopy 54: 41–80. Aske, N., Kallevic, H., and Sjöblom, J. (2002). Water-in-crude oil emulsion stability studied by critical electric field measurements. Correlation to physico-chemical parameters and near-infrared spectroscopy. Journal of Petroleum Science and Engineering 36: 1–17. Behar, F., Beaumont, V., De B. Penteado, H.L. (2001). Rock-Eval 6 technology: performances and developments. Oil & Gas Science and Technology 56: 111–134. Blümich, B. (2019). Low-field and benchtop NMR. Journal of Magnetic Resonance 306: 27–35. Bushnev, D.A. and Burdel’naya, N.S. (2012). Organic matter in Silurian rocks from the Chernov uplift. Geochemistry International 50: 683–691.

75

76

2.2 NMR-Spectroscopy and NMR-Relaxometry

Bushnev, D.A., Burdel’naya, N.S., Mokeev, M.V., and Gribanov, A.V. (2010). Chemical structure and 13 C NMR spectra of the kerogen of carbonaceous rock masses. Doklady Earth Sciences 430: 210–213. Carvajal-Ortiz, H. and Gentzis, T. (2015). Critical considerations when assessing hydrocarbon plays using Rock-Eval pyrolysis and organic petrology data: data quality revisited. International Journal of Coal Geology 152: 113–122. Chen, J., Feng, J., Chen, F. et al. (2019). Characterizing oils in oil-water mixtures inside porous media by Overhauser dynamic nuclear polarization. Fuel 257: 116107. Corzilius, B. (2020). High-field dynamic nuclear polarization. Annual Review of Physical Chemistry 71: 143–170. Da Silva Oliveira, E.C., Neto, A.C., Junior, V.L. et al. (2014). Study of Brazilian asphaltene aggregation by nuclear magnetic resonance spectroscopy. Fuel 117: 146–151. Duarte, L.M., Filgueiras, P.R., Silva, S.R.C. et al. (2016). Determination of some physicochemical properties in Brazilian crude oil by 1 H NMR spectroscopy associated to chemometric approach. Fuel 181: 660–669. Eccles, C. (2017). Low field NMR methods and applications. In: Encyclopedia of Spectroscopy and Spectrometry (ed. J.C. Lindon, G.E. Tranter, and D.W. Koppenaal), 611–626. Oxford: Academic Press. Espitalié, J., Laporte, J.L., Madec, M. et al. (1977). Méthode rapide de caractérisation des roches mètres, de leur potentiel pétrolier et de leur degré d’évolution. Revue de l’Institut Francais du Petrole 32: 23–42. Fleury, M. and Romero-Sarmiento, M. (2016). Characterization of shales using T 1 –T 2 NMR maps. Journal of Petroleum Science and Engineering 137: 55–62. Freed, D.E. (2007). Dependence on chain length of NMR relaxation times in mixtures of alkanes. Journal of Chemical Physics 126 (17): 174502. Galeev, R.I., Sakharov, B.V., Khasanova, N.M. et al. (2022). Novel low-field NMR method for characterization content and SARA composition of bitumen in rocks. Journal of Petroleum Science and Engineering 214: 110486. Ganeeva, Y.M., Yusupova, T.N., and Okhotnikova, E.S. (2020). Thermal analysis methods to study the reservoir bitumens. Journal of Thermal Analysis and Calorimetry 139: 273–278. Gao, Y., Zou, Y.-R., Liang, T., and Peng, P. (2017). Jump in the structure of type I kerogen revealed from pyrolysis and 13 C DP MAS NMR. Organic Geochemistry 112: 105–118. Gizatullin, B., Gafurov, M., Rodionov, A. et al. (2018). Proton-radical interaction in crude oil – a combined NMR and EPR study. Energy and Fuels 32: 11261–11268. Gizatullin, B., Gafurov, M., Vakhin, A. et al. (2019). Native vanadyl complexes in crude oil as polarizing agents for in situ proton dynamic nuclear polarization. Energy and Fuels 33: 10923–10932. Gupta, P.L., Dogra, P.V., Kuchhal, R.K., and Kumar, P. (1986). Estimation of average structural parameters of petroleum crudes and coal-derived liquids by 13 C and 1 H NMR. Fuel 65: 515–519. Hasan, M.U., Ali, M.F., and Bukhari, A. (1983). Structural characterization of Saudi Arabian heavy crude oil by NMR spectroscopy. Fuel 62: 518–523. Holzgrabe, U. (2017). Quantitative NMR, methods and applications. In: Encyclopedia of Spectroscopy and Spectrometry (ed. J.C. Lindon, G.E. Tranter, and D.W. Koppenaal), 816–823. Oxford: Academic Press. Hudebine, D. and Verstraete, J.J. (2004). Molecular reconstruction of LCO gasoils from overall petroleum analyses. Chemical Engineering Science 59: 4755–4763. Kalabin, G.A., Kanitskaya, L.V., and Kushnarev, D.F. (2000). Quantitative NMR Spectroscopy of Natural Organic Raw Materials and Its Processing Products. Moscow: Chemistry. ISBN: 5-7245-1169-X. Kenyon, W.E. (1992). Nuclear magnetic resonance as a petrophysical measurement. Nuclear Geophysics 6 (2): 153–171.

References

Kenyon, W.E. (1997). Petrophysical principles of applications of NMR logging. The Log Analyst 38 (2): 21–40. Khatibi, S., Ostadhassan, M., Xie, Z.H. et al. (2019). NMR relaxometry a new approach to detect geochemical properties of bitumen in tight shales. Fuel 235: 167–177. Kok, M.V., Varfolomeev, M.A., and Nurgaliev, D.K. (2017). Crude oil characterization using TGA-DTA, TGA-FTIR and TGA-MS techniques. Journal of Petroleum Science and Engineering 154: 537–542. Kushnarev, D.F., Afonina, T.V., Kalabin, G.A. et al. (1989). Investigation of the composition of crude oils and condensates from the south of the Siberian platform using 1 H and 13 C NMR spectroscopy. Petroleum Chemistry U.S.S.R. 29: 149–159. Li, J., Jiang, C., Wang, M. et al. (2020). Adsorbed and free hydrocarbons in unconventional shale reservoir: a new insight from NMR T 1 –T 2 maps. Marine and Petroleum Geology 116: 104311. Ma, Y., Wang, H., Wang, W. et al. (2020). The application of nuclear magnetic resonance T 1 –T 2 maps in the research of sedimentary bitumen: a case study of early mature shale with type I kerogen. Journal of Petroleum Science and Engineering 194: 107447. Marshall, C.P., Kannangara, G.S.K., Wilson, M.A. et al. (2002). Potential of thermogravimetric analysis coupled with mass spectrometry for the evaluation of kerogen in source rocks. Chemical Geology 184: 185–194. Masili, A., Puligheddu, S., Sassu, L. et al. (2012). Prediction of physical–chemical properties of crude oils by 1 H NMR analysis of neat samples and chemometrics. Magnetic Resonance in Chemistry 50: 729–738. Mehana, M. and El-Monier, I. (2016). Shale characteristics impact on nuclear magnetic resonance (NMR) fluid typing methods and correlations. Petroleum 2: 138–147. Mirotchnik, K.D., Allsopp, K., Kantzas, A. et al. (2001). Low-field NMR method for bitumen sands characterization: a new approach. SPE Reservoir Evaluation and Engineering 4 (2): 88–96. Mitchell, J., Gladden, L.F., Chandrasekera, T.C., and Fordham, E.J. (2014). Low-field permanent magnets for industrial process and quality control. Progress in Nuclear Magnetic Resonance Spectroscopy 76: 1–60. Molina, D.V., Uribe, U.N., and Murgich, J. (2010). Correlations between SARA fractions and physicochemical properties with 1 H NMR spectra of vacuum residues from Colombian crude oils. Fuel 89: 185–192. Mondal, S., Yadav, A., Kumar, R. et al. (2017). Molecular-level structural insight into clarified oil by nuclear magnetic resonance (NMR) spectroscopy: estimation of hydrocarbon types and average structural parameters. Energy and Fuels 31: 7682–7692. Mueller, E. and Philp, R.P. (1998). Extraction of high molecular weight hydrocarbons from source rocks: an example from the Green River Formation, Uinta Basin, Utah. Organic Geochemistry 28 (9, 10): 625–631. Peksoz, A., Almaz, E., and Yalciner, A. (2010). The characterization of asphaltene behavior in some aromatic solvents by dynamic nuclear polarization technique. Journal of Petroleum Science and Engineering 75: 58–65. Poindexter, E. (1958). An overhauser effect in natural crude oil. Nature 182: 1087. Poveda, J.-C., Molina, D.-R., and Pantoja-Agreda, E.-F. (2014). 1 H- and 13 C-NMR structural characterization of asphaltenes from vacuum residua modified by thermal cracking. Ciencia, Tecnología i Futuro CT&F 5: 49–60. Rakhmatullin, I.Z., Efimov, S.V., Margulis, B.Y., and Klochkov, V.V. (2017). Qualitative and quantitative analysis of oil samples extracted from some Bashkortostan and Tatarstan oilfields based on NMR spectroscopy data. Journal of Petroleum Science and Engineering 156: 12–18.

77

78

2.2 NMR-Spectroscopy and NMR-Relaxometry

Rakhmatullin, I.Z., Efimov, S.V., Tyurin, V.A. et al. (2020). Qualitative and quantitative analysis of heavy crude oil samples and their SARA fractions with 13 C nuclear magnetic resonance. Processes 8: 995. Romero-Sarmiento, M.F., Ramiro-Ramirez, S., Berthe, G. et al. (2017). Geochemical and petrophysical source rock characterization of the Vaca Muerta Formation, Argentina: implications for unconventional petroleum resource estimations. International Journal of Coal Geology 184: 27–41. Rudszuck, T., Förster, E., Nirschl, H., and Guthausen, G. (2019). Low-field NMR for quality control on oils. Magnetic Resonance in Chemistry 57 (10): 777–793. Rudszuck, T., Nirschl, H., and Guthausen, G. (2021). Perspectives in process analytics using low field NMR. Journal of Magnetic Resonance 323: 106897. Sanchez-Minero, F., Ancheyta, J., Silva-Oliver, G., and Flores-Valle, S. (2013). Predicting SARA composition of crude oil by means of NMR. Fuel 110: 318–321. Smirnov, M.B. and Vanyukova, N.A. (2014). Relations between the main structural-group composition parameters of Western Siberia crude oils according to NMR data. Petroleum Chemistry 54: 355–365. Song, Y.Q. and Kausik, R. (2019). NMR application in unconventional shale reservoirs – a new porous media research frontier. Progress in Nuclear Magnetic Resonance Spectroscopy 112, 113: 17–33. Straley, C., Rossini, D., Vinegar, H. et al. (1997). Core analysis by low-field NMR. The Log Analyst 38 (2): 84–93. Tikhonova, M.S., Ivanova, D.A., Kalmykov, A.G. et al. (2019). Methods of step extraction of rocks of high-carbon formations for the study of the component distribution of bitumen and variability of their basic geochemical parameters. Georesources 21: 172–182. Timur, A. (1969). Pulsed nuclear magnetic resonance studies of porosity, movable fluid, and permeability of sandstones. Journal of Petroleum Technology 21 (6): 775–786. Verstraete, J.J., Schnongs, P., Dulot, H., and Hudebine, D. (2010). Molecular reconstruction of heavy petroleum residue fractions. Chemical Engineering Science 65: 304–312. Volkov, V.Y., Sakharov, B.V., Khasanova, N.M., and Nurgaliev, D.K. (2018). Analysis of the composition and properties of heavy oils in situ by low field NMR relaxation method. Georesursy 20 (4): 308–323. Volkov, V.Y., Al-Muntaser, A.A., Varfolomeev, M.A. et al. (2021). Low-field NMR-relaxometry as fast and simple technique for in-situ determination of SARA-composition of crude oils. Journal of Petroleum Science and Engineering 196: 107990. Yang, K., Connolly, P.R.J., Li, M. et al. (2020). Shale rock core analysis using NMR: effect of bitumen and water content. Journal of Petroleum Science and Engineering 195: 107847. Yao, Y., Liu, D., Che, Y. et al. (2010). Petrophysical characterization of coals by low-field nuclear magnetic resonance (NMR). Fuel 89 (7): 1371–1380. https://doi.org/10.1016/j.fuel.2009.11.005. Zauer, E. (2018). Modern automatic CHNS/O/X organic compound analyzers. Analytics and Control 22: 6–19. ISSN: 2073-1442 (Print), ISSN: 2073-1450 (Online).

79

2.3 FTIR-Spectroscopy Irek I. Mukhamatdinov Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia

Among a diversity of studying methods of the structure of organic and inorganic compounds, including oils, the important place is taken by the infrared spectroscopy (IR) method. This method is almost universal in its opportunities. Researches of oils and oil products are conducted in the near, middle, and far region of IR spectrum (Ivanova et al. 2008). For the characteristic of oils by IR method, the following absorption bands are often used: 720–730 cm−1 band corresponds to deformation pendulum oscillations of the methylene group; 1200–1300 cm−1 band accounts for fan and rotating oscillations of CH2 -group; Bands in the field of 1375–1460 cm−1 correspond to the symmetric and antisymmetric deformation oscillations of methyl and methylene groups. Oscillations of aromatic fragments are shown on different sections of a spectrum, where corresponding bands with frequencies are 1600, 750, 810, and 870 cm−1 (Bellamy 1975; Petrova et al. 2011; Grin’ko and Golovko 2014; Svarovskaya et al. 2009). Because of the complexity of composition of the oils, IR spectra of oil and bitumen systems represent a difficult pattern of overlapping and superimposing absorption bands with distortion of their form and intensity. Therefore, direct identification and moreover the quantitative determination of the functional groups on absorption intensities in IR spectra are impossible. Thus, indirect techniques of identification and determination of structural fragments content of hydrocarbon and nonhydrocarbon compounds of oils and oil products are used. One of the widely used approach for the characterization of oils and bitumens by IR method is the use of so-called “spectral coefficients,” (SCs) that is the relations of optical density of characteristic absorption bands of different types of bonds: aromaticity coefficient of Car = D1600 /D720 , aliphaticity Cal = D720 +D1380 /D1600 , branching Cb = D1380 /D720 (Svarovskaya et al. 2009; Abdrafikova et al. 2013; Agaev et al. 2006; Kayukova et al. 2017; Yusupova et al. 2017). IR spectroscopy is considered as one of the crucial methods among various methods used to study the structure of oil and oil products. It is mainly based on the absorption, reflection, and scattering of infrared energy when passing through a substance. This method is also more reliable to determine functional groups in their qualitative and quantitative analyses than chemical analyses. The structure-group composition of oil, resins, and asphaltenes samples were studied by FTIR spectroscopy. IR spectra of resins and asphaltenes were recorded on a PERKIN ELMER Spectrum two with a UATR (Single Reflection Diamond) attachment in the range from 2000 to 450 cm−1 with a resolution of 4 cm−1 . In order to study the structure-group composition of the investigated products, the spectral coefficients were calculated, defined as the ratio of the optical density values at the maxima of the corresponding absorption bands: C1 = D1600 /D720 (aromaticity); C2 = D1710 /D1465 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (aliphaticity); C5 = D1030 /D1465

80

2.3 FTIR-Spectroscopy

(sulfurization index) using a baseline in the 2000–450 cm−1 spectrum. The first coefficient is aromaticity due to stretching vibrations of C=Carom bonds of aromatic rings. Oxidation of (C2) is associated with the presence of carbonyl groups (CO); the structure of paraffin chains can be estimated by their branching (C3) CH3 /CH2 . The fourth factor is aliphaticity (CH3 + CH2 )/C = Carom, , and it shows the proportion of methyl and methylene groups to aromatic, whereas C5 shows the proportion of sulfoxide groups. The condensing degree of asphaltenes (C6) is a ratio of optical density of band adsorption at 880 cm−1 , which stands for angular condensation of aromatic rings to the sum of optical adsorption bands at 750 and 820 cm−1 (D880 /D750 + D820 ) standing for aromatic C—H bonds in linear and perpendicular to the molecule. In the work of Vakhin et al. (2017), a synthetic oil-soluble iron-based catalyst was studied experimentally. A physical model of the catalytic transformation of high-viscosity oil at 200 ∘ C was developed. The composition and physicochemical and rheological characteristics of the thermocatalysis products were studied. IR spectroscopy found that the compositions of individual fractions changed. It was shown that the fraction of high-molecular-mass components could be substantially reduced by using the synthetic catalyst in combination with a hydrogen donor. This reduced the viscosity and therefore, increased the degree of oil extraction. IR spectroscopy was used to analyze the structure and group composition of resins and asphaltenes from starting Ashal’cha oil and noncatalytic and catalytic aquathermolysis products after thermal treatment (200 ∘ C, six hours). The spectral coefficients (Table 2.3.1) of resins from Ashal’cha field oils were calculated using the IR spectra shown in Figure 2.3.1. Hydrothermal transformations of oil increase the aromaticity and decrease the aliphaticity. IR spectroscopy showed that resins in oil with a hydrogen donor and catalyst had higher aromaticity coefficients than those in oils without catalysts. This could be seen from the increasing intensities of absorption bands at 1600 and 900–730 cm–1 . Also, the aliphaticity decreased as compared with starting oil and reached a maximum of 3.54 units in sample 2 because of the reduced aromaticity. It should be noted that the resin oxidation increased characteristically with a catalyst and hydrogen donor because carbon-heteroatom bonds were hydrolyzed to form mostly phenols and alcohols. The noticeable increase of the sulfation parameter, which characterized the –SO content, was considered just as important and resulted from oxidation of sulfides to sulfoxides. The decrease of the resin branching parameter with a catalyst and hydrogen donor in the oil was a consequence of the increased length of the alkyl chains as compared with the experiment without the additives. Thus, the IR spectral studies indicated that the structures of resins and asphaltenes changed significantly in the presence of a catalyst and hydrogen donor (Vakhin et al. 2017). The next work (Sitnov et al. 2017) is devoted to the study of the influences of catalytic (introduction of catalytic precursors and [H]-donor with amount of 1 wt.% on oil) and noncatalytic Table 2.3.1

Spectral coefficients of from Ashal’cha oil resins. Products of aquathermolysis

Resins spectral coefficients

1

2

3

4

C1

1.86

1.58

0.58

1.93

C2

0.17

0.56

0.31

0.30

C3

0.57

0.76

0.65

0.64

C4

3.06

2.43

3.54

2.56

C5

0.21

0.54

0.38

0.35

666.40

753.56

809.01

867.57

1020.94

1215.04

1304.69

1376.83

1455.96

1602.21

3.0 3.5 4.0 2.5 0.5 1.0 1.5 2.0

4 3 2 1

0.0

Absorbance units

1697.45

2.3 FTIR-Spectroscopy

2000

1800

1600

1400 1200 Wavenumber (cm–1)

1000

800

600

Figure 2.3.1 IR spectra of resins from studied oils: 1 – initial oil, 2 – control experiment, 3 – with catalyst-based Fe, 4 – with catalyst-based Fe and hydrogen donor. Table 2.3.2

Number of experience in accordance with the experimental conditions.

Experiment number

Condition of experiment

Initial oil 1 2

Product of the noncatalytic aquathermolysis (Temperature = 150 ∘ C, initial pressure = 3 bar, water:oil 30 : 70) Product of the catalytic aquathermolysis (Temperature = 150 ∘ C, initial pressure = 3 bar, water:oil 30 : 70)

3 4

Product of the noncatalytic aquathermolysis (Temperature = 180 ∘ C, initial pressure = 3 bar, water:oil 30 : 70) Product of the catalytic aquathermolysis (Temperature = 180 ∘ C, initial pressure = 3 bar, water:oil 30 : 70)

aquathermolysis to the group composition and properties of extra-heavy oil from Ashal’cha oil reservoir (Republic of Tatarstan) in reservoir conditions (Table 2.3.2). As revealed from the results of IR spectroscopy of the resin fraction, an increase in the aromaticity factor (C1) occurs as a result of the temperature increase and the introduction of the catalyst and the hydrogen donor. The intensity of the absorption bands increases at 1600 cm−1 , with a corresponding decrease in aliphaticity (C4). The maximum effect is observed at a temperature of 180 ∘ C (Exp. 4). Under these conditions, the oxidation index of the resins also decreases, and there is no absorption band at 1710 cm−1 . There is a marked increase in the sulfuring degree characterizing the content of the –SO groups. This is a consequence of the oxidation of sulfides to sulfoxides. When using a catalyst and a hydrogen donor at 150 ∘ C, the branching degree of the resins is reduced due to the increase in the length of the alkyl chains (Figure 2.3.2, Table 2.3.3). In Table 2.3.4 and in Figure 2.3.3 are presented the results of IR spectral analysis of asphaltenes deposited from oil samples after the process of aquathermolysis. According to Table 2.3.4, the characteristic increase in absorption bands of 1710 cm−1 corresponding to the C=O vibrations of acid groups occurs in asphaltene fractions in catalytic

81

Absorbance units 3 4

Exp. 1

2

498.89 471.61 463.72

602.87

806.37 744.28 728.35 694.79

862.93

1019.83

1154.32

1305.99 1261.13

1376.30

1455.60

1603.73

5

1668.13

2.3 FTIR-Spectroscopy

Exp. 2

1

Exp. 3 Exp. 4 0

82

2000

Figure 2.3.2 Table 2.3.3

1800

1600

1400 1200 Wavenumber (cm–1)

1000

800

600

IR spectra of resins after thermal treatment. Spectral coefficients of resins after thermal treatment.

Spectral coefficients

Experiment 1

Experiment 2

Experiment 3

Experiment 4

C1

0.59

0.74

0.76

1.01

C2

0.22

0.17

0.17

—

C3

0.66

0.64

0.68

0.70

C4

4.71

4.91

4.49

3.64

C5

0.39

0.76

0.58

0.60

Table 2.3.4

Spectral coefficients of asphaltenes after thermal treatment.

Spectral coefficients

Experiment 1

Experiment 2

Experiment 3

Experiment 4

C1

0.95

0.91

1.07

0.79

C2

0.10

0.17

0.12

0.20

C3

0.73

0.76

0.74

0.78

C4

3.34

2.84

2.82

2.98

C5

0.38

1.16

0.42

0.43

C6

0.42

0.26

0.41

0.44

aquathermolysis (Exp. 2, Exp. 4). This is the result of the hydrolysis of carbon-heteroatom bonds with the formation of phenols and alcohols. The aromaticity index (C1) of asphaltenes decreases insignificantly at temperatures of 150 and 180 ∘ C due to the C=Car bond destruction and decomposition of aromatic rings. In the case of catalytic aquathermolysis at 150 ∘ C (Exp. 2), an increase in the sulphidity index (C5) is noticeable,

Exp.3

861.24

1028.82

1260.22

1337.47 1312.34

1375.11

1455.20

1600.11 1574.52 1557.11

745.13 724.07 666.31 624.69 603.14 579.44 553.90 524.81 496.43 478.41 465.56 459.24 456.43

2

Exp.2

806.89

Absorbance units 3 4 5

Exp.1

1

1696.44

2.3 FTIR-Spectroscopy

800

600

0

Exp.4 2000 Figure 2.3.3

1800

1600

1400 1200 Wavenumber (cm–1)

1000

IR spectra of asphaltenes after thermal treatment.

which is the reason for the oxidation of sulfides to sulfoxides. Also, the condensation index (C6) of asphaltenes is reduced by a factor of 2 due to decomposition and hydrogenation of aromatic rings and partially destruction of C—C bonds in asphaltene molecules in the presence of a hydrogen donor and a catalyst. Thus, the results of the studies have shown the possibility of using a Co-based catalyst for steam-thermal treatment of the reservoir to intensify the processes of in situ refining of high-viscosity oil. The main effect of using the catalyst, which manifests itself already at 150 ∘ C, is to reduce the content of resins and asphaltenes, which ensure an irreversible decrease in the viscosity of the produced oil, facilitates further transportation and processing (Sitnov et al. 2017). Onishchenko et al. (2017) presents an analysis of the change in the composition of the bitumoid of Domanic sediments of the Volga–Ural oil and gas province with thermal effects on the specimen of the kerogen-containing formation. The laboratory modeling of the katagenesis process on Domanic rock sample was conducted in autoclave at a pressure of 50 atm, at a temperature range of 200–350 ∘ C with the addition of water, and the duration of exposure of six hours. They used a sample of Domanic rock with a high kerogen content from Sarmanovskoya area (Tatarstan Republic) of Franco-Famenian carbonate complex. The IR spectral analysis was also conducted for initial resins and after the thermal treatment. The results are presented in Table 2.3.5 and Figure 2.3.4. Table 2.3.5

IR-analysis of the initial resins and after the thermal treatment with water.

Spectral coefficients

Initial

C1

0.64

C2 C3 C4 C5

200 ∘ C

250 ∘ C

300 ∘ C

350 ∘ C

0.29

0.28

1.20

0.97

0.28

0.37

0.43

0.39

0.48

0.62

0.63

0.65

0.70

0.71

4.04

5.34

5.43

2.63

2.39

1.69

14.52

6.78

0.82

0.86

83

675.47

729.74

861.99 799.75

1015.59

1091.78

1259.87

1376.80

1478.77 1455.62

4 3

1 2

Absorbance units

5

1651.17

2.3 FTIR-Spectroscopy

2

1

3 4 5 0

84

2000

1800

1600

1400

1200

1000

800

600

Wavenumber (cm–1)

Figure 2.3.4 IR spectra of initial resins (1) and after thermal treatment with water: 2 – 200 ∘ C; 3 – 250 ∘ C; 4 – 300 ∘ C; 5 – 350 ∘ C.

The IR-spectra of the resins are characterized by an increase in the intensity of the aromatic triplet in the range of 900–730 cm−1 , which indicates the occurrence of processes with the formation of condensed aromatic structures. In the IR-spectrum of resins, bands in the 1200–1100 cm−1 region corresponding to alcohols and ethers appear which also confirms the course of oxidation processes under the given experimental conditions. An increase in the oxidation index of resins is observed after water treatment at 350 ∘ C. It should be emphasized that the extreme change in the index of sullenness with a maximum value of 14.52 units. At a temperature of 200 ∘ C, characterizing the content of the –S=O groups, with a further decrease in the coefficient at a temperature of 300 ∘ C, which is the reason for the reduction of sulfoxides to sulfides and hydrogen sulfide. The IR spectral analysis was also conducted for initial asphaltenes and after the thermal treatment. The results are presented in Table 2.3.6 and Figure 2.3.5. A significant difference in the spectra of asphaltenes is an increase in the intensity of the absorption bands of 1710 cm−1 , corresponding to the vibrations of the C=O carbonyl groups. This is the result of the hydrolysis of carbon-heteroatom bonds with the formation of, in particular, phenols and alcohols, reaching a maximum at 350 ∘ C characterizing the content of the –S=O groups, with Table 2.3.6 IR-analysis of the initial asphaltenes and after the thermal treatment with water. Spectral coefficients

Initial

200 ∘ C

250 ∘ C

300 ∘ C

350 ∘ C

C1

1.75

0.72

0.97

1.24

1.26

C2

0.24

0.64

0.57

0.40

0.65

C3

0.75

0.80

0.79

0.80

0.89

C4

2.10

2.77

2.25

2.02

1.89

C5

0.42

0.66

0.58

0.47

0.77

C6

0.34

0.42

0.37

0.33

0.28

1030.14

805.88

746.72 728.34 677.00 622.96 590.48 542.89 517.98 493.02 471.19 464.94

1261.60

1375.88

1454.84

1000

800

600

4 3

1 2

Absorbance units

5

1643.91 1605.70

2.3 FTIR-Spectroscopy

2

1

3 4

0

5 2000

1800

1600

1400

1200

Wavenumber (cm–1)

Figure 2.3.5 IR spectra of initial asphaltenes (1) and after thermal treatment with water: 2–200 ∘ C; 3–250 ∘ C; 4–300 ∘ C; 5–350 ∘ C.

a further decrease in the coefficient at a temperature of 300 ∘ C, which causes the reduction of sulfoxides to sulfides and hydrogen sulfide. Aromaticity reduced asphaltenes at 200 ∘ C due to the reactions of degradation of C=Sar and decompressor aromatic rings. In the same way, the aliphaticity index of the sample is increased after the steam-thermal effect at 200 ∘ C. The branching index asphaltenes varies slightly compared to the original asphaltene bitumoid and reaches a maximum for the sample after steam-heat exposure at 350 ∘ C due to the decrease of straight-chain methylene groups (CH2 ) in paraffins. Fixed increase in the sulfurized representing the content of sulfoxide groups (–S=O), which is the cause of oxidation of sulfides to sulfoxides with increasing steam-temperature exposure. Also, the condensation index of asphaltenes decreases slightly at 350 ∘ C (Onishchenko et al. 2017). In the experiments of Salih et al. (2018a), asphaltene fractions extracted from samples of high-viscosity Ashal’cha (Tatarstan), Boca de Jaruko (Cuba), and Tahe (China) deposits and a sample of oxidized road asphalt with trademark BND 50/70 produced by OJSC TAIF-NK (Tatarstan) were studied. Five asphaltene fractions of these crude oils and oil products were obtained by step-by-step extraction with solvents of various compositions and dissolving powers. The fractional compositions of asphaltenes from various types of oil and of a product that had been subjected to refinery processing were found to vary. Spectral coefficients were determined from the IR spectroscopic data for individual fractions characterizing the structural-group composition of the subjects of investigation. The results of the IR spectral analysis of the asphaltene fractions of the investigated oils and asphalts are presented in Table 2.3.7 and Figure 2.3.6. A significant difference in the spectra of the asphaltenes from the oil of the Ashal’cha deposit is the absence of the absorption bands at 1710 cm−1 corresponding to the C=O vibrations of the carbonyl groups in the A1, A2, and A5 fractions, indicating a small content of aromatic and oxygen-containing fragments. The absorption band at 1710 cm1 is only present in the A3 and A4 fractions, and it has an identical spectral coefficient of 0.22. The aromaticity of the asphaltenes decreases extremally with a minimum in the A4 fraction (0.67) and a marked increase to 1.15 in A5, which probably indicates closer packing of the aromatic rings. Analysis of the aliphaticity and branching coefficients indicates ambiguity in the variation of these parameters, but they correspond to maximum aromaticity index for the A5 fraction with minimum aliphaticity index

85

86

2.3 FTIR-Spectroscopy

Table 2.3.7 Spectral coefficients

Results of IR spectral analysis of asphaltene fractions.

A1

A2

A3

A4

A5

Carbene carboids

1.15

1.09

Oil of Ashal’cha deposit C1 C2

1.01 —

0.84 —

0.73

0.67

0.22

0.22

—

—

C3

0.75

0.79

0.62

0.71

0.80

0.82

C4

2.98

3.25

2.98

3.86

2.78

2.82

C5

0.47

0.43

1.15

0.40

0.43

0.45

C6

0.39

0.30

—

0.46

0.40

0.39

Oil of Boca de Jaruko deposit C1

0.05

0.02

0.66

0.48

0.65

1.06

C2

0.36

0.22

0.18

0.26

0.26

0.12

C3

0.64

0.62

0.70

0.72

0.76

0.65

C4

23.36

47.85

3.96

4.28

4.39

3.18

0.33

0.42

0.47

0.55

0.30

0.28

0.08

0.42

0.31

0.38

1.75

C5 C6

0.59 —

—

Oil of Tahe deposit C1

1.39

1.03

1.45

1.38

C2

0.09

0.21

0.24

0.25

0.05

0.22

C3

0.80

0.82

0.84

0.82

0.62

0.89

C4

2.22

2.49

2.16

2.22

7.43

2.05

C5

0.38

0.44

0.42

0.40

0.12

0.60

C6

0.38

0.35

0.41

0.45

0.16

0.51

OJSC TAIF-NK BND 50/70 road bitumen C1

1.37

1.27

1.26

1.40

1.58

1.46

C2

0.22

0.24

0.31

0.29

0.36

0.40

C3

0.76

0.79

0.79

0.79

0.78

0.80

C4

2.22

2.41

2.36

2.27

2.10

2.03

C5

0.36

0.43

0.38

0.44

0.36

0.41

C6

0.36

0.34

0.26

0.35

0.39

0.30

compared with the other fractions. Of no lesser importance is the appreciable increase of the sulfurization index characterizing the content of sulfoxide groups (–S=0), the highest value of which is observed in the A3 fraction. The condensation index of the asphaltenes is also ambiguous and has a maximum at the A4 fraction. In contrast to the asphaltenes from oil of the Ashal’cha deposit, the spectra of the asphaltenes from the Boca de Jaruko oil do contain absorption bands at 1710 cm−1 , indicating a significant content of aromatic- and oxygen-containing fragments. The oxidizability decreases in the series of asphaltene fractions, reaching a minimum in the carbene–carboid fraction. The aromaticity of the asphaltenes increases significantly from the Al fraction (0.05) to the A3 fraction (0.66) and reaches

1 2

1 2 3 4 5 6

3 4 5 6 (a)

(c)

1 2 3

1 2 3 4 5 6 2000 1800

4 5 6 1600 1400 1200 1000 Wavenumber (cm–1) (b)

800

600

2000 1800

1600 1400 1200 1000 Wavenumber (cm–1) (d)

800

600

Figure 2.3.6 IR spectra of asphaltene fractions of oil from the Ashal’cha (a), Boca de Jaruko (b), and Tahe (c) deposits and BND 50/70 road bitumen (d): (1) A1; (2) A2; (3) A3; (4) A4; (5) A5; (6) carbene–carboids.

88

2.3 FTIR-Spectroscopy

the highest value in the carbene–carboid fraction (1.06), indicating strongest condensation of the naphthene and aromatic rings. A characteristic inverse relationship is observed for the aliphaticity index, which decreases from 23.36 to 3.18 in the transition to the heavier nonpolar fraction. Analysis of the sulfurization, branching, and condensation coefficients indicates ambiguity in the variation of these characteristics, the highest values of which are observed in the A5 fraction. As shown by the results of IR spectral analysis of the asphaltenes of BND 50/70 grade asphalt produced by OJSC TAIF-NK, the aromaticity coefficient decreases from Al (1.37) to A3 (1.26) and then increases to A5 (1.58). The aliphatic index is inversely proportional to the changes in aromaticity index. The oxidization, branching, and condensation indices of the asphaltenes change little. In the studies of Salih et al. (2018b), the resins extracted from heavy oil sample from Ashal’cha oilfield (Tatarstan, Russia) and the products of Co-based catalytic and noncatalytic aquathermolysis are investigated. Saturates and four subfractions of resins (R1–R4) are extracted by liquid adsorption chromatography method with individual and solvents. Fractionation method leads to obtain more detail information about the structure of high molecular weight compounds due to subdividing large molecule and its aggregates into small ones. The extracted subfractions of resins are different in color (R1 – yellow, R2 – orange, R3, R4 – dark brown). The pattern of adsorption chromatographic separation of resin components allows us to substantially supplement the results of IR spectral analysis. The pattern of adsorption–chromatographic separation of resin components allows us to interoperate the IR spectra results more considerably. Table 2.3.8 and Figure 2.3.7 present the results of IR spectra of resin fractions of initial oil. IR spectra indicate an ambiguous character of change in the spectral coefficients of resin fractions. The aromaticity index, like that of aliphatic, has maxima and minima, which confirms the fundamentally different structure of the resin fractions as elution with solvents, possessing different dissolving power. The heavier fraction of R4 is characterized by the minimum aliphaticity value (3.62 units) and the maximum aromatic value (0.71 units). The coefficient of sulfidity increases, probably because of the increasing amount of sulfoxides in the resin fractions. Figure 2.3.8 and Table 2.3.9 illustrate the results of IR spectral analysis of resin fractions of a catalytic aquathermolysis product with the use of a hydrogen donor. The absence of an aromaticity factor in R1 and the minimum value for R2 and R3 confirms the hydrothermal-catalytic effect with the course of reactions of destruction of C—S and C—O bonds of aromatic rings and side hydrocarbon chains. The aliphaticity index C1 of the R2 fraction increases fivefold and for the fraction of R3 – increases by a factor of 2.5 compared to the corresponding index value of the original oil. The highest value of aromaticity is again characterized by a heavier fraction of R4, in view of the transition in the composition of resins of the most high-molecular Table 2.3.8 Spectral coefficients obtained from the IR spectra of resin fractions of the initial oil. Spectral coefficients

R1

R2

R3

R4

C1

0.34

0.42

0.06

0.71

C2

0.03

0.07

—

—

C3

0.59

0.63

0.62

0.69

C4

8.67

6.66

19.07

3.62

C5

0.17

0.27

0.32

0.37

858.35

806.62 746.54 729.78 719.55 666.27 633.98 603.74 588.34 578.34 567.64 557.79 527.02 498.81 482.43 474.24 463.32

1027.93

1111.94

1310.04

1375.68

1461.32 1455.09

4 3

R1

2

R3

1

R2

R4

0

Absorbance units

5

1599.26

2.3 FTIR-Spectroscopy

1600

1200 1400 Wavenumber (cm–1)

1000

800

600

667.48

753.55

809.74

1031.43

4 3

R1

2

R3

1

R2

R4

0

Absorbance units

5

1215.08

IR spectra of resin fractions of the original oil.

1456.01

Figure 2.3.7

1800

1376.99

2000

2000

1800

1600

1400

1200

Wavenumber Figure 2.3.8

1000

800

600

(cm–1)

IR spectra of resin fractions of catalytic aquathermolysis with a hydrogen donor.

compounds from fractions with a lower molecular weight of R1–R2 to R4 during steam-thermal treatment together with the catalyst. The work of Vakhin et al. (2018a) is devoted to the study of aquathermolysis processes of heavy oil produced by CSS technology on Boca de Jaruco oil field. Various catalysts based on cobalt, nickel, iron, and copper were used for intensification of in situ upgrading processes of heavy oil. The first paper in a series presented the results of transformation of crude oil and its saturate fraction after thermal treatment with and without catalysts. Spectral coefficients (SC) of initial crude oil, as well as thermo catalytic treated one are presented in Table 2.3.10 and their spectra are in Figure 2.3.9a.

89

90

2.3 FTIR-Spectroscopy

Table 2.3.9 Spectral coefficients obtained from the IR spectra of resin fractions gained using catalytic aquathermolysis in the presence of hydrogen donor. Spectral coefficients

R1

R2

R3

R4

C1

—

0.04

0.02

0.53

C2

—

0.13

—

0.22

C3

0.56

0.63

0.59

0.65

C4

—

30.45

47.65

3.91

C5

0.15

0.28

0.30

0.34

Table 2.3.10 Characterization of initial crude oil and treated crude oil after catalytic aquathermolysis according to IR spectroscopy data. Spectral indicesa) Objects and experiment conditions

The initial crude oil

C1

C2

C3

C4

C5

1.50

0.14

0.56

4.17

0.04

0.93

0.23

0.56

3.65

0.03

1.05

0.22

0.56

3.48

0.03

1.00

0.28

0.58

3.21

0.03

Samples of crude oil after treatment After thermal treatment After thermal treatment with hydrogen donor Co After thermal treatment with catalyst

Ni

1.50

0.31

0.73

2.84

0.07

Fe

1.13

0.22

0.68

3.62

0.10

Cu

1.70

0.25

0.56

2.93

0.04

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D146 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization).

According to IR-spectral analysis, the content of aromatic structures after noncatalytic thermal treatment (SC = 0.93) decreases comparing with initial crude oil (SC = 1.50). In contrast, addition of hydrogen donor slightly increases the aromatic content (SC = 1.05). However, addition of various catalysts further increases this coefficient (Table 2.3.10). The maximum value of this parameter corresponds to the experiment with catalyst based on copper. This is probably due to destruction of aliphatic substituents and condensation of aromatic rings. The increase in oxidation and sulfurization parameters at 250 ∘ C is observed, which reaches maximum in the case of nickel catalyst. These are indications of thermooxidative reactions. Another aspect is increasing the branching parameters because of alkyl chains elongation compared with blank run. The IR-spectroscopy analysis of investigated crude oil’s individual components (saturates, aromatics, resin, and asphaltene) revealed the common features in their structures, as well as their differences. The spectrums and their coefficients of saturate hydrocarbons are provided in Figure 2.3.9b and Table 2.3.11, respectively. According to IR-spectroscopy of saturate hydrocarbons, the content of sulfoxide group decreases after thermocatalytic treatment

Absorbance (a.u)

2.3 FTIR-Spectroscopy

6 5 4 3 2 1 0 800

1000 1200 1400 Wavenumber (sm–1) (a)

1600

1800

6

Absorbance (a.u)

600

5 4 3 2 1

600

Figure 2.3.9 fraction (b).

800

1000 1200 1400 Wavenumber (sm–1) (b)

1600

0 1800

IR spectra of oils before and after thermal-catalytic treatment (a) and saturated hydrocarbons

of oil at 250 ∘ C due to reduction reaction of sulfoxides to hydrogen sulfides. The branching parameter changes to some extent. For saturate hydrocarbons, the share of paraffinic structure (aliphatic coefficient) increases according to intensive absorption band of 720 cm−1 standing for methylene (CH2 ) group in paraffin. The maximum value is observed for the products of cobalt-based aquathermolysis. In the second part of work of Vakhin et al. (2018b), the conversion of aromatic hydrocarbons depending on thermocatalytic conditions and composition of catalysts are investigated. The thermocatalytic treatment destructs resin and asphaltene molecules. The destruction products transform into aromatic fractions.

91

92

2.3 FTIR-Spectroscopy

Table 2.3.11 Characterization of saturates fraction of initial crude oil and treated crude oil after catalytic aquathermolysis according to the data of IR spectroscopy. Spectral indicesa) Objects and experiment conditions

The initial crude oil

C3

C4

C5

0.56

4.17

0.04

0.55

—

—

Samples of crude oil after treatment After thermal treatment After thermal treatment with hydrogen donor

After thermal treatment with catalyst

0.55

21.00

0.05

Co

0.54

21.14

0.04

Ni

0.51

16.53

0.06

Fe

0.55

20.47

0.04

Cu

0.51

12.90

0.07

a) C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization).

Table 2.3.12

IR-analysis of the aromatic hydrocarbon fraction. Spectral coefficientsa)

Objects and experiment conditions

0

The initial oil

C1

C3

C4

C5

1.06

0.53

4.78

0.20

Experimental products 1

After thermal treatment

1.05

0.53

4.66

0.22

2

After thermal treatment with hydrogen donor

0.37

0.56

6.68

0.22

3

Co

1.00

0.53

4.93

0.18

4

Ni

0.37

0.55

6.96

0.22

Fe

1.12

0.54

4.51

0.21

Cu

0.35

0.55

7.08

0.20

5 6

After thermal-catalytic treatment by catalyst

a) C1 = D1600 /D720 (aromaticity); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization).

The spectral coefficients of aromatic hydrocarbons and their specters are demonstrated in Table 2.3.12 and Figure 2.3.10, correspondingly. The analysis of aromatic hydrocarbons doesn’t account for oxidation coefficient of C2. According to IR-spectroscopy results of aromatic hydrocarbons, the least value of aromaticity index corresponds to the samples 2 (with hydrogen donor), 4 (Ni-based catalyst), and 6 (Cu-based catalyst). This confirms the destruction reaction of C=C bonds in aromatics, as well as aromatic rings become denser. The reverse characteristic dependency of aliphaticity value is observed for the same samples. This value increases by 1.5 times due to reduction of aromaticity value. Branching and sulfurization parameters do not vary much. The third part of study by Vakhin et al. (2018c) discusses conversion of resins and asphaltenes. The influence of thermocatalytic conditions and composition of catalysts is also studied.

Absorbance (a.u)

2.3 FTIR-Spectroscopy

6 5 4 3 2 1 0

600

800

1000 1200 1400 Wavenumber (sm–1)

1600

1800

Figure 2.3.10 IR spectra of aromatic hydrocarbon fraction of initial oil (0), experimental products after thermal treatment (1), and after thermal treatment with hydrogen donor (2), after thermal-catalytic treatment by various catalyst based on (3) Co, (4) Ni, (5) Fe, (6) Cu.

The destruction of resins and asphaltenes are observed after thermocatalytic treatments. The changes in composition of resins and asphaltenes are revealed by IR-spectroscopy data. The spectral coefficients of resins corresponding to the crude oil of Boca de Jaruco reservoir are presented in Table 2.3.13 and their specters in Figure 2.3.11a. According to IR-spectral analysis of resins, nonlinear changes in the content of aromatic structures occur. This is observed from intensity of absorption band at 1600 cm−1 . The maximum value corresponds to the sample 5 (iron-based catalyst). In this sample, the destruction reactions of aliphatic substitutes take place in polar components of resins, which is justified by the least coefficient of aliphaticity. The lower C1 and higher C4 values are detected in samples 3 (cobalt-based catalyst), 4 (nickel-based catalyst), and 6 (copper-based catalyst). It is necessary to underline the nonlinear change of Table 2.3.13

IR-analysis of the resins. Spectral coefficientsa)

Objects and experiment conditions

C1

C2

C3

C4

C5

0

0.15

0.35

0.64

9.96

0.58

0.55

0.17

0.62

6.13

0.29

1.54

0.48

0.77

2.95

0.67

0.32

0.23

0.76

5.44

0.36

Initial oil

Experimental products 1

After thermal treatment

2

After thermal treatment with hydrogen donor

3 4 5 6

Co After thermalcatalytic treatment with catalyst

Ni

0.47

0.20

0.60

6.10

0.36

Fe

3.93

0.52

1.00

1.96

0.85

Cu

0.40

0.13

0.59

6.58

0.24

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D146 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization).

93

2.3 FTIR-Spectroscopy

IR intensity

6 5 4 3 2 1 0

600

800

1000 1200 1400 Wavenumber (sm–1) (a)

1600

1800

6

IR intensity

94

5 4 3 2 1 0

600

800

1000 1200 1400 –1 Wavenumber (sm ) (b)

1600

1800

Figure 2.3.11 IR spectra of resins (a) and asphaltenes (b), initial oil (0), experimental products after thermal treatment (1), and after thermal treatment with hydrogen donor (2), after thermal-catalytic treatment with various catalyst based on (3) Co, (4) Ni, (5) Fe, (6) Cu.

oxidizing parameter of resins (adsorption band is 1710 cm−1 , which stands for C=O carbonyl groups), maximum of which also corresponds to 5 as a result of content increase in composition of aromatic and oxygen-containing fragments. In resin specters, bands are detected in the range of 1200–1100 cm−1 , corresponding to alcohols and esters, which also justifies oxidizing processes. The branching parameter negligibly changes the maximum of which coefficient = 1 is observed in sample 5 as well. The critical change of sulfurization parameter with a minimum value of 0.24 unit for the sample 6 corresponds to the content of –S=O group. However, sample 5 has a maximum value, which is explained by oxidation of sulfides into sulfoxides.

2.3 FTIR-Spectroscopy

Table 2.3.14

IR-analysis of the asphaltenes. Spectral coefficientsa)

0

Objects and experiment conditions

C1

C2

C3

C4

C5

C6

The initial oil

0.15

0.35

0.64

9.96

0.58

0.53

Experimental products 1

After thermal treatment

0.85

0.16

0.67

3.87

0.28

0.31

2

After thermal treatment with hydrogen donor

0.39

0.18

0.67

5.26

0.30

0.08

3 4 5 6

After thermal-catalytic treatment with catalyst

Co

0.40

0.15

0.64

5.09

0.24

0.05

Ni

0.74

0.15

0.66

4.10

0.24

0.24

Fe

0.31

0.17

0.68

6.10

0.30

0.46

Cu

0.79

0.20

0.65

3.86

0.27

0.27

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D146 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization); C6 = D880 /D750 +D820 (asphaltene condensation).

In Table 2.3.14 and Figure 2.3.11b, the results of IR-spectroscopy analysis of asphaltenes, extracted from Cuban oil (Boca de Jaruco reservoir) are illustrated. To evaluate the nature of processes, as a result of which polycondensed structures are formed, the condensing degree of asphaltenes were determined. According to IR-spectroscopy of asphaltenes, the decrease in the content of aromatic structures after thermocatalytic influences is observed. This might be the result of C=Carom bond destruction processes and uncompression of aromatic rings. The minimum value corresponds to sample №5 (iron-based catalyst) and №3 (cobalt-based catalyst). The specific inverse correlation is observed for aliphaticity of the same samples, which rises almost six times. Because of uncompression and possible hydrogenation of aromatic rings and particularly destruction of C—C bonds in asphaltenes, the condensing degree of asphaltenes also significantly changes from initial oil sample. The minimum condensing coefficients are observed for noncatalytic and cobalt-based catalytic aquathermolysis. Oxidizing, branching, and sulfurization coefficients from specters of asphaltenes change negligibly. The work by Sitnov et al. (2019) discusses the physical simulation of aquathermolysis for heavy oil in the presence of minerals – dolomite and calcite. According to the SARA analysis and IR-spectroscopy, the presence of calcite provides significant effects on the reduction of resins and asphaltenes, their condensability and aromaticity by destructing the least stable C—S—C bonds. The results of IR-spectroscopy analysis of asphaltenes that have been extracted from model system are presented in Table 2.3.15 and Figure 2.3.12. The IR-data correlates with the results of SARA analysis. Hence, it indicates the significant changes in the structure of asphaltenes that are extracted from model systems. According to the spectral coefficient values, the sample that has been treated at 300 ∘ C in the presence of calcite is the most effective run. The decreasing of aromaticity and condensability of asphaltenes is related with the conduction of destructive processes of C—C and C—S—C bonds that resulted in the formation of smaller condensed structures, as well as detachment of alkyl substitutes and increase in the length of alkyl chains. Thus, this improves the aliphaticity parameter of n-paraffinic hydrocarbon fragments (Kayukova et al. 2017). In the work by Mukhamatdinov et al. (2019), we investigated the influence of steam treatment on structural group composition of resins and asphaltenes of heavy oil. The object of investigation was oil-saturated rocks from Riphean–Vendian complex. The extracted crude oil was determined

95

2.3 FTIR-Spectroscopy

Table 2.3.15

IR-analysis of the asphaltenes after experiments. After thermal treatment in the presence of minerals

Spectral

Initial oil

coefficientsa)

Dolomite

Calcite

200 ∘ C

300 ∘ C

200 ∘ C

300 ∘ C

C1

2.09

1.11

1.11

1.18

0.58

C2

0.17

0.22

0.38

0.19

0.28

C3

0.63

0.73

0.79

0.77

0.81

C4

2.48

2.78

2.44

2.70

3.26

C5

0.18

0.52

0.52

0.44

0.48

C6

0.61

0.40

0.42

0.41

0.13

806.94 745.54 724.41 698.64 662.91 614.19 596.48 579.09 559.30 531.18 519.29 506.69 485.99 474.00 463.25 455.34

862.61

1028.52

1111.88

1259.30

1310.53

1375.39

1454.83

1557.77

1599.49

1645.45

3

4

5

1696.23

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D1465 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 + D1380 )/D1600 (aliphaticity); C5 = D1030 /D1465 (sulfuration), C6 = D880 /D750 + D820 (condensed of asphaltenes).

D200 2

Absorbance units

D300

1

K200 K300

0

96

2000

1800

1600

1400 1200 1000 Wavenumber (cm–1)

800

600

Figure 2.3.12 IR spectra of asphaltenes after thermal treatment in the presence of minerals: D – dolomite (CaCO3 ⋅MgCO3 ), C – calcite (CaCO3 ).

as a high-viscous fluid. The resins and asphaltenes destructed to a small extent due to thermal treatment. The oil-soluble iron-based catalyst intensified the destructive processes. The content of sulfur compounds (–SO) in resins and asphaltenes drastically decreased due to reduction reaction of sulfoxide to sulfide and hydrogen sulfide. The results showed that catalytic aquathermolysis, even at low temperature ranges, promoted the cracking reaction of most macromolecular components and increased the content of light fractions of heavy oil. Consequently, it reduced its viscosity. The specters and spectral coefficients of resins are presented in Figure 2.3.13 and Table 2.3.16 accordingly. In the presence of catalyst, the coefficient of aromaticity was increased 1.5 times (150 ∘ C) and 3.5 times (200 ∘ C). It was seen from increasing intensity of absorption band at 1600 cm−1 . Inversely, the aliphaticity decreased due to condensation of aromatic rings. The oxidation coefficient did not change as no oxidation processes occurred. The sulfur compounds (containing –SO group) drastically decreased because of the reduction reaction of sulfoxide to

Catalyst Control

601.10

728.52

806.05

869.36

1022.12

1311.49 1261.97

1377.09

1458.90

1653.84 1603.65

Initial 150 °C 200 °C 150 °C 200 °C

0

1

2

Absorbance (a.u.) 3 4 5

1703.23

2.3 FTIR-Spectroscopy

2000

1800

1600

1400

1200 1000 Wavenumber (sm–1)

800

600

400

Figure 2.3.13 IR spectra of resins of initial extract and after thermal and thermal-catalytic treatment without and with catalyst. Table 2.3.16 Characterization of initial resins and treated resins after catalytic aquathermolysis according to IR spectroscopy data. After thermal treatment The initial extract

150 ∘ C

200 ∘ C

150 ∘ C

C1

1.74

1.89

0.77

2.86

2.75

C2

0.32

0.27

0.33

0.31

0.38

C3

0.71

0.60

0.64

0.63

0.68

C4

2.80

3.07

3.24

2.58

2.38

C5

0.48

0.39

0.34

0.29

0.33

Objects and experiment conditions

Spectral indicesa)

After thermal treatment with catalyst 200 ∘ C

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D146 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 + D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization).

sulfide and hydrogen sulfide. Besides, the catalysts increase alkyl side-chain length in contrast with blank run (without catalyst). Table 2.3.17 and Figure 2.3.14 present the results of IR-spectroscopy analysis of asphaltenes. The aromaticity of asphaltenes also increased in the presence of catalyst. The most increase corresponds to the treatment temperature of 150 ∘ C due to condensation of aromatic rings at the given temperature. The branching value of asphaltenes varies little in comparison with initial bitumoid. The aliphaticity value decreased with catalyst as the coefficient of aromaticity structures increased. The distinction of asphaltene specters is increasing absorption band of 1710 cm−1 . The distinct difference between the specters of asphaltenes is in increasing the absorption band of 170 cm−1 that stands for C=O oscillation of carbonyl group. It reaches its maximum at 200 ∘ C in the presence of catalyst. The content of sulfur compounds significantly reduced in resins and asphaltenes that are characterized by the content of –SO group. The reduction is due to redox reaction of sulfoxides to sulfides and hydrogen sulfides. The condensing degree of asphaltenes also changes due to aromatic ring condensation of asphaltene molecules in the presence of

97

2.3 FTIR-Spectroscopy

Table 2.3.17 Characterization of initial asphaltenes and treated asphaltenes after catalytic aquathermolysis according to IR spectroscopy data. After thermal treatment The initial extract

150 ∘ C

200 ∘ C

150 ∘ C

C1

2.39

2.57

2.67

3.79

3.39

C2

0.23

0.24

0.29

0.33

0.35

C3

0.73

0.75

0.75

0.74

0.74

C4

1.82

1.78

1.73

1.51

1.54

C5

0.37

0.45

0.44

0.34

0.39

C6

0.47

0.48

0.54

0.57

0.52

Objects and experiment conditions

Spectral indicesa)

After thermal treatment with catalyst 200 ∘ C

463.71

808.01 745.64 727.72 694.57

864.31

1029.24

1375.35

1454.57

1600.28

Initial Catalyst control

1

2

Absorbance (a.u.) 3 4

5

1694.18

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D146 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 + D1380 )/D1600 (paraffinicity); C5 = D1030 /D1465 (sulfurization); C6 = D880 /D750 + D820 (asphaltene condensation).

150 °C 200 °C 150 °C 200 °C

0

98

2000

1800

1600

1400

1200

Wavenumber

1000

800

600

400

(sm–1)

Figure 2.3.14 IR spectra of asphaltenes of initial extract and after thermal and thermal-catalytic treatment without and with catalyst.

catalyst. The most increase during the treatment was observed at a temperature of 150 ∘ C. Thus, the IR-specters of resin and asphaltene reveals the significant changes due to catalytic intensification of destructive processes. The chemical changes in the composition of asphaltene subfractions of heavy oil under the influence of oil-soluble Co-based catalyst were investigated (Mukhamatdinov et al. 2020a). The IR spectroscopy is one of the significant method among various methods of investigating the oil and oil product structures. It is based on absorbance, reflection, and dissipation of infrared energy passing through a substance. This method is more reliable than chemical analysis in the determination of functional groups. The IR spectral coefficients of the asphaltene subfractions are presented in Table 2.3.18, and their IR spectra are provided in Figures 2.3.15–2.3.17.

2.3 FTIR-Spectroscopy

Table 2.3.18 IR-analysis of the initial, after thermal treatment, after thermal-catalytic treatment by co carboxylate, after thermal-catalytic treatment by co carboxylate, and hydrogen donor asphaltene subfractions. Spectral coefficients

A1

A2

A3

A4

A5

Carbene– carboids

C1

1.01

0.84

0.73

0.67

1.15

1.09

C2

—

—

0.22

0.22

—

—

C3

0.75

0.79

0.62

0.71

0.80

0.82

C4

2.98

3.25

2.98

3.86

2.78

2.82

C5

0.47

0.43

1.15

0.40

0.43

0.45

C6

0.39

0.30

—

0.46

0.40

0.39

Initial

After thermal treatment C1

1.15

1.05

1.19

1.07

0.97

1.08

C2

—

—

—

0.43

—

—

C3

0.78

0.80

0.77

0.74

0.81

0.87

C4

2.65

3.01

2.73

2.98

3.11

2.85

C5

0.52

0.37

0.45

0.66

0.48

0.43

C6

0.24

0.26

0.34

0.29

0.41

0.31

After thermal-catalytic treatment by co carboxylate C1

1.27

1.11

1.05

1.19

0.91

1.39

C2

0.32

—

0.18

0.24

0.21

0.15

C3

0.76

0.80

0.79

0.78

0.75

0.76

C4

2.51

2.90

2.70

2.56

3.05

2.12

C5

0.44

0.44

0.60

0.54

0.39

0.95

C6

0.38

0.39

0.33

0.39

0.44

0.40

After thermal-catalytic treatment by co carboxylate and hydrogen donor C1

1,23

1,21

1,13

1,30

1,04

1,30

C2

0,29

0,17

—

—

—

0,09

C3

0,77

0,74

0,79

0,80

0,78

0,83

C4

2,68

2,70

2,73

2,57

2,76

2,42

C5

0,54

0,42

0,43

0,42

0,42

0,70

C6

0,34

0,42

0,37

0,43

0,41

0,34

A significant difference among the IR spectra of asphaltenes is the absence of 1710 cm−1 absorption band in the A1, A2, and A5 subfractions, corresponding to the fluctuations of C=O bond of carbonyl groups that reveals the low content of aromatic- and oxygen-containing compounds in the structural composition. The 1710 cm−1 absorption band, which is only presented in the A3 and A4 subfractions, have the constant spectral coefficient −0.22. Aromaticity of asphaltenes is drastically decreasing in the A4 subfraction −0.67, and a sharp increase in the A5 subfraction up to 1.15, that is probably indicating a denser arrangement of the aromatic rings. Analysis of aliphaticity coefficient and branching testify the heterogeneous change in their property. However, A5 fraction stands for

99

2.3 FTIR-Spectroscopy

1465 C‒H in CH2 groups

7

1600 C=C in benzene rings

720 CH2 > 4

A1 Initial Thermal Co-based catalyst +hydrogen donor A2 Initial Thermal

2

Absorbance (a.u.) 3 4 5 6

866; 804; 745 aromatic triplet

1030 S=O 1380 C‒H in CH3 groups

1

Co-based catalyst

0

+hydrogen donor

2000

1800

1600

1400

1200

1000

800

600

Wavenumber (cm–1) Figure 2.3.15

IR spectra A1–A2 subfractions.

7

1465 C–H in CH2 groups A3

6

C–O–C ethers 1030 1380 C–H in S=O CH3 groups

Initial

866; 804; 745 aromatic triplet

1600 C=C in benzene rings

720 CH2 > 4

5

Thermal

Absorbance (a.u.) 3 4

Co-based catalyst +hydrogen donor A4 Initial

2

Thermal

1

Co-based catalyst +hydrogen donor

0

100

2000

1800

1600

1400

1200

1000

800

600

Wavenumber (cm–1) Figure 2.3.16

IR spectra A3–A4 asphaltene subfractions.

the highest aromaticity index at minimum aliphaticity value comparing with another subfractions. A noticeable increase in the sulfurization occurred, which is characterized by –S=O groups, and the maximum value of which is observed in the A3 subfraction. The degree of condensation of asphaltenes with the maximum value for A4 subfraction is also attractive. The IR spectral analysis was also carried out for the asphaltenes of the oil after noncatalytic aquathermolysis. It should be noted that absorption band of the 1710 cm−1 in the A1, A2, A3, A5

2.3 FTIR-Spectroscopy

7

1465 C–H in CH2 groups

A5

1380 C–H in CH3 groups

1600 C=C in benzene rings

866; 804; 745 aromatic triplet

1030 S=O

720 CH2 > 4

6

Initial

Co-based catalyst +hydrogen donor C–C Initial Thermal

2

Absorbance (a.u.) 3 4 5

Thermal

1

Co-based catalyst

0

+hydrogen donor

2000

1800

1600

1400

1200

1000

800

600

Wavenumber (cm–1) Figure 2.3.17

IR spectra A5–carbene–carboids asphaltene subfractions.

subfractions and in the carbene–carboids is also absent as in the case of initial crude oil, which corresponds to the fluctuations of C=O carbonyl groups. The 1710 cm−1 absorption band is only presented in the A4 subfraction and has a spectral coefficient equal to 0.43. Aromaticity of asphaltenes has an extreme nature with the highest value in the A3 subfraction (1.19), probably indicating a denser arrangement of the aromatic rings. Moreover, the minimum aliphaticity value corresponds to A3 fraction. The sulfuring degree (–S=O groups) is significantly increasing in A1–A4, indicating the maximum peak (0.66) in A4 subfraction, as Ashal’cha oil is sour (Kayukova et al. 2016). The IR spectral analysis was also conducted for asphaltenes of oils after the thermal treatment using a Co-based catalyst. There are significant differences between the spectral ratios of maltenes of oils after catalytic aquathermolysis and noncatalytic aquathermolysis. First, the sulfurization indicator, which characterizes the content of –SO groups, is decreasing thus causing the reduction of sulfoxides to hydrogen sulphide in a presence of catalyst. Second, aromaticity is also decreasing and, as a result, aliphaticity is increasing, reflecting the destruction reactions of C-heteroatom (S, O, N) bonds in the maltene molecules. In the IR spectrum of the A1 subfraction (solvation shell) of asphaltenes, one can observe that the maximum absorption band is 1710 cm−1 , that corresponds to the fluctuations of C=O carbonyl groups reflecting the high content of aromatic and oxygenated compounds in their composition. Aromaticity of asphaltenes is decreasing in the A5 (0.91) subfraction with a sharp increase (up to 1.39) in residue on the filter (coke), indicating a denser arrangement of the aromatic rings. For the carbene–carboids, the maximum sulfuring degree and minimum aliphaticity value is noticed. Among all the five subfractions, the tendency of increasing sulfuring degree and decreasing aliphaticity disrupts only for A5 subfraction. The IR-analysis was carried out for asphaltenes from oil treated by catalyst precursor and hydrogen donor. In hydrothermal-catalytic conversion of oil, the increase in aromaticity and decrease in aliphaticity are observed. The comparison of IR-spectroscopy results of asphaltene subfractions revealed the decrease in intensity of absorption band at 1600 cm−1 with minimum in

101

2.3 FTIR-Spectroscopy

A3 subfraction with further increase toward carbene–carboid. The same changes are specific for aliphaticity as it is inversely proportional to aromaticity. The branching coefficient varies slightly. In the spectra of asphaltenes in the presence of a catalyst and a hydrogen donor, there is no absorption bands of 1710 cm−1 , corresponding to C=O carbonyl groups, starting from A3 and beyond. It should be noted that there is a noticeable decrease in the sulfurization index that characterizes the content of –S=O group. The minimum values were detected in both A2 and A5 subfractions. The changes in both coefficients justifies the cracking of C-heteroatom (S, O, N). The condensability coefficients of asphaltenes are ambiguous. In the work of Mukhamatdinov et al. (2020b), we isolated resins from a sample of extra heavy oil and from the products of catalytic aquathermolysis in the presence of a cobalt-based catalyst and studied their structure. In fact, an oil fraction and four resin fractions (R1–R4) have been extracted by liquid adsorption chromatography method in the presence of pure solvents and their binary mixtures in various ratios. MALDI data have showed that the average molecular weight of resins for each of the fractions has decreased, whereas IR spectroscopy has highlighted a change in their structural group composition. The picture of adsorption-chromatographic separation of resins components can significantly complement the results of IR spectral analysis. Table 2.3.19 and Figure 2.3.18 show the results Table 2.3.19 The results of IR spectral analysis of resins fractions of the control experiment (six hours).

0.52

—

—

C3

0.59

0.63

0.61

0.65

C4

7.90

6.37

34.93

4.26

Absorbance units 4 2 3

5

577.83 545.05 476.44 459.81

0.03

0.04

1304.92

0.44

—

1455.56

0.37

C2

1601.14

C1

747.42 724.23

R4

864.15 808.32

R3

1185.31 1165.24

R2

1376.24

R1

1032.71

Spectral coefficients

R1 R2

1

R3 R4

0

102

2000

Figure 2.3.18

1800

1600

1400 1200 1000 Wavenumber (cm–1)

800

600

IR spectra of resins fractions of the control experiment (six hours).

2.3 FTIR-Spectroscopy

of infrared spectral analysis of oil resins fractions of the control experiment after heat and steam exposure for six hours. According to IR spectroscopy results, all indicators are ambiguous due to the different nature of resin fractions constituent components. Literally, during heat and steam exposure for six hours, the aromaticity index possessed maximums and minimums and the oxidation coefficient was negligible in fraction R2. However, the heavier fraction R4 was characterized by a minimum aliphatic value (4.26 units) and a maximum aromaticity value (0.52 units), which characterizes the proximity to asphaltenes nature. Table 2.3.20 and Figure 2.3.19 provide IR spectral analysis results of oil resins fractions after heat and steam exposure with a cobalt-based catalyst precursor for six hours. Adding the catalyst to the system during heat and steam exposure for six hours leads to a significant improvement of the structural-group composition. Thus, the aliphaticity index of R3 decreased as a result of the detachment of alkyl substituents in aromatic fragments under the influence of catalyst and passes into lighter fractions of resins, as well as toward saturated and aromatic hydrocarbons. On the other hand, the oxidation coefficient also manifests itself in fractions R1 and R2 as a result of oxidative catalytic cracking processes, which also indicate a redistribution of the group composition.

—

C3

0.60

0.63

0.62

0.73

C4

7.88

7.33

27.87

6.64

Absorbance units 4 3 2

5

574.82 535.09 499.00 479.84 475.34 455.27

0.21

—

747.51 724.20

0.04

0.13

808.60

0.37

0.11

1033.14

0.38

C2

1247.38 1202.61 1166.64

C1

1305.94

R4

1376.33

R3

1455.57

R2

1601.60

R1

1712.92

Spectral coefficients

864.43

Table 2.3.20 The results of oil resins fractions IR spectral analysis in the presence of catalyst precursor (six hours).

R1 R2

R4

0

1

R3

2000

1800

1600

1400

1200

Wavenumber Figure 2.3.19

1000

800

600

(cm–1)

IR spectra of oil resins fractions in the presence of catalyst precursor (six hours).

103

2.3 FTIR-Spectroscopy

Mukhamatdinov et al. (2021b) investigate the effect of various solvents on the upgrading of heavy oil from the Ashal’cha field under thermal steam injection by means of aromatic and industrial solvents. IR spectral data have established that a significant difference between all spectra of oils is the absence of absorption bands at 1710 cm−1 , which corresponds to the vibrations of C=O carbonyl groups. The spectral coefficients and spectra of the original Ashal’cha oil as well as the control experiment and oil samples with aromatic solvents after thermal steam treatment are presented in Table 2.3.21 and Figure 2.3.20, respectively. The absence of absorption bands at 1710 cm−1 , which corresponds to the vibrations of C=O carbonyl groups is the only difference between all spectra of oil samples. During hydrothermal transformations of oil, there is a decrease in the aromaticity of oil and an increase in its aliphaticity. The results of IR spectroscopy of oil samples show a decrease in the intensity of absorption bands at 1600 cm−1 , when comparing the aromaticity coefficients in a series of aromatic solvents. For toluene, this indicator is the lowest (0.14 units), which probably indicates a less-significant condensation of aromatic rings compared to the original oil. The same changes are characteristic of the aliphatic index, since it is inversely proportional to aromaticity. In the Table 2.3.21 Spectral coefficients

Results of IR spectral analysis of oil samples. Without solvent

Initial

+ benzene

+ benzene + toluene

+ toluene

+ Solvesso-150

0.61

0.61

0.61

0.60

0.62

0.61

C4

7.43

8.03

7.32

12.24

10.20

6.85

C5

0.14

0.13

0.19

0.16

0.19

0.15

1

Absorbance units 4 3 2

5

471.53

C3

595.99

—

744.90 724.16

0.53

—

808.51

0.21

—

865.73

0.14

—

1031.84

0.50

—

1376.40

0.41

—

1455.93

0.49

C2

1602.60

C1

1 2 3 4 5 6

0

104

2000

1800

1600

1200 1400 Wavenumber (cm–1)

1000

800

600

Figure 2.3.20 IR spectra of Aschal’cha oil samples: 1 – initial, 2 – without solvent; 3 – with benzene; 4 – with toluene; 5 – with a mixture of toluene and benzene; 6 – with Solvesso-150.

2.3 FTIR-Spectroscopy

series of benzene-mixture of toluene and benzene–toluene, there is a significant increase in the aliphaticity coefficient, which indicates an intense detachment of side aliphatic substituents at aromatic rings as a result of steam thermal effect. The branching and sulfur content factors for all samples change insignificantly. The spectral coefficients and the spectra of the Aschal’cha initial oil as well as the control experiment and oil samples with industrial solvents after thermal steam effect are presented in Table 2.3.22 and Figure 2.3.21, respectively. The results of IR spectral analysis indicate that the addition of IS under steam thermal effect reduces significantly the aromaticity coefficient and increases the aliphaticity one, which confirms the directed action of the solvent on the destruction of resins and asphaltenes with an increase in the number of alkyl groups. It is worthy to note the significant decrease of oil sulfur content in the presence of petroleum ether which characterizes the content of –SO groups and which is the reason for the reduction of sulfoxides to sulfides. Changes in the structure-group composition of oil samples under thermal steam effect in the presence of solvents are ambiguous due to the completely different nature and class of the used individual compounds and their mixtures. The greatest changes detected by IR spectroscopy were found in oil samples with additives of toluene and IS. Results of IR spectral analysis of oil samples.

C1

0.49

0.41

0.48

0.42

0.30

0.46

C2

—

—

—

—

—

—

C3

0.61

0.61

0.61

0.63

0.63

0.61

C4

7.43

8.03

7.14

7.29

8.99

7.91

C5

0.14

0.13

0.11

0.16

0.17

0.12

1800

1600

Absorbance units 2 3 4

5

595.99

+ petroleum ether

808.51 744.90 724.16

+ IS

865.73

+ white spirit 5%

1031.84

+ white spirit 3%

1376.40

Without solvent

1455.93

Initial

1602.60

Spectral coefficients

471.53

Table 2.3.22

1 2 3 5 6

0

1

4

2000

1200 1400 Wavenumber (cm–1)

1000

800

600

Figure 2.3.21 IR spectra of Aschal’cha oil samples: 1 – initial, 2 – without solvent; 3 – with white spirit 3%; 4 – with white spirit 5%; 5 – with IS; 6 – with petroleum ether.

105

106

2.3 FTIR-Spectroscopy

In this work of Vakhin et al. (2021), we studied the catalytic performance of an oil-soluble, nickel-based catalyst during aquathermolysis of oil-saturated crushed cores from Boca de Jaruco extra-heavy oil field. The decomposition of nickel tallate and some aspects of in situ transformation of the given catalyst precursor under the steam injection conditions were investigated in a high-pressure batch reactor using XRD and SEM analysis methods. The changes in physical and chemical properties of core extracts after the catalytic aquathermolysis process with various duration were studied using gas chromatography for analyzing gas products, SARA analysis, GC-MS of saturated and aromatic fractions, FT-IR spectrometer, elemental analysis, and MALDI. The spectrometric indices of crushed core extracts before and after the catalytic upgrading depending on the duration of the process are summarized in Table 2.3.23, and their spectra are presented in Figure 2.3.22. The results exhibit a slight decrease in aromaticity index (C1) with increasing the duration of the catalytic upgrading due to loss of peripheral alkyl radicals and Table 2.3.23

FT-IR spectrometric indices of crushed core extracts. Spectral coefficientsa)

Experimental conditions

C1

C2

C3

C4

C5

Initial rock extracts

0.33

0.12

0.58

7.38

0.20

48 h

0.39

0.11

0.58

7.00

0.15

0.37

0.10

0.58

7.32

0.15

0.34

0.11

0.57

7.63

0.13

0.33

0.04

0.57

7.52

0.14

72 h 96 h

Hydrothermal treatment in the presence of catalyst and [H]-donor Without catalyst

a) C1 = D1600 /D720 (aromaticity); C2 = D1710 /D1465 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 +D1380 )/D1600 (aliphatic); and C5 = D1030 /D1465 (sulfurization).

4

3 Initial rock extract 2 48 hours 72 hours

1

96 hours 96 hours [h]-donor

0 600

800

1000

1200

Wavenumber Figure 2.3.22 300 ∘ C.

1400

1600

1800

(cm–1)

FT-IR of core extracts depending on the duration of catalytic hydrothermal treatment at

2.3 FTIR-Spectroscopy

long hydrocarbon chains from resins and asphaltenes. The previous statement is true if only aliphatic indices (C4) are increasing with decreasing aromaticity indices. This is explained by the production of low molecular weight alkanes after catalytic upgrading. The insignificant changes in oxidation (C2) and branching (C3) indices indicate the conduction of weak oxidative reactions initiated by water during catalytic aquathermolysis processes. Sulfurization index (C5) is decreasing with time due to reduction of sulfoxides to sulfides and hydrogen sulfides, which is inconsistent with gas chromatography results. The index of aliphatic (C4) is a maximum (7.63) after catalytic aquathermolysis for 96 hours due to the minimum aromaticity (0.34) achieved in this sample. This indicates the conduction of bond destruction reactions and ring opening of polycyclic aromatic hydrocarbons. In this work by Mukhamatdinov et al. (2021a), we studied the hydrogen-donating capacity of naphthenic and polar solvents during hydrothermal treatment of heavy oil from Ashal’cha reservoir (Republic of Tatarstan, Russia). It was established that naphthenic solvents significantly influence the viscosity reduction due to the destructive hydrogenation of resins and asphaltenes. Among naphthenic solvents, decalin significantly increases the amount of evolved gases, particularly C4–C10 isomers and aromatics. However, the maximum evolved gases among all the used solvents correspond to formic acid. The results of elemental analysis revealed that H/C ratio rises in crude oil samples after hydrothermal treatment in the presence of cyclohexane and decalin. FT-IR spectral indices revealed increase in crude oil aliphaticity in case of using solvent from cyclohexane-tetralin-decalin series due to the cleavage of carbon-heteroatom bonds in aliphatic substitutes of resins and asphaltenes. The significant changes in FT-IR spectra of crude oil are observed in the presence of tetralin. The spectrometric indices of initial crude oil from Ashalcha field, control sample, and crude oil after hydrothermal treatment in the presence of naphthenic solvents are presented in Table 2.3.24 and their spectra in Figure 2.3.23. Usually, C=O carbonyl group is uncharacteristic for Ashalcha reservoir crude oil and its aquathermolysis products. Therefore, the absorption band at 1710 cm−1 even in the presence of both aromatic and naphthenic solvents was not detected. The aliphaticity index rises in case of cyclohexane-tetralin-decalin due to destruction of carbon-heteroatom bonds in aliphatic substitutions of resins and asphaltenes. The branching parameter changes insignificantly. The sulfurization index, which characterize the content of –SO group, decreases. This decrease is due to oxidation of sulfides to sulfoxides in crude oil samples with tetralin. The spectrometric indices for initial crude oil sample, control sample, and oils with polar solvents are presented in Table 2.3.25 and their spectra in Figure 2.3.24. During hydrothermal-catalytic conversion of oil, aromaticity index increases, while aliphaticity decreases. The impact of polar solvents on spectrometric indices is negligible as the chemical nature of solvents differ from hydrocarbons. However, aliphaticity in samples with formic acid and glycerol Table 2.3.24

Results of FT-IR spectroscopy analysis of oil samples.

Spectral index

Initial

Control

+ cyclohexane

+ tetralin

+ decalin

C1

0.49

0.41

0.49

0.44

0.47

C2

—

—

—

—

—

C3

0.61

0.61

0.60

0.59

0.61

C4

7.43

8.03

7.25

7.51

7.61

C5

0.14

0.13

0.13

0.21

0.12

107

1

Absorbance units 2 3 4

471.53

595.99

744.90 724.16

808.51

865.73

1031.84

1376.40

1455.93

5

1602.60

2.3 FTIR-Spectroscopy

1 2 3 4 5

0

108

2000

1800

1600

1400 1200 Wavenumber (cm–1)

1000

800

600

Figure 2.3.23 FT-IR spectra of crude oil samples: 1 – initial (parent), 2 – control; 3 – with cyclohexane; 4 – with tetralin; 5 – with decalin. Table 2.3.25

FT-IR analysis of crude oil samples.

Spectral index

Initial

Control

+ formic acid

+ ethanol

+ glycerol

C1

0.49

0.41

0.47

0.47

0.48

C2

—

—

—

—

—

C3

0.61

0.61

0.63

0.61

0.62

C4

7.43

8.03

7.78

7.38

7.72

C5

0.14

0.13

0.14

0.12

0.13

are independent of aromaticity, rather than with ethanol. This indicates that the concentration of methylene group in crude oil sample is high. The branching and sulfurization parameters change insignificant. The changes in the structural-group composition of crude oil after hydrothermal treatment in the presence of solvents are diverging due to the chemical nature and class of applied individual solvents as well as their mixtures. The most change in FT-IR spectra is observed in crude oil sample with tetralin. The object of this study in the work of Sitnov et al. (2021) is a bituminous sandstone sample from the Ashal’cha reservoir. The catalytic (iron tallate) hydrothermal simulation was carried out under reservoir conditions (200 ∘ C, 30 bar). After the thermal processes, all aquathermolysis products from the model system (Table 2.3.26) and the initial rock sample are extracted by a warm mixture of organic solvents in Soxhlet extractor: chloroform, benzene, and alcohol are taken in equal ratios to obtain oil extracts. The composition and physicochemical characteristics of the products were studied using elemental and SARA analysis, MALDI, GC-MS, and FT-IR. The introduction of catalyst in combination with a hydrogen donor reduces the content of resins by 22 wt.% with an increase in the share of saturated hydrocarbons by 27 wt.% The destructive hydrogenation leads to a decrease in the sulfur content of upgrading products. This is crucial for the oil reservoirs of the Tatarstan Republic, as their crude oils are characterized by high sulfur content.

2

Absorbance units 3 4

471.53

595.99

744.90 724.16

808.51

865.73

1031.84

1376.40

1455.93

5

1602.60

2.3 FTIR-Spectroscopy

1 2 3

1

4

0

5 2000

1800

1600

1400 1200 Wavenumber (cm–1)

1000

800

600

Figure 2.3.24 FT-IR spectra of crude oil samples: 1 – initial (parent), 2 – control; 3 – with formic acid; 4 – with ethanol; 5 – with glycerol. Table 2.3.26

Model system.

Sample number

Subject of Research

1

Initial rock extract

2

Product of the noncatalytic aquathermolysis

3

Product of the catalytic aquathermolysis

4

Product of the catalytic aquathermolysis with hydrogen donor

Spectral coefficients of resins and asphaltenes of the studied oil extracts, calculated based on IR spectra (Figures 2.3.25 and 2.3.26), are presented in Tables 2.3.27 and 2.3.28. According to the results of IR spectroscopy, the structural changes in the molecules of resins and asphaltenes were characterized by a similar orientation, depending on the conditions taken to influence the initial extract in the rock. All considered samples showed that the aromaticity coefficient decreased (this can be seen from the decreasing intensity of absorption bands at 1600 cm−1 ) with an increasing trend in the aliphatic index for the fractions of resins and asphaltenes (for resins to a greater extent). A significant effect (C4 increases almost six times) was achieved during the aquathermolysis of oil in the presence of a mixture of a catalyst and a hydrogen donor. This was due to destructive processes in the molecules of resinous-asphaltene compounds that led to the formation of smaller condensed structures, the separation of alkyl substituents, and an increase in the length of alkyl chains. As a result, there is a decrease in the branching index of fragments of n-paraffin hydrocarbons (Ganeeva et al. 2014). A decrease in the degree of aromaticity due to thermocatalytic processes indicates a decrease in the stability of the colloidal state of resins and asphaltenes and, possibly, aggregation of asphaltenes to coarser particles up to precipitation from the system (Ganeeva et al. 2014). The oxidation coefficient, as expected, does not change since the process of steam treatment proceeds in the system of an inert gas – nitrogen. An equally important change that should be considered was a noticeable decrease in the sulfurization

109

666.40

753.56

809.01

1020.94

1000

867.57

1215.04

1304.69

1376.83

1200

3.0 2.5 2.0 1.5

Absorbance units

3.5

4.0

1697.45

1602.21

1455.96

2.3 FTIR-Spectroscopy

4

2 1

0.0

0.5

1.0

3

2000

1800

1600

1400

800

600

Wavenumber (cm–1)

752.32 728.31 666.16 600.14 574.79 550.81 537.79 525.89 515.53 494.24 481.40 469.89 456.94

808.13

861.94

1031.01

1096.20

1283.81 1261.76

1375.28

1454.23

1595.21

IR spectra of resins of oil extracts.

1727.41 1696.42

Figure 2.3.25

Absorbance units 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

110

4 3 2 1

2000

1800

1600

1400

1200

1000

800

600

Wavenumber (cm–1) Figure 2.3.26

IR spectra of asphaltenes of oil extracts.

index, characterizing the content of –SO groups, which is the reason for reducing sulfoxides to sulfides and hydrogen sulfide. This indicates the intensification of the hydrogenolysis reactions of carbon-heteroatom bonds. The condensity of asphaltenes also significantly changes due to the decompaction of their structure and partial destruction and hydrogenation of C—C bonds in the presence of a catalyst and a mixture of a hydrogen donor and a catalyst.

References

Table 2.3.27

Results of IR spectral analysis of resins. Subject of research

Spectral coefficients

1

2

3

4

C1

0.72

0.24

0.15

0.04

C2

0.22

0.23

0.22

0.22

C3

0.63

0.64

0.62

0.62

C4

4.37

7.05

9.31

25.45

C5

0.65

0.53

0.43

0.37

C1 = D1600 /D720 (aromaticity); C2 = D1710 /D1465 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 + D1380 )/D1600 (aliphaticity); C5 = D1030 /D1465 (sulfurization index).

Table 2.3.28

Results of IR spectral analysis of asphaltenes. Subject of research

Spectral coefficients

1

2

3

4

C1

1.18

0.33

0.94

0.97

C2

0.30

0.35

0.31

0.28

C3

0.77

0.78

0.75

0.76

C4

2.63

4.63

2.73

2.76

C5

0.63

0.54

0.43

0.42

C6

0.38

0.45

0.28

0.28

C1 = D1600 /D720 (aromaticity); C2 = D1710 /D1465 (oxidation); C3 = D1380 /D1465 (branching); C4 = (D720 + D1380 )/D1600 (aliphaticity); C5 = D1030 /D1465 (sulfurization index); C6 = D880 /D750 + D820 (condensity).

References Abdrafikova, I.M., Ramazanova, A.I., Kayukova, G.P. et al. (2013). Structural-group composition of the products of heavy Ashal’cha oil conversion by the method of IR Fourier spectroscopy. Bulletin of Kazan Technological University 16 (7): 237–242. (in Russian). Agaev, S.G., Zemlyansky, E.O., and Gultyaev, S.V. (2006). Paraffin deposits of the Verkhnesalatskoye oil field of the Tomsk region. Oil Refining and Petrochemistry (3): 8–12. (in Russian). Bellamy, L.J. (1975). The Infrared Spectra of Complex Molecules, 433. Dordrecht: Springer https://doi .org/10.1007/978-94-011-6017-9. Ganeeva, Y.M., Yusupova, T.N., Romanov, G.V., and Bashkirtseva, N.Y. (2014). Phase composition of asphaltenes. Journal of Thermal Analysis and Calorimetry 115: 1593–1600. Grin’ko, A.A. and Golovko, A.K. (2014). Thermolysis of petroleum asphaltenes and their fractions. Petroleum Chemistry 54: 42–47. https://doi.org/10.1134/S0965544113040051. Ivanova, L.V., Safieva, R.Z., and Koshelev, V.N. (2008). IR spectroscopy in the analysis of oil and oil products. Bulletin of the Bashkir University 13 (4): 869–874. (in Russian).

111

112

2.3 FTIR-Spectroscopy

Kayukova, G.P., Uspensky, B.V., Abdrafikova, I.M., and Musin, R.Z. (2016). Characteristic features of the hydrocarbon composition of Spiridonovskoe (Tatarstan) and Pitch Lake (Trinidad and Tobago) asphaltites. Petroleum Chemistry 56: 572–579. Kayukova, G.P., Foss, L.E., Feoktistov, D.A. et al. (2017). Transformations of hydrocarbons of Ashal’hinskoe heavy oil under catalytic aquathermolysis conditions. Petroleum Chemistry 57: 657–665. https://doi.org/10.1134/S0965544117050061. Mukhamatdinov, I.I., Sitnov, S.A., Slavkina, O.V. et al. (2019). The aquathermolysis of heavy oil from Riphean-Vendian complex with iron-based catalyst: FT-IR spectroscopy data. Petroleum Science and Technology 37 (12): 1410–1416. https://doi.org/10.1080/10916466.2019.1587464. Mukhamatdinov, I.I., Salih, I.S.S., Rakhmatullin, I.Z. et al. (2020a). Influence of Co-based catalyst on subfractional composition of heavy oil asphaltenes during aquathermolysis. Journal of Petroleum Science and Engineering 186: https://doi.org/10.1016/j.petrol.2019.106721. Mukhamatdinov, I.I., Salih, I.S.S., Ziganshina, M.R., and Onishchenko Ya.V., Feoktistov D.A. (2020b). Fractionation of Aschalcha extra heavy oil resins and studying their structure. IOP Conference Series: Earth and Environmental Science 516: 012034. https://doi.org/10.1088/1755-1315/516/1/012034. Mukhamatdinov, I.I., Salih, I.S.S., Ismael, M. et al. (2021a). Influence of naphthenic hydrocarbons and polar solvents on the composition and structure of heavy oil aquathermolysis products. Industrial & Engineering Chemistry Research 60 (36): 13191–13203. https://doi.org/10.1021/acs.iecr.1c02341. Mukhamatdinov, I.I., Salih, I.S.S., Khelkhal, M.A., and Vakhin, A.V. (2021b). Application of aromatic and industrial solvents for enhancing heavy oil recovery from the ashalcha field. Energy and Fuels 35 (1): 374–385. https://doi.org/10.1021/acs.energyfuels.0c03090. Onishchenko, Y.V., Vakhin, A.V., Chemodanov, A.E. et al. (2017). Laboratory modeling of catagenesis of organic matter from a Domanic shale of Tatarstan in hydrous pyrolysis experiments. In: 17th International Multidisciplinary Scientific GeoConference – SGEM, vol. 17, issue 15, 281–288. 10.5593/sgem2017H/15/S06.036. Petrova, L.M., Abbakumova, N.A., Foss, T.R., and Romanov, G.V. (2011). Structural features of asphaltene and petroleum resin fractions. Petroleum Chemistry 51: https://doi.org/10.1134/ S0965544111040062. Salih, I.Sh.S., Mukhamatdinov, I.I., and Akhmetov, B.R. (2018a). Redistribution of subfractional composition of resins after hydrothermal catalytic influences.In: 18th International Multidisciplinary Scientific GeoConference, SGEM, vol. 18, issue 14, 735–742. 10.5593/sgem2018/1.4/S06.096. Salih, I.S.S., Mukhamatdinov, I.I., Garifullina, E.I., and Vakhin, A.V. (2018b). Study of fractional composition of asphaltenes in hydrocarbon material. Chemistry and Technology of Fuels and Oils 54 (1): 44–50. https://doi.org/10.1007/s10553-018-0896-9. Sitnov, S.A., Vakhin, A.V., Mukhamatdinov, I.I. et al. (2017). Aquatermolysis of heavy oil in reservoir conditions with the use of ultradispersed Co-based catalyst. In: 7th International Multidisciplinary Scientific GeoConference – SGEM, vol. 17, issue 15, 45–52. 10.5593/sgem2017H/15/S06.006. Sitnov, S.A., Vakhin, A.V., Mukhamatdinov, I.I. et al. (2019). Effects of calcite and dolomite on conversion of heavy oil under subcritical condition. Petroleum Science and Technology 37 (6): 687–693. https://doi.org/10.1080/10916466.2018.1564766. Sitnov, S.A., Mukhamatdinov, I.I., Feoktistov, D.A. et al. (2021). Underground upgrading of the heavy crude oil in content-saturated sandstone with aquathermolysis in the presence of an iron based catalyst. Catalysts 11 (10): 1255. https://doi.org/10.3390/catal11101255. Svarovskaya, L.I., Filatov, D.A., Gerelmaa, T., and Altuninat, L.K. (2009). IR and 1 H NMR assessments of the biodegradation of oil. Petroleum Chemistry 49: 136–141. https://doi.org/10.1134/ S0965544109020066.

References

Vakhin, A.V., Sitnov, S.A., Mukhamatdinov, I.I. et al. (2017). Aquathermolysis of high-viscosity oil in the presence of an oil-soluble iron-based catalyst. Chemistry and Technology of Fuels and Oils 53 (5): 666–674. Vakhin, A.V., Aliev, F.A., Kudryashov, S.I. et al. (2018a). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: part I–changes in composition of saturated hydrocarbons. Petroleum Science and Technology 36 (21): 1829–1836. https://doi.org/10.1080/ 10916466.2018.1514411. Vakhin, A.V., Mukhamatdinov, I.I., Aliev, F.A. et al. (2018b). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: part II–changes in composition of aromatic hydrocarbons. Petroleum Science and Technology 36 (22): 1850–1856. https://doi.org/10.1080/ 10916466.2018.1514412. Vakhin, A.V., Sitnov, S.A., Mukhamatdinov, I.I. et al. (2018c). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: part III–changes in composition resins and asphaltenes. Petroleum Science and Technology 36 (22): 1857–1863. https://doi.org/10.1080/10916466 .2018.1514413. Vakhin, A.V., Aliev, F.A., Mukhamatdinov, I.I. et al. (2021). Extra-heavy oil aquathermolysis using nickel-based catalyst: some aspects of in-situ transformation of catalyst precursor. Catalysts 11 (2): 189. https://doi.org/10.3390/catal11020189. Yusupova, T.N., Ganeeva, Y.M., Romanov, G.V. et al. (2017). Change in the structural-group composition of bitumen asphaltenes upon thermal bitumen recovery. Petroleum Chemistry 57: 198–202. https://doi.org/10.1134/S0965544117020256.

113

114

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods Roman S. Borisov, Anton L. Maximov, Anastasiia Yu. Kanateva, and Vladimir G. Zaikin A.V. Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, 29 Leninsky prospect, 119991 Moscow, Russia

2.4.1

Introduction

Being one of the most important and versatile natural resources for fuels and various chemicals, oil has an extremely complex composition and contains millions of organic and organo-element compounds that differ in molecular weight, structure, chemical and physical properties, and quantitative contents in oils. A rather complex composition is also preserved in petroleum fractions isolated for specific needs by distillation or extraction as well as in oil refining products. Among the wide variety of oils, heavy (density 0.921–1.000 g/cm3 ) and super-heavy oils and natural bitumen (density above 1.000 g/cm3 ) have their own characteristics owing to their chemical compositions. In contrast to light oils, they contain much larger amounts of resinous and asphaltene substances, nitrogen-, chlorine-, oxygen-, sulfur-containing compounds, and even metals. Due to such an extraordinary variety of organic compound compositions of different oils, it has not yet been possible to apply any single physicochemical method to their complete qualitative and quantitative description. However, at present, it is quite obvious that among the known spectral methods, mass spectrometry (MS) has turned out to be one of the most effective structural-analytical technique for the investigation of such a complex mixture as petroleum. In fact, MS frequently provides the fundamental knowledge on the petroleum compositions that usually serves as the basis for creating standardized approaches for oil refining processes in industry, for their monitoring and evaluating the quality of products, usually representing extremely complex mixtures of organic substances. MS turned out to be a virtually irreplaceable tool. In addition, the tasks for the analysis of various crude oils are being challenged and monitoring to address geochemical studies such as oil origin, its migration, diagenesis, and catagenesis in deposits. Another important issue is due to the need to provide suitable examination of oil spills on soil and water surfaces that is crucial for a rapid and complete biodegradation of oils and oil products and can be used for arbitration purposes. It should be underlined that a variety of MS techniques has various capabilities and limitations but each of them allows studying only a certain part of the components of petroleum feedstock and none of them is completely universal. Indeed, at present, there is no single mass spectrometric method by which one could comprehensively characterize the composition of crude oil, usually consisting of hundreds of thousands of compounds of various structures, elemental compositions, volatility, and MS ionizability. Therefore, the study of petroleum is usually based on the use of a set of mass spectrometric methods, which require the development of approaches for complete interpretation of such huge composite data. However, intensive development of mass spectral technology lays the foundation for obtaining ever wider arrays of data on the composition of oils in

2.4.2 Chromatography Methods

a single experiment. This was facilitated by the development of both new ionization techniques and methods for analyzing ions, implemented in high-resolution mass spectrometers. As a result, a new field in petroleum chemistry named “petroleomics” was born. This term was proposed by A. Marshall for the description of the aggregate of data obtained using ultrahigh-resolution mass spectrometry (uHRMS) (Marshall and Rodgers 2004). However, comprehensive investigation of petroleum in such a way requires the use of a set of relatively recent and traditional mass spectrometric options. In this case, the choice of a particular mass spectrometric approach is determined by the specificity and limitations of the ionization methods employed. In addition, until now, a fairly general methodology for the complete characterization of a wide variety of oils (including heavy and super-heavy) and petroleum products by the mass spectrometric method necessitates frequently the off-line or online use of various separation techniques (fractionation, chromatography, etc.) and mass spectrometric equipment that can be hyphenated with each other. As is known, chromatographic separation mainly by gas chromatography (GC) and highperformance liquid chromatography (HPLC) in online combination with mass spectrometric analysis is a fairly common methodology in the study of a wide variety of oils. However, these methods manage to investigate only those fractions that are volatile and separated by such chromatography methods. At the same time, other mass spectrometry techniques are necessary to investigate polar, low volatility nonpolar, and unresolved and high molecular petroleum components for their further refining and processing into useful products. “Soft” desorption/ionization methods, which are especially applicable to the analysis of heavy, poorly soluble components of oils, have great potential in such studies. They usually operate at atmospheric pressure and permit the desorption/ionization of solid samples which are very suitable for the investigation of heavy petroleum fractions. In such analyses, uHRMS plays a fundamental role because it allows the high-precision determinations of molecular weights and elemental compositions of individual components, even in unseparated mixtures. This chapter is devoted to this issue, where the possibilities of various mass spectrometric methods for the study of heavy and super-heavy oils, their fractions and industrial products are considered. The potential of applying the combined chromatography-mass spectrometry techniques is also the focus of the present review. Work in the development and practical application of MS for the analysis and comprehensive study of crude oils is continuously progressing. Some common aspects of such an application of MS have been discussed in a number of reviews (see, for example, Marshall and Rodgers 2004; Rodgers and McKenna 2011; Niyonsaba et al. 2019; Borisov et al. 2019; Qian 2021).

2.4.2

Chromatography Methods

Chromatography is one of the most powerful analytical tools dealing with complex mixtures such as crude oils or oil fractions. Combining chromatographic separations to MS allows receiving the time larger information for subsequent analysis. The prevalent chromatographic method for petroleum and oil fractions separation is GC combined with different detectors depending on the task; however, HPLC is also used. Here, we will discuss the most common applications of these chromatographic methods to crude oil and fraction analysis.

2.4.2.1

Gas Chromatography

GC is the first-choice method for investigation of complex mixtures of volatile thermotolerant compounds. GC separation supposes the evaporation of the sample in the injector with subsequent separation in the column containing stationary phases of different nature varying in the polarity

115

116

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

and detection with (pseudo)universal or selective detectors. The heart of the GC system is chromatographic column ensuring separation of the analytes. Unfortunately, none of the existing chromatographic columns allows separation such as complex mixtures as oil and its fractions to individual compounds. The decision is usually either in utilization of “fingerprinting” when the samples are compared according to the total profile of analytes without identification and quantification of all the individual compounds, or in group analysis, or in selective detection utilization. 2.4.2.1.1 Fingerprinting

Oil fingerprinting is essential while working with oil spills for designation of the source of spill and for response to the extraordinary situations (Yang et al. 2016). However, fingerprinting of oil spills can be rather challengeable to an analyst due to the complex composition of crude oils and fractions, so capillary columns with low-polar stationary phases are usually used. The main aim of fingerprinting analysis is usually determination of total petroleum hydrocarbons (TPHs) and/or target petroleum hydrocarbons in the crude oil or oil fraction. TPHs might be determined using GC system equipped with either of both a flame-ionization detector and mass spectrometer, or FID only. Yang et al. (2022) demonstrated the possibility of oil spills fingerprinting using low polar 5% phenyl polydimethylsiloxane capillary column (DB-5 ms, 30 m × 0.25 mm i.d., 0.25 μm film thickness). The carrier gas was helium, and 1.0 μl of each sample solution was injected in splitless mode. After column separation, the sample was split at a ratio of 1 : 1, and each part directed into the MS and FID detector simultaneously. The oven temperature was programmed from 42 to 320 ∘ C. However, the injector temperature was set only at 300 ∘ C, which is undesirable as some high-boiling compounds might be deposited in the injector. The injector temperature usually is recommended to be equal or higher than the column temperature at the same moment. The FID detector and MS ion source were both set at 300 ∘ C, which was also lower than the maximal column temperature. This may result in condensation of some analytes from the high-boiling samples and deterioration of detector signal. From the good practices point of view, detectors temperatures in GC should be set from 10 to 20 ∘ C higher than the maximal column temperature during the separation. Target petroleum hydrocarbons were detected by MS using selected ion monitoring (SIM) mode. The quantitation of target compounds was successfully performed based on FID signal. It is worth noting that splitless mode is essential for investigation of crude oil due for the possibility of sample composition discrimination with split injection. It is known (Zenkevich and Olisov 2019; Barwick 1999) that the effect of the discrimination of sample composition, or the nonlinearity of splitting, which interferes with the results of quantitative determinations, has been known since the appearance of split injection technique. Most often, the composition discrimination is subjected as a decrease in the peak areas of high-boiling components compared to the peak areas of low-boiling components. If the mass concentrations of a series of homologues in the injected sample are equal, their corresponding peak areas can decrease with increase of the number of carbon atoms in the molecules with split injection. These effects are expressed to the greatest extent on the manual injection of the samples with a wide range of boiling points with a syringe into a heated injector to exclude this effect on-column or programmed temperature vaporization injection may be used (Engewald et al. 1999). Cold large volume injection may be used for the injection of crude oil and oil fractions solutions which is the most often laboratory practice. This injection method allows utilization of both split/splitless injection in one cycle: the solvent peak is excluded from the chromatogram via cold injection with subsequent selective evaporation of a solvent being the most volatile compound of the mixture in the split mode of the PTV injector, and the splitless mode then used at higher injector temperatures for transportation of the other compounds into the GC column (Pavón et al. 2008).

2.4.2 Chromatography Methods

Chua et al. (2020) demonstrated the possibilities of fingerprinting of the artificially weathered crude oil spill samples for matching them to unweathered possible parent oils. The study proved the possibility for utilization of the method in crude oil forensics using a combination of principal component analysis in tandem with traditional forensic oil fingerprinting tools to ensure the additional confidence in challenging oil spill cases. The four parent oils and two weathered samples were investigated, and each of them was injected on gas chromatograph equipped with a flame ionization detector and standard split injector (Figure 2.4.1). A traditional low polar 5% phenyl polydimethylsiloxane (Restek Rtx-5DB-5ms)-fused silica column (30 m × 0.32 mm id × 0.25 μm film thickness) was used for separation. Rather large volume of two μls was injected, so the split mode had to be used with a split ratio of 1 : 25. The temperature programming was from 45 to 325 ∘ C with detector temperature equal to maximal column temperature of 325 ∘ C. In the study, weathering was considered as physical and chemical changes to the spilled oils composition, thereby decreasing the reliability of both GC/FID and GC/MS diagnostic ratios in source attribution of the spills, especially taking into consideration that percentage of high-boiling compounds is objectively increased with time, and these are mainly discriminated in wide samples. The shortcomings of these traditional methods were overcome by applying multivariate statistical tools that enabled accurate characterization of the crude oil spill samples. Three goals in crude oils and spills GC fingerprinting were achieved: established a working approach to crude oil processing using principal component analysis, developed a three-step approach consisting of GC/FID and GC/MS diagnostic ratios, and multivariate statistics, and have successfully applied the developed method to match two weathered spill samples to their unweathered parent oil sources. 2.4.2.1.2 Group Analysis and Simulated Distillation

Giordano et al. (2021) suggested characterization of crude oil and fractions by GC with flame ionization detector to control the process of laboratory model distillation of small volumes of crude oil. The solution (1 : 10 oil fraction to dichloromethane) with split injection was also used to decrease the amount of the sample due to low-sample capacity of capillary columns, so the standard sample of hydrocarbons (C8 –C40 ) separation was necessary for semiquantification of the most important hydrocarbons of the target oil fractions. For chromatographic analyses, the injection was 1 μl with a split ratio of 1 : 15. The low-polar column was also used (5% phenyl polydimethylsiloxane) with dimensions of 30 m × 0.25 mm × 0.25 μm with oven temperature programming from 40 to 330 ∘ C. Using the received data, Giordano et al. (2021) have characterized the oil fractions with distillation (true boiling point [TBP]) curves: the six distillate fractions generated by the suggested system were plotted versus their measured total percent mass (% w/w). The resulting TBP curve, plotted based on the chromatographic data, showed the potential of the suggested mini-distillation platform to attain TBP curves from 70 to 150 ∘ C (Figure 2.4.2). Li et al. (2019) demonstrated the qualitative and quantitative analysis of the light fractions of coal-based crude oils, which were conducted by GC-MS and GC-FID methods, correspondingly, under the conditions of material balance calculation. A rather simple and accurate quantitative method was developed by using the relative response factor prediction formula based on the molecular formula of the compounds (GC-MS) combined with the area normalization method (GC-FID). The methods were based on the determination of about three hundred compounds. The suggested method allowed quantification and balancing of approximately 70% (w/w) of the components in the whole light weight fractions of coal-based crude oils. According to the received results, the carbon number of the coal-based crude oils was mainly below 16. The amount of hydrocarbons (∼61% w/w) was larger than that of compounds with heteroatoms (∼17%) in coal liquefied oil, which was mainly composed of alkyl tetrahydronaphthalene, alkyl naphthalene, and phenols. As for the coal tar, where phenols and alkyl naphthalene were the main components, the amount of compounds

117

118

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

×102 1

+ EI TIC Scan S Bunker C 1.D

Bunker C

0.8 0.6 0.4 0.2 0 10

×102

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

+ EI TIC Scan S Burnaby oil spill 1.D

60

65

Burnaby oil spill

1

0.8 0.6 0.4 0.2 0

×102

10

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

60

+ EI TIC Scan S Cold Lake 1.D

65

Cold Lake

1

0.8 0.6 0.4 0.2 0

×102

10

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

+ EI TIC Scan S Gulf of Mexico 1.D

60

65

Gulf of Mexico

1

0.8 0.6 0.4 0.2 0

×102

10

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

60

+ EI TIC Scan QA1.D

65

QA1

1

0.8 0.6 0.4 0.2 0

×102 1

10

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

60

65

QA2

+ EI TIC Scan QA2.D

0.8 0.6 0.4 0.2 0

10

15

20

25

30 35 40 45 50 Counts (%) versus Acquisition time (min)

55

60

65

Figure 2.4.1 GC-MS TIC chromatograms – the profiles of the samples used for the screening purposes. The results of the TIC analysis supported those of the GC/FID analysis. Each of the four potential source oils showed a distinctive pattern of peaks. The weathered sample QA1 was best matched to the Burnaby oil spill sample; the QA2 sample was best compared to the Gulf of Mexico oil. Source: Chua et al. (2020)/ Reproduced with permission from Elsevier.

Relative quantity(%)

370 min

319 min

289 min

239 min

2

4

6 8 10 12 14 16 Retention time (min)

(b) 30 Naphtha 25 20 15 10 5 0 25 20 15 10 5 0 30 24 18 12 6 0 40 32 24 16 8 0 60 50 40 30 20 10 0 6 7 8

18

Kerosene

Diesel 400 min

370 min

(c)

100

Relative quantity (%)

400 min

Intensity (arbitrary unit)

(a) 1600 1200 800 400 0 1600 1200 800 400 0 2000 1500 1000 500 0 2000 1500 1000 500 0 2400 1800 1200 600 0

Diesel collection Start

80

40

Naphtha

20 0

Diesel 200

319 min

Kerosene

60

250 300 350 Distillation time (min)

400

(d)

289 min

239

319

400

239 min

9 10 11 12 13 14 15 16 Number of carbons

Figure 2.4.2 The chromatographic separation results (GC-FID) for distillate oil fractions (a) and relative quantities of hydrocarbons in relation to the total number of carbons in five fractions collected along the distillation of a crude oil sample as pointed in the figure (from 239 to 400 minutes) (b). The received data allowed plotting the gelative quantity plots of naphtha, kerosene, and diesel fractions versus distillation time (c). Photos of samples distilled after 239, 319, and 400 minutes of distillation demonstrate the visual difference between distillate oil fractions (d). Source: Giordano et al. (2021)/with permission from ELSEVIER.

120

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

containing heteroatoms (∼44%) was in contradiction to liquefied oil higher than those of the hydrocarbons (∼30%). Rodrigues et al. (2018) have used the high temperature GC separation data of 98 petroleum samples from the Brazilian coast sedimentary basin to access to group composition of the oil samples which may be then correlated to the physicochemical properties of petroleum and oil fractions, which in turn is essential for decision-making during the production process. TBP curve is one of the most important parameters in the crude oil characterization, and this parameter can be estimated by high temperature gas chromatography (HTGC). Rodrigues et al. (2018) have used the HTGC technique together with partial least squares regression to estimate the values of API gravity (density compared to water), kinematic viscosity, pour point, carbon residue, saturated, and aromatic content in crude oil. This method uses the analytical signal (chromatogram peak area) and the information of interest (the parameters of the petroleum samples) enabling a mathematical relation between instrumental response and a determined property of the sample. The samples from light to heavy with API gravity ranging from 11.4 to 54.0 were investigated, where crudes with API gravity lower than 22.3 are classified as heavy oils, and with API gravity from 22.3 to 31.1 as medium oils. The developed methods were then applied for investigation of nine unknown samples from a field of production of the Brazilian coast. PLS model for API gravity, carbon residue, saturates, and aromatics contents showed the acceptable prediction accuracy with significance level of 5%. One of the classical GC methods for group analysis is simulated distillation of petroleum, oil fractions, and heavy residues. It is known that distillation is the most used method for determining the boiling temperatures range of crude oil and oil fractions. However, classical distillation methods are time and reagent consuming demanding several days and several liters of products for TBP characterization. In 1960s, Eggerston et al. (1960), and Green et al. (1964) proposed utilization of GC separation for simulation of the TBP conventional method. The concept was called simulated distillation and was based on the idea that the components of the sample are eluted from the chromatographic column in order of their boiling point, which is true for hydrocarbon mixtures on the nonpolar sorbents. Utilization of mixture of normal alkanes with known boiling points allows correlation of the retention times of analytes to their boiling temperatures. Furthermore, based on the chromatographic zone area, the curve describing the dependence of the boiling point on the percentage of the corresponding fraction can be constructed. GC method was in good correlation with the classical TBP methods results and ensured significant decrease of analysis time (the separation process was finished in an order of one hour) and sample consumption (microliters versus liters), and due to these reasons simulated distillation has become a predominant analytical tool for the characterization of petroleum products in research and refinery laboratories. Many works and articles have been published and several methods depending on the boiling point range were developed and standardized. At the present time, several methods are used in laboratories (Durand et al. 1999). The main of them are conventional simulated distillation (CSD) method for carbon number distribution from C5 to C72 (35–650 ∘ C); high temperature simulated distillation (HTSD) method for hydrocarbons from C10 to C100 (174–720 ∘ C), and residual simulated distillation (RSD) method for carbon numbers from C5 to C60 (35–615 ∘ C) on samples containing nondistillable fractions (asphaltenes) by using the capillary column of the CSD method with a precolumn. Capillary columns for simulated distillation are usually shorter than those for other analytical applications (10–15 m) and contain the nonpolar stationary phase such as 100% polydimethylsiloxane. Sample injection in simulated distillation methods is usually organized via program temperature vaporization detector with cool injection to avoid sample discrimination in the high-boiling area.

2.4.2 Chromatography Methods

2.4.2.1.3 Selective Detection

Among the selective detectors for GC, the most used are sulfur and phosphorus selective (pulsed) flame photometric detectors (FPD/PFPD) which are widely used for investigation of mercaptans, thiophenes, and so forth; electron capture detector (ECD) being highly sensitive to electron-acceptable compounds, for example, halogenated hydrocarbons, nitrogen-phosphorus detector (NPD) utilizing K or Rb inorganic salt for N- and P-containing compounds detection. Separately stands the MS detection giving the analyst opportunity of SIM and tandem MS/MS modes. The MS detection will be in detail described in Sections 2.4.4 and 2.4.5 (e.g. Section 2.4.4), and here, we will give only one example. Dias et al. (2021) have presented the optimization of a GC method with tandem MS/MS detection for the determination of nitrogen-containing heterocyclic aromatic compounds assumed to be geochemistry markers in crude oils. The parameters for the operation of tandem mode MS/MS were optimized to provide multiple reaction monitoring (MRM) data. Dias et al. (2021) also have evaluated the matrix effect and have used for quantification the matrix-matched calibration. The application of the developed method to eight crude oil samples for their geochemical evaluation without preliminary fractionation steps which allowed shortening analysis time and increasing the energy efficiency of method was demonstrated. The results indicated that the GC-MS/MS method was promising for fast and efficient quantitative analysis of N-markers in crude oils, and it minimized sample preparation time and solvent consumption. Flame photometric detector (FPD) and pulsed flame photometric detector (PFPD) are among the major selective detectors in GC for sulfur detection in petroleum (Yue et al. 2015; Wang et al. 2009). Both sulfur detectors are equipped with photomultiplier tube as photo detector since it provides fast response and extremely high sensitivity to photons emitted from S2 * particles due to quadratic signal. However, these detectors have some disadvantages (Geng et al. 2017, 2018), including relatively high cost, long equilibration time up to 50 minutes for some configurations, fragile to strong light radiation and to electromagnetic interference and vibration. Recently, Ni et al. (2020) have suggested the FPD with a silicon photodiode assembly instead of a photomultiplier tube for sulfur detection to overcome the disadvantages of the classical sulfur photometric detection. The photosensitive area was optimized for the suggested photodiode, its optical design, and bandpass filters. The newly suggested design allowed full utilization of all the wide emission spectrum of S2 *. Ni et al. have determined the detection limits for nine sulfur-containing compounds which are the model compounds of sulfur (2-propanethiol, 1-methyl-1-propanethiol, diethyl sulfide, 1-butanethiol, tetrahydrothiophene, dipropylsulfide, isopropyl disulfide, thianaphthene, and methyl-parathion) and were between 5.8 × 10−12 for thianaphthene to 9.5 × 10−12 g S/s for 2-propanethiol. The developed detector ensured the linear response of three orders of magnitude for the tested sulfur compounds and S/C selectivity of 105 . It was demonstrated that the overall performance of the FPD integrated with a silicon photodiode assembly and was comparable to a conventional detector coupled with a photomultiplier tube, ensuring simultaneously the rather shorter equilibration time together with robust to electromagnetic interference and vibration and lower cost. Clark and Thurbide (2015) have also suggested the modification of classical multiple flame photometric detector (mFPD) suitable for investigation of petroleum products. The new device was based on the interconnecting fluidic channels within stainless steel plate. Relative to the quartz tube mFPD, the stainless steel mFPD provided a 50% reduction in background emission levels, an orthogonal analytical flame, and more sensitive operation. As a result, sulfur response in the stainless steel mFPD spanned four orders of magnitude with a minimum detectable limit of 9 × 10−12 g S/s and had S/C selectivity about 104 . According to Clark et al., the stainless steel mFPD allowed analyte emission monitoring in the multiple worker flames. This mode might improve the analytical flame response of sulfur (both linear HSO*, and quadratic S2 *) and phosphorus.

121

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

Secondary retention time (s) 0.5 1

Phosphorus detection in the suggested devise provided six orders of linearity with a detection limit of about 2.0 × 10−13 g P/s. The stainless steel design of mFPD improved the analytical performance of the detector and provided a versatile tool for GC. The additional advantage of the described device design was in the fact that simultaneous elution of sulfur-containing compounds and hydrocarbons which are the case for petroleum products, didn’t result in signal suppression, while in classic FPD, coelution of target compounds and hydrocarbons might cause the detector signal blanking. Nizio and Harynuk (2012) have reported the approach using comprehensive two-dimensional GC with nitrogen phosphorus detection (GC × GC–NPD) for qualitative and quantitative profiling of alkyl phosphates in industrial petroleum samples at levels comparable to or below those achievable by ICP-OES. Alkyl phosphates are used for intensification of petroleum production and contain substituents with hydrocarbon chains from 6 to 20 carbon atoms – this is a required condition for the formation of stable microemulsions with oil. The recovery study was presented in a fracturing fluid sample and a profiling study of alkyl phosphates in four recovered fracturing fluid/crude oil mixtures (flowback). Installation of the Deans switch after the separation columns allowed using the GC system with two detectors (with FID as the second detector) and have significantly increased the overall life of the NPD bead from about 5 chromatographic runs to more than 1000 due to redirection to NPD of only the components of interest. The resulting GC × GC-FID/NPD system was capable of qualitative and quantitative analysis of individual alkyl phosphate compounds in industrial petroleum samples. The developed method also ensured detection and quantification of alkyl phosphate standards at levels comparable to those achievable by the currently accepted ICP-OES methodology. All the industrial petroleum samples which were tested by Nizio and Harynuk (2012) contained two known trialkyl phosphates (triethyl phosphate and tributyl phosphate) and numerous other compounds that were likely alkyl phosphates or nitrogen-containing compounds (Figure 2.4.3).

TPhP

TEP TBP

0

122

0

200

400 Primary retention time (s)

600

Figure 2.4.3 GC × GC-NPD chromatogram of a flowback sample, containing TEP (triethyl phosphate), TBP (tributyl phosphate), TPhP (triphenyl phosphate). Additionally, several other probable phosphorus (possibly nitrogen) compounds are also detected; however, remaining unidentified due to a lack of available alkyl phosphate standards. Source: Nizio and Harynuk (2012)/Reproduced with permission from Elsevier.

2.4.2 Chromatography Methods

2.4.2.2

Liquid Chromatography

Liquid chromatography is the less-common method for petroleum characterization. The main area for this method is characterization of molar masses of the compounds and some quick qualitative separations for primary evaluation of the samples. However, shifting of focus of petroleum products from fuels to petrochemical synthesis demands the detailed characterization of most important components, such as aromatic fraction, which will help in the conversion of aromatics to valuable petrochemical products using different processing conditions. Panda et al. (2022) have applied an aromatic-selective size exclusion chromatographic method for the characterization of aromatic compounds in high-boiling fractions (cuts), distilled from three crude oils differing in density (or API gravity) values. The fractionation of crude oils was performed with subsequent determination of number average molecular weight M n which may be correlated to the analyte colligative properties, e.g. freezing point depression and weight average molecular weight M w values which may be correlated to properties, such as melt viscosity, for each fraction. It has been found that the compositions of the distilled cuts displayed significant similarities, irrespective of the crude oil’s origin. The compositional information was then added with the data from HPLC and GC × GC separations. The suggested method allowed fast (lower than 30 minutes) separations; however, the analysis time might be further reduced (down to 10 minutes) with smaller sorbent particle size (i.e. moving to ultra-HPLC), similar to polymer analysis. Azinfar et al. (2018) described investigation of high MW hydrocarbon mixtures such as heavy oil, bitumen, and vacuum residues, using gel permeation chromatographic method to get information on the boiling point and molecular weight distributions. The GPC and simulated distillation separation results to characterize very heavy hydrocarbon samples have been combined. The LC method has been simplified excluding the utilization of standard sample which is necessary for calibration. Instead of this, Azinfar et al. (2018) suggested using each working sample individually as a standard sample for GPC calibration by matching the simulated distillation and GPC results. The reason for this was the fact that standard sample in GPC is usually the polymer compound with known MW and molar weight distribution, and such a standard is not relevant to oil samples. The GPC and simulated distillation results were coupled resulting in the calibration curve retention time versus molecular weight of the compound. The obtained calibration curve was then utilized for each sample together with the GPC results, and the whole molecular weight and boiling point distributions were obtained for all the extra heavy samples. The tandem GPC-SD characterization method was validated with the known standard sample and was applied to access the molecular weight distribution of bitumen and the bitumen fractions samples. The developed characterization method provides a tool to find a better understanding of the molecular weight and boiling point distributions of complex mixtures. The results demonstrated that the very small range of components with molar masses from 100 to 1400 Da might be characterized with simulated distillation only. Using the combined GPC-SD method was however able to extend the distribution and provide more detailed information about the molecular weight distribution of asphaltene and the extra heavy cuts up to 9000 Da for whole bitumen and 10 000 Da for asphaltenes (Figure 2.4.4). Kim et al. (2015) also suggested a combination of three analytical methods including HPLC, two-dimensional nuclear magnetic resonance spectroscopy (2D NMR), and uHRMS, for molecular characterization of crude oil, which was deasphalted using n-heptane. HPLC system was used for separation of the crude oil to five fractions, and it was demonstrated that the first one contained mostly monoaromatic compounds, the second diaromatics, the third triaromatics, the fourth tetraaromatics, and more, and the fifth fraction polar compounds. Utilization of information obtained from 2D NMR and HR MS cleared that the first to third fractions received from HPLC

123

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

Cut 4 SD

This study

1.4

100

1.2

80

80

Cut 4-deasphalted This study SD

0.8

40

0.6

0.4 20

1.0

%Off

0.6

40

1.2 60

wt.%

%Off

0.8

0.4

20

0.2

0.2

0 100

1000 MW (g/mol) Whole bitumen 100 This study SD

1.6 1.4

1.0 60

1.8

wt.%

100

0 100

0.0 10000 2.5

100

2.0

80

1000 MW (g/mol)

Asphaltene SD

This study

0.0 10000

0.7 0.6

80

40

1.0

40

20

0.5

20

%Off

1.5

wt.%

60

60

0.4 0.3

wt.%

0.5 %Off

124

0.2

0 100

1000 MW (g/mol)

0.0 10000

0 100

0.1 1000 10000 MW (g/mol)

0.0

Figure 2.4.4 The %Off versus molecular weight (the left axis) and the molecular weight distribution obtained by the combined GPC-SD method (the right axis). The open symbols are the predicted molecular weights of the compounds which were received using the GPC-SD method, the filled symbols correspond to the results of simulated distillation only. Source: Azinfar et al. (2018)/Reproduced with permission from Elsevier.

separation contained hydrocarbon compounds with an aromatic ring core and multiple saturated cyclic rings. Groups of compounds with sulfur atom and any number of carbon and hydrogen atoms were abundant in the first two fractions (Figure 2.4.5). Examining of double-bond equivalents (DBEs) distribution for the received fractions provided the further additional information. DBE corresponds to the number rings plus the number of double bonds in the molecular formula and is calculated with the simple equation: DBE = c + h∕2 + n∕2 + 1 for elemental formula Cc Hh Nn Oo Ss . Calculations for each fraction have shown that compounds with thiophene (DBE = 3) and benzothiophene (DBE = 6) functional groups were abundant in the first fraction and those with benzodithiophene (DBE = 8) in the second fraction. In contrast, a significant prevalence of nonaromatic sulfur compounds was observed in the fifth (polar) fraction. Kim et al. (2015) have demonstrated that combination of HPLC, 2D NMR, and HR MS may help to overcome the limitations of individual techniques in the investigation of complex heavy crude oil samples. Sim et al. (2015) have used HPLC combined with HR-MS for the detailed characterization of the heavy crude oil fractions prepared by reversed-phase HPLC. The possibility of the method was

2.4.3 Mass Spectrometry Methods

Summed relative abundance (%)

35

Fraction 1 Fraction 2

30

Fraction 3

25

Fraction 4

20

Fraction 5

15 10 5 0

S1

S2

S3

HC

N1

O1

O1S1 N1O1 N1S1

Chemical class Figure 2.4.5 Class distribution plots of the five fractions received in HPLC separations. Here, the HC class corresponds to hydrocarbons containing only hydrogen and carbon atoms. Other classes such as S1 correspond to compounds having a sulfur atom and an unlimited number of carbon and hydrogen atoms, and so forth. Source: Kim et al. (2015)/Reproduced with permission from Elsevier.

demonstrated on the HPLC fractionation of heavy crude oil and an oil shale pyrolysate, derived from a sample of outcrops rock. For the HPLC separation, the cyanopropyl standard column with 5 μm particles with 100 Å pore size and CN guard column were used resulting in four fractions which were then investigated separately using high resolution mass spectrometry. The analyses of the individual fractions showed that the elution order of the analytes in HPLC system was determined mostly with the carbon number of alkyl chains and the DBE value of the compounds of sample. The compounds with larger values of DBE, which are considered to be the more condensed aromatic structures, and smaller carbon number, assumed to have short side chains in the molecule structure, were eluted earlier compared to the analytes which were characterized with lower DBE values, which corresponds to the less-aromatic structures, and higher carbon number (presumably compounds with longer alkyl chains). This elution order is classical to the reversed-phase HPLC separations, so reversed-phase HPLC might be considered as the effective tool for separation of crude oil compounds.

2.4.3

Mass Spectrometry Methods

It is well known that the main elements of mass spectrometry, which predetermine the diversity of its structural and analytical possibilities, are based on the use of different ionization principles and methods of ion analysis. Various combinations of corresponding instrumental devices can be used in the study of organic mixtures by hyphenated versions of mass spectrometry with chromatographic methods and desorption/ionization mass spectrometry, in particular, for the determination of qualitative and quantitative composition of crude oils, heavy oil fractions, and of products manufactured on their basis. In the following, we will briefly describe the most common and especially effective mass spectral devices that are used for the comprehensive study of heavy oils and their products.

125

126

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

2.4.3.1

Ionization Methods

A large number of different ionization methods have been developed for mass spectrometers, which are currently used to analyze whole heavy oils, their fractions, and refined products. Among them, one should single out rather early, but still-used methods, as well as relatively new techniques that ensure an effective comprehensive study of oils. It should be mentioned that each of the methods considered below has its own advantages and disadvantages, and unfortunately, none of them allows one to describe in detail the whole petroleum, consisting of various classes of compounds differing in molecular weights, elemental composition, physical chemical properties, reactivity, and ionization capacity. 2.4.3.1.1 Electron Ionization (EI)

Electron ionization (EI) is the earliest and till now popular ionization technique. It operates on vapors and is based on the interaction of analyte molecules by an electron beam. Usually, employed electron energy 70 eV is significantly higher than the ionization energy of organic analyte molecules that provide the formation of molecular radical ions. The latter ions undergo decay to form fragment ions which form a mass spectrum. An important advantage of the method is that it is universal and is capable of ionizing molecules of any elemental composition. Another advantage of the method is the high inter-instrument reproducibility of the spectra which make it possible to use databases of standard EI mass spectra for identification purposes. Being used in combination with chromatography, special modes of mass spectrometric operation allow the observation of target compounds with low detection limits even within chromatography unresolved mixtures. One of the disadvantages of EI is that it produces sometimes extremely unstable M+⋅ ions which make the structural identification difficult. To improve the situation, EI with vibrationally cold molecules (supersonic molecular beam mass spectrometry, SMB-MS) may be employed. It is based on the insertion of an analyte and carrier gas through a small nozzle into the ion source in the form of a SMB (Amirav et al. 2008). At supersonic flow velocity, the vibrational and rotational energies of analyte molecules decrease, the molecular radical cations produced have a much lower degradation rate and are recorded in the mass spectrum as more intense M+⋅ ion peaks along with a characteristic set of fragment ion peaks. The efficiency of such experimental approach has been demonstrated by qualitative and quantitative determination of n-alkanes in jet fuels (Islam et al. 2017). 2.4.3.1.2 Chemical Ionization (CI)

Chemical ionization (CI) provides another opportunity to increase the stability of a molecular ion. The method is usually referred to as “soft” or low-energy ionization. In this case, the ion source operates at an elevated pressure of a reactant gas (methane, isobutane, ammonia) at ∼1 torr. Under EI, its molecules yield primary ions that interact with neutral reactant molecules to give rise to a variety of secondary ions. Subsequent interaction of the latter ions with simultaneously introduced analyte molecules is accompanied by protonation, deprotonation, or cluster formation. The analyte ions thus formed decompose only slightly and the ion peaks in molecular region become more intense. It should be noted that modern mass spectrometer can operate with fast-switching EI/CI. 2.4.3.1.3 Atmospheric-Pressure Chemical Ionization (APCI)

Atmospheric-pressure chemical ionization (APCI) is another and rather old “soft” ionization method in MS. It is also based on the interaction of analyte molecules with ionized reagent gas molecules that proceeds in the ion source at atmospheric pressure. Further reaction of reagent ions with analyte molecules gives rise to protonation and deprotonation products as well as to the formation of clusters and molecular radical ions. Solvent ions, ionized water clusters, and

2.4.3 Mass Spectrometry Methods 4 000 000

Abundance

100% toluene

10 000 000

Abundance

127

3 000 000 2 000 000 1 000 000

5 000 000

0

301.015 301.056 301.097 301.138 301.179 301.22 301.261 301.302 301.343 301.384

m/z

300

400

500 m/z

600

800

(a)

Toluene/MeOH 50:50

4 000 000

10 000 000

3 000 000 Abundance

Abundance

700

2 000 000 1 000 000

5 000 000 0

300

400

500 m/z

600

301.056

700

301.138

301.22 m/z

301.302

301.384

800

(b) Figure 2.4.6 Effect of solvent compositions on APCI MS sensitivity for petroleum. Mass spectrum (a) registered using toluene as a solvent, whereas mass spectrum (b) acquired for a solution in toluene/ methanol mixture. Source: Kim et al. (2016)/Reproduced with permission from American Chemical Society.

dopant ions, formed in a corona discharge, are the main ionized participants in such reactions. During APCI, most petroleum compounds form positive ions among which the M+⋅ and [M + H]+ ions are particularly important. The ionization efficiencies of the analytes can be increased using charge-transfer agents such as toluene (Figure 2.4.6). This ionization technique is mostly suitable for combinations of MS with HPLC. Alternatively, however, the samples may be introduced into the ion source by direct injection. 2.4.3.1.4 Field Ionization (FI)

Field ionization (FI) is based on the detachment of an electron from a molecule via the quantum-mechanical tunneling effect in a strong electric field. Primary molecular radical cations thus formed have a relatively low internal energy and do not decompose further which provides determining the molecular weights. It should be noted that under FI conditions, saturated and aromatic hydrocarbons have similar ionization efficiencies which make the method convenient for quantitative analysis. Hyphenated FI-MS and chromatography method may be used for studying petroleum compounds.

128

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

2.4.3.1.5 Photoionization (PI), Atmospheric Pressure Photoionization (APPI), Atmospheric Pressure Photochemical Ionization (APPCI)

Rather old photoionization (PI) method is based on the interaction of analyte molecules in the gas phase with photons resulting in the formation of molecular radical cations. At low-photon energies (8–12 eV), corresponding peaks are practically the only ones in the spectra. However, with raising the energy, the spectrum can be increasingly populated by peaks of fragment ions. Note that PI using high-energy photons from UV laser and excimer lamp light sources was designated as single-photon ionization (SPI) or vacuum photon ionization (VPI). These versions are helpful for the detection of hydrocarbons of various structures with low-detection limits. Atmospheric pressure photoionization (APPI) and atmospheric pressure photochemical ionization (APPCI) techniques were developed by Revelskii et al. (1986). Both methods involve the interaction of organic molecules with photons at atmospheric pressure. In both the cases, analyte M+⋅ ions along with ionic products of secondary ion–molecule reactions are formed. In the case of APPCI version, the additional introduction of dopants into the ionization region is used which gives rise to increasing the probability of secondary ions formation. In contrast to APCI, APPCI technique makes it possible to impart a charge to nonpolar and weakly polar compounds giving rise to molecular cation radicals, protonated, deprotonated molecules, and cluster ions. In addition, the introduction of selected dopant molecules can substantially alter the probability of occurrence of certain ionization processes. All the named PI techniques can be used for the analysis by hyphenated chromatography-mass spectrometry methods. 2.4.3.1.6 Electrospray Ionization (ESI)

Electrospray ionization (ESI) belongs to a group of ionization methods that employ the ionization at atmospheric pressure occurring in an ions source. The method is based on the principle developed by Dole et al. (1968) saying that if a solution of a compound in a volatile solvent passes through a thin capillary, then charged drops are formed at the capillary end under the action of a strong positive or negative electric field. Upon the evaporation of solvent molecules, necessary charged ions of the analyte molecules are released in a gaseous state and enter the mass spectrometer. Main ionization mechanisms of ESI involve primary formation of closed-shell ions owing to protonation (cationization) or deprotonation of analyte molecules. In this case, the energy supplied to the analyte molecules is low, and primary ions do not undergo fragmentation which is convenient for determining their molecular weights. In order to gain structural information from ESI mass spectra, simulated decomposition of primary or fragment ions is utilized. ESI method is mainly applicable to the detection of polar oil compounds. A further strong increase in the popularity of ESI is associated with the work of the Russian research group headed by Gall’ (Aleksandrov et al. 1984; Yamashita and Fenn 1984). At present, the method has received the widest distribution in the analysis of a wide variety of organic compounds, both by direct insertion of a sample into an ion source, and especially when using a HPLC-MS combination. Note that ESI-MS was first applied to the analysis of nonseparated petroleum compounds by Zhan and Fenn (2000) who showed that ESI mass spectra of deasphalted oils are a complex superposition of ion peaks of a large number of polar compounds. A special kind of ionization mass spectrometric method is based on the principles of desorption/ionization and is especially effective in the study of heavy and rather high molecular and nonvolatile polar petroleum components. 2.4.3.1.7 Atmospheric Solid Analysis Probe (ASAP) Mass Spectrometry

Atmospheric solid analysis probe-mass spectrometry (ASAP-MS) is another version of open-air ionization mass spectrometry. It is based on the ionization of analyte molecules by their interaction

Corona discharge

9000 nA

Hea ted g as

gas ted Hea

≈ 1mg of crude oil, saturated hydrocarbon fraction and five condensed aromatic standards

Glass capillary 150–650 °C

2.4.3 Mass Spectrometry Methods

M M M [M+H]+ M e– M– M [M+H]+ M●+ + e (H O) M● 2 n M●+ [M+H]+ (H2O)n

N2 (150 °C)

Synapt G2-S HDMS

M = HCs, PAH, polar species Figure 2.4.7 ASAP probe scheme and the ionization mechanisms via proton ([M + H]+ ) and electron (M+⋅ ) transfers. Source: Tose et al. (2017)/Reproduced with permission from American Chemical Society.

with the primary corona-generated ions from atmospheric nitrogen and water vapors (McEwen et al. 2005). The sample solution can be introduced into the ion source in a tube, the end of which is blown with a stream of nitrogen at 50–500 ∘ C or through LC. The method practically does not require the use of sample preparation procedures and relates to ambient mass spectrometry. ASAP-MS can be readily applied to the analysis of petroleum hydrocarbons that are ionized in this mode, giving rise to protonated molecules and [M + N]+ ions. By varying the gas temperature, the fractionation of the samples can be accomplished, thereby facilitating the interpretation of the obtained spectral data. In particular, during the analysis of crude oils, this approach made it possible to separate the components according to their boiling points and to detect condensed aromatics standards and paraffinic fraction (Figure 2.4.7). 2.4.3.1.8 Laser Desorption/Ionization (LDI), Matrix-Assisted Laser Desorption/Ionization (MALDI), Surface-Activated Laser Desorption/Ionization (SALDI)

These desorption/ionization methods are based on the ion formation during pulsed laser irradiation of solid samples. Being the pulsed techniques, all of them are optimally combined with time-of-flight (TOF) mass analyzer and Fourier-transform ion cyclotron resonance (FT-ICR) analyzer, and it is with these devices that most efficient commercial instruments are equipped. Among them, matrix-assisted laser desorption/ionization (MALDI) gained the most popularity in the selective detection of certain classes of compounds. It is based on the discovered enhancement of laser desorption/ionization (LDI) efficiency of an analyte in the presence of another compound (matrix) (Karas et al. 1985). Since the efficiency of desorption/ionization of analytes of various structures depends on the nature of organic matrix used, it is possible to select those of them that achieve a better yield of ions for one class of analytes and their suppression for the other one. Surface-activated laser desorption/ionization (SALDI) does not require the addition of any organic matrix to a sample. However, it uses the phenomenon of desorption/ionization enhancement by a surface with particular chemical and physical features. LDI techniques utilize the ion formation during laser irradiation of analytes without addition of any other compounds. It is acceptable for the analysis of nonvolatile polar compounds and polyaromatic hydrocarbons. An additional version of LDI, so-called L2 MS, has to be noted. This variety

129

130

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

involves independent operation of two lasers: primary use of IR laser ensures the desorption of analyte molecules from the target surface, whereas the UV laser ensures their ionization in the gas phase. Another variant of two-stage desorption/ionization method employs the laser energy transfer by thermal and acoustic waves (laser-induced acoustic desorption, LIAD) (Borton et al. 2013). In this case, an analyte is deposited on a metal foil, and the laser beam is directed to its back side. Further ionization of the desorbed molecules can be achieved by various techniques. 2.4.3.1.9 Direct Analysis in Real Time (DART)

Direct analysis in real time (DART) is the first technique related to the type of ambient mass spectrometry that usually requires no special sample preparation (Cody et al. 2005). The method is based on the Penning ionization effect, which involves the ion formation during the interaction of air components (water vapor, ammonia, or nitrogen, and oxygen) with excited metastable atoms. The generated ions participate in further converting the analyte molecules into charged species. One of the advantages of DART-MS is that it usually requires no special sample preparation. DART ionization allows detecting a number of hydrocarbon classes including paraffins, which are hardly ionized by other soft ionization methods. In DART source, paraffins produce ion corresponding to products of their oxygenation (Figure 2.4.8). 2.4.3.1.10 Other Prospective Desorption Ionization Techniques

In addition to DART, other versions of ambient MS including the desorption ionization in open air have been tested in the analysis of oil. For example, desorption electrospray ionization (DESI) (Wu et al. 2010) and nanospray DESI (Eckert et al. 2012) can be used for recording the mass spectra of petroleum constituents. In both cases, the flow of atomized solvent microdroplets is directed to the C27H53O (2)

C26H51O2 (2)

393.40774

395.38682

C26H49O2 (3)

393.37131

Paraffin A

C25H47O3 (3) C H O (1) 27 51

C25H45O3 (4)

C24H43O4 (4)

393.33495

395.35039

395.42304

395.31382

C27H53O (2) 393.40731

C26H51O2 (2) 395.38647

C26H49O2 (3)

Paraffin B

393.37098

C25H47O3 (3) C H O (1) 27 51 395.35005 395.42275 C24H43O4 (4) 395.31357

C25H45O3 (4)

393.33438

393.40728

Paraffin C

395.38645

C26H49O2 (3) 393.37100

C25H45O3 (4)

393.33450

200

C26H51O2 (2)

C27H53O (2)

300

400

500

600

m/z

C H47O3 (3) 394.41048 25 395.35006 394.37403

700

800

900

C27H51O (1) 395.42263

1000

Figure 2.4.8 DART(+)-Orbitrap mass spectra for the paraffin samples. The right-side inserts show paraffin detection in oxygenated form (Ox classes). Source: Romão et al. (2015)/Reproduced with permission from American Chemical Society.

2.4.3 Mass Spectrometry Methods

surface of the object under study that provides the desorption and ionization of analyte. Aromatic compounds can also be detected with the use of the so-called “desorption APCI,” where a nitrogen stream passing through a corona discharge area and containing ionized particles is directed to the analyte surface which carries out the desorption/ionization process (Jjunju et al. 2013).

2.4.3.2

High and Ultrahigh-Resolution Mass Spectrometry (uHTMS)

In oil analysis, it is usually necessary to determine an accurate mass, and hence, elemental composition, for all the compounds present in a sample. This means that a mass spectrometer should include such an ion mass analyzer that provides a high-resolution throughout the whole recorded mass range. Usually, the mass resolving power (R) is equal to M/ΔM(50%), where ΔM(50%) is full width at half maximum (FWHM) of particular peak for the ion having mass M. Mass accuracy indicates the accuracy of the m/z provided by the mass analyzer which is indicated by the difference between the theoretical m/z (mtheor) and the measured m/z (mmeasured) and expressed in millimass units (mmus) or in parts per million (ppm). The highest resolving power, and hence, the highest analytical value is provided by uHRMS which makes it possible to resolve ions having the same nominal, but different exact mass, allowing one to determine the elemental composition of the ion. Currently, several mass analyzers provide high resolution. 2.4.3.2.1 Fourier-Transform Ion Cyclotron Mass Spectrometry (FT-ICR-MS)

The FT-ICR analyzers are the most efficient technique for uHRMS. They are based on the determination of the m/z ratio according to the cyclotron frequency for an ion in a fixed magnetic field (Cho et al. 2015). Corresponding instruments allow getting today the highest mass resolution (more than 107 ) and have a potential for further increasing the resolving power on the basis of novel open dynamically harmonized cells (Figure 2.4.9). This is of particular importance for the determination of elemental composition of unresolved petroleum molecules having rather high molecular weights. 2.4.3.2.2 Fourier Transform Orbitrap Mass Spectrometry

Orbitrap mass analyzer developed by Makarov (2000) is an orbital ion trap. It includes inner and outer electrodes and the ions, generated with the aid of one or another ionization method, are injected into the space between them. Under action of static electric field, the ions orbit around the inner electrode and oscillate along the z-axis. Using Fourier transformation, the m/z values of ions Plug

Imbedded cell

Flange Atmosphere vacuum

Isolating ring

Isolating ring

Pump

Figure 2.4.9 Schematic representation of the vacuum tube with an incorporated open dynamically harmonized cell. Source: Nikolaev and Lioznov (2020)/Reproduced with permission from American Chemical Society.

131

132

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

are determined by measuring the frequency of their oscillation. FT orbitrap mass spectrometers (FT-Orbitrap) provide ultrahigh resolution that exceeds 200 000 and even 400 000. Owing to the lower cost and the absence of a superconducting magnet, which is used in FT-ICR mass spectrometers, they have become the most common instruments in analytical practice, in particular, with the study of petroleum (Pomerantz et al. 2011). 2.4.3.2.3 Multireflection Time-of-Flight (TOF) Mass Analyzer

Ultrahigh-resolution multireflection TOF mass analyzers are based on an innovative multireflecting “zig-zag” design (Verentchikov et al. 2005) or on a spiral ion trajectory technique. The main purpose of using such devices is to increase the ion path length between the ion source (usually MALDI) and the detector, which ensures high resolution of TOF analyzers. Corresponding mass spectrometers (HRT) are characterized by high speed, simplicity, and nondiscriminative mass analysis. Another important advantage of HRT is that its resolution power increases with m/z, which means that they make it possible to determine the exact mass and elemental composition of rather heavy ions. Now, a mass resolution higher than 55 000 can be achieved using such a mass analyzer, and this type of TOF is mostly used in conjunction with MALDI technique. 2.4.3.2.4 Presentation of High-Resolution Mass Spectrometry Data

High-resolution MS and, particularly, ultrahigh-resolution FT-ICR-MS, applying to the study of crude oil at a molecular level, allows recording mass spectra containing hundreds of thousands of ion peaks with exact m/z values for each oil sample. Because such mass spectra are difficult for both visualization and interpretation of a huge array of data, special processing methods were suggested. Among them the methods producing the Kendrick mass scale and the Kendrick mass defect (KMD) plots (Kendrick 1963 and van Krevelen diagrams Islam et al. 2012) have been particularly efficient.

2.4.4 Particular Application of Combined Chromatography-Mass Spectrometry Methods to Analysis of Heavy Oils, Their Fractions, and Petroleum Products 2.4.4.1 Capabilities of Gas Chromatography-Mass Spectrometry Techniques in the Analysis of Heavy Oils Combination of GC and MS has become the most important tool in the study of complex mixtures of organic compounds, including oils. For heavy and extra-heavy oils, this hyphenated method can only be used to analyze their sufficiently volatile liquid fractions, which are naturally present in them, although in a smaller amount than in more light oils. Naturally, it can be efficiently applied to analysis of quite volatile petroleum products as well. The main purpose of the analysis is the detection and identification of alkanes, cycloalkanes, aromatic, and heteroaromatic compounds. Combination of one-dimensional GC and MS (GC-MS) is the oldest and still widely used in the analysis of oils. However, in recent years, conjugated techniques based on multidimensional GC and MS (MDGC-MS), mainly comprehensive two-dimensional GC (GC × GC-MS) have become more widespread. 2.4.4.1.1 Combined One-Dimensional GC and MS Method (GC-MS)

In the majority of such analyses, MS is used in EI mode (GC-EIMS) since the latter method allows obtaining a large amount of information from the spectra and, in particular, providing the structural identification of each chromatographically resolved compound by using standard databases. In

2.4.4 Particular Application of Combined Chromatography-Mass Spectrometry Methods to Analysis of Heavy Oils

Intensity 7000000 6000000 5000000 4000000

Ortho-xylene C8 C9

C11

S2 pyrolysis GC traces KS3 C14

C19

3000000 2000000 1000000 4500000 4000000 3500000 3000000 2500000 2000000 1500000 1000000

C26

C33 C40

S Ortho-xylene C8 C11 C9

C14

M33-1 C19

C26

C33

500000 S 1e+07 9000000 Ortho-xylene 8000000 7000000 6000000 C8 C9 C11 C14 5000000 M20-3 4000000 C19 C26 3000000 C33 2000000 S 1000000 30.00 40.00 50.00 60.00 70.00 10.00 20.00 Retention time (min)

C40

C40 80.00

90.00

Figure 2.4.10 Py–GC pyrograms of asphaltenes. Source: Makeen et al. (2015)/Reproduced with permission from Elsevier.

fact, despite the creation of new ionization and compound separation methods, such combines are still widely used in routine petrochemical and geochemical studies. As is clear, the objects of analysis by GC-MS are traditionally relatively volatile components (various hydrocarbons, aromatic, and heteroorganic compounds), the range of which increases with the involvement of high-temperature GC-MS, which is especially important in the study of heavy oils. The problems with volatility of analytes can be also solved using pyrolysis. Thus, for example, Py-GC/MS was used for characterization of asphaltenes (Figure 2.4.10). As is known, before the creation of the GC-MS method, MS was used to analyze only unresolved mixtures of petroleum compounds. This approach was mainly directed to structural group analysis of the main oil hydrocarbon classes on the basis of mathematical processing of the total mass spectrum. Recently, the structural group analysis was suggested to be performed using GC-MS which permitted the contents of the main hydrocarbon classes in super complex oil mixtures (heavy distillates and residua) to be determined (Brodskii et al. 2014). It should be noted that one can meet a huge array of publications concerning the use of GC-EIMS in petroleum chemistry. Here we will mention only the most interesting results obtained in recent years. One of the studies dealt with a comparison of oils from traditional reservoirs of various

133

134

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

geneses in the South China (Zhu et al. 2018). The same method was used for the detection and identification of saturated and aromatic compounds produced during a simulated biodegradation conversion of heavy components (resins) in heavy oil in the presence of biosurfactants (Wang et al. 2018). Of particular importance is the detection and relative quantification of so-called “biomarkers” (terpanes, steranes, and their aromatic analogues) which are widely used in establishing some geochemical parameters, such as the origin of oil, thermal maturity, metamorphism, biodegradation of oils, typing oil, and transformation during spills of oils (Peters et al. 2005). In addition, detection of typical biomarkers (in particular, triterpanes) and their ratios can be used for solving specific environmental problems (Yang et al. 2017). It should be noted that work continues to search for new types of oil biomarkers. In particular, rather recently a new set of hexacyclic benzohopanes C33 –C35 was identified in oil with the aid GC-EIMS (Nytoft et al. 2016). It is interesting to pay attention to the discovery of biomarker-like compounds produced by thermolysis of the insoluble prokaryote biomass of various origin (Gordadze et al. 2018). High-temperature GC-EIMS was used for detecting and identifying metalloporphyrins (Figure 2.4.11). Comparatively low-molecular weight S-containing compounds that are present in volatile fractions of heavy oil can be effectively analyzed by GC-EIMS (Lobodin et al. 2015a). For more efficient detection and identification of individual compounds, belonging to reactive (mercaptanes, thiacyclanes, aliphatic sulfides, and disulfides) and nonreactive (polycyclic thiophenes) series, special preliminary separation of such fractions was shown to be suitable. The use of CI in GC-MS (GC-CIMS) appeared to be also helpful in the determination of molecular weights of hydrocarbons that produce unstable molecular ions under EI conditions. Of a considerable interest is the application of GC-MS system including the combined ion source with fast switching between EI/CI modes. For example, such a technique in high-temperature mode was used for the identification of a group of wax hydrocarbons of unknown genesis (Huang et al. 2003). Owing to recording of the [M + H]+ ions in CIMS mass spectra, these compounds were unambiguously assigned to the structure of branched alkanes. In other cases, combination of GC with CIMS was successfully used for the selective detection of particular compounds in petroleum. For example, sulfur- and nitrogen-containing heterocyclic compounds in crude oils and their refining products were determined by using ammonia gas reagent that provides efficient and selective protonation of such heteroorganic compound within a hydrocarbon mixture (Creaser et al. 1993). Application of MS on the basis of APCI in such a combination (GC-APCIMS) allowed the enhancement of sensitivity to be made in the detection of biomarker compounds in oils at the trace level. Additional use of MRM mode reduced the limits of detection of the corresponding components (Wu et al. 2015). All this is particularly important for the determination of bio-markers in oil spills containing lower concentration of targeted compounds, as compared to initial oils, owing to multifold dilution and biodegradation. The use of the FI in GC-MS technique (GC-FIMS) has also considerable prospects. In particular, this method was used for the analysis of various diesel hydrocarbons and for qualitative and quantitative determination of alkanes and olefins in coal tar (Ni et al. 2013). Hyphenated method GC and MS on the basis of PI is also a fairly popular technique for studying hydrocarbon mixtures, in particular, polycyclic aromatic hydrocarbons (Revelsky et al. 2015). The main advantage of PI and its SPI and VPI versions is that they predominantly allow recording the M+⋅ ion peaks. These methods were shown to be effective in the observation as well as in separation and identification of isomeric hydrocarbons (Isaacman et al. 2012). Note also that the GC-SPIMS method in combination with thermogravimetry made it possible to identify products formed during thermolysis of crude oils of various geographical origins (Wohlfahrt et al. 2013).

2.4.4 Particular Application of Combined Chromatography-Mass Spectrometry Methods to Analysis of Heavy Oils

Abundance

4.0E+07

2.0E+07

0.0E+00 18

19

20

21 22 Retention time (min) (a)

23

24

Abundance

1.8E+06

1.2E+06

6.0E+05 M - (CH2)n/CH3)n 2

M - (CH2)n/CH3)n

0.0E+00 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 m/z (b) 2.0E+06

Abundance

1.5E+06

1.0E+06

5.0E+05

M - phen 2 M - 2 phen 2

M - phen M -2 phen

0.0E+00 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 m/z (c) Figure 2.4.11 (a) High temperature GC-ToF-MS total ion chromatogram of porphyrin mixture, (b) background subtracted mass spectrum of 2,3,7,8,12,13,17,18-octaethyl-21H,23H-porphine zinc(II) (18.852 minutes in (a)) and (c) background subtracted mass spectrum at 430 ∘ C of 5,10,15,20-tetraphenyl21H,23H-porphine copper(II) (23.228 minutes in (a)). Source: Sutton and Rowland (2012)/Reproduced with permission from Elsevier.

135

136

2.4 Analysis of Heavy Crude Oil and Its Refined Products by Various Chromatographic and Mass Spectrometry Methods

Although ultrahigh resolution MS, such as Orbitrap and FT-ICR-based MS, is very effective in compound identification, its combination with GC is less commonly used. However, in the literature one can find some applications of such hyphenated technique to analyze volatile oil compounds of petroleum. For example, a combined EI/APPI ion source in such a system permitted a wider range of hydrocarbons in gas condensate to be identified (Kondyli and Schrader 2019). In the other case, polyaromatic hydrocarbons in various fossil oil and petroleum crude oil were characterized by a combination of GC with atmospheric-pressure laser ionization FT-ICR-MS (Benigni et al. 2016). 2.4.4.1.2 Combined Two-Dimensional GC and MS (GC × GC-MS)

As it was already mentioned in Section 2.4.2, the most important advantage of the GC × GC-MS method is the ability to separate, detect, and identify components that coelute in a one-dimensional GC-MS. Naturally it gained a great success in the investigation of complex mixtures of hydrocarbons. In particular, it is effective in the differential determination of isomeric petroleum compounds that can co-elute in one-dimensional GC-MS. Analysis of literature data indicates an increasing involvement of GC × GC-MS in the study of various aspects of petroleum chemistry and oil refining (Pollo et al. 2018). The most routine EI sources are most popular in mass spectrometers included in such techniques. This is especially important, since available databases of standard EI mass spectra can be used for automatic identification of substances. Although the GC × GC-EIMS method allows obtaining sufficiently reproducible and reliable results, in most cases, the dependence of the effective ionization cross section on the structure of the compound complicates quantitative determinations without using the appropriate internal standards. To ensure the necessity for providing the high speed of recording mass spectra, TOF mass spectrometers are frequently used. Attention should be paid to the possibility of realizing a high-temperature version of the method which allowed the separation and identification of hydrocarbons with the carbon numbers up to 60. This principle was the basis of the new approach to two-dimensional simulated distillation of crude oils (Figure 2.4.12). Using GC × GC-TOFMS, many classes of aliphatic, alicyclic, aromatic, sulfur compounds were identified in Brazilian oils (Vanini et al. 2018). In other studies, this technique was efficiently applied to a detailed description of individual classes of sulfur compound series (Han et al. 2018) and nitrogen heterocycles in heavy gas oil petroleum fractions (Von Mühlen et al. 2010). In addition to conventional oils, GC × GC-TOFMS was also used in the analysis of bio-oils (biomass pyrolysis products) (Tessarolo et al. 2014) and heavy hydrocarbon shale (Li et al. 2015). Several works have been devoted to the separation and identification of various conventional and unconventional biomarker compounds in oil by this method. For example, substituted tetracyclic (Araújo et al. 2018), pentacyclic, and other triterpenoid hydrocarbons (Santos et al. 2018; Silva et al. 2011) were identified and quantified. High chromatographic resolution of GC × GC-TOFMS made possible the detection and characterization of such specific compounds such as diamondoids (Silva et al. 2013) and heteroatomic adamantanes (Zhu et al. 2020) which are present in oil in minor amounts. It should certainly be noted that application of selective and sensitive GC × GC-TOFMS in EI mode makes a significant contribution to understanding of oil degradation processes of various origins, in particular, bio-degradation as well. In order to increase the abundance of molecular ions, the application of FI mass spectrometers in such combines is also of interest. For example, the hydrocarbon population in the middle distillate fraction of crude oil was studied with the use of GC × GC-FIMS (Genuit and Chaabani 2017) and, owing to the presence of intense M+⋅ peaks, unambiguous differentiation of various hydrocarbon structures even within overlapping elution zones was made. The same approach was successfully used for detailed characterization of heavy petroleum distillates (Figure 2.4.13).

2.4.4 Particular Application of Combined Chromatography-Mass Spectrometry Methods to Analysis of Heavy Oils

1

2nd Dimension Retention Time [s] 2

3

Massws: TIC

9 – 61 610 09 –6 600 99 –5 590 89 –5 580 79 –5 570 69 –5 560 59 –5 550 49 –5 540 39 –5 530 29 –5 520 19 –5 510 09 –5 500 99 –4 490 89 –4 480 79 –4 470 69 –4 460 59 –4 450 49 –4 440 39 –4 430 29 –4 420 19 –4 410 09 –4 400 399 – 390 89 –3 380 379 – 370 69 –3 360 359 – 350 49 –3 340 39 –3 330 329 – 320 19 –3 310 09 –3 300 99 –2 290 89 –2 280 279 – 270 69 –2 260 259 – 250 49 –2 240 39 –2 230 229 – 220 219 – 210 9 – 20 200 99 –1 1 9 0 18 9 – 180 79 –1 170 69 –1

150

160

59

0

0

Fe2 O3 /clay > Fe2 O3 /sandstone, while clay-based iron oxide catalysts turned out to be the most selective. In another study, Chen et al. (2011) have taken reservoir minerals from Daqing and Liaohe oilfields oil sand. The experiments were carried out with the addition of the same oil. Reservoir minerals included quartz, potash feldspar, plagioclase, some calcite, and dolomite (about 1%). The total content of clay minerals was about 10%. In contrast to simple aquathermolysis, the addition of formation minerals resulted in a decrease in the average molecular weight of higher saturated and aromatic content, lower resins, and asphaltenes content. The coefficient of viscosity of the two oil samples has been found to increase from 7.41% and 12.95% up to 16.05% and 25.29%, respectively. With the addition of NiSO4 , the average molecular weight of the reaction products and the contents of resin and asphalt formation are further reduced, saturated, and aromatic; with a further increase in the content of hydrocarbons, the viscosity is significantly reduced, with a recovery factor of up to 84.39%. In the work performed by Tavakkoli Osgouei and Parlaktuna (2018), the effect of carbonate and sandstone rock, temperature, and clay and nonclay minerals was evaluated during the steam distillation of Chamurlu and Bati Raman heavy oil samples. The density and viscosity of the remaining oil were measured, and SARA analysis was performed. It has been found that kaolinite particles allow the volatile components to evaporate, creating an inert surface, which can be considered as a kind of catalytic effect. As for other minerals, especially bentonite, the behavior of clay particles seems to be different. The swelling of bentonite in water due to special mineralogical composition prevents the evaporation of volatile components of oil. Consequently, the density of the remaining oil decreases. The behavior of sepiolite was similar to that of kaolinite as the sepiolite sample contains kaolinite as a clay mineral. The content of asphaltenes in the remaining oil after steam distillation was found to be higher than that of the original crude oil. It increases somewhat in the presence of minerals. On the other hand, the addition of clay minerals leads to a significant increase in the content of asphaltenes due to a decrease in the amount of volatile components. Zhang et al. (2010) investigated the effect of minerals together with nickel sulfate used as a catalyst and added it to the reaction system. The viscosity reduction rate was found to reach 50%. It was shown that

Ph

Ph

Initial extract

Initial extract Pr

Pr

19

19

Ph

Ph

Pr

Pr 32

200 °C

14

17

18

22

24

26

28

32

12

13

15

200 °C 30

24

22

17 18

35

Intensity

35

Intensity

14

15

30

28

26

19 20 19 20

Ph

30 15

16

Pr

20 17 18 19

14

21

22

24

26

28

300 °C

28

30

27

32 13

Ph 19

Pr

5

10

15

20

n-alkanes

25

30 35 40 45 50 55 60 65 70 Retention time (min) Regular isoprenoids Pr pristane Ph phitane

300 °C

25

35

75

5

10

15

20

25

n-alkanes

21

23

30 35 40 45 50 55 60 65 70 Retention time (min) Regular isoprenoids Pr pristane Ph phitane

75

Figure 4.17 GC–MS chromatograms (TIC = total ion chromatography) of saturated hydrocarbon fraction of initial oil and after thermal treatment in the presence of calcite (CaCO3 ) – left and in the presence of calcite (CaCO3 ) – right.

196

4 Fundamentals of In Situ Upgrading

if formic acid is used as a hydrogen donor, the rate of viscosity reduction can increase even more to a total amount of up to 71.8%. The reaction of heavy oil aquathermolysis under steam injection conditions was affected by the reaction temperature, reaction time, amount of added minerals, catalyst density, and hydrogen donor. Moreover, it was reported that minerals, a catalyst, and a hydrogen donor can jointly enhance the aquathermolysis reaction of heavy oil in the presence of high-temperature steam. In fact, the mineral composition included quartz, kaolinite, montmorillonite, illite, and others. The applied value of the study lay in the possibility of regulating the process of chemical transformation of oil. In the process, minerals can electrify metal ions – nickel, vanadium, iron, and molybdenum. This process speeds up the breaking of the C—S bond. In this case, acid cents are formed, accelerating the reactions of aquathermolysis. For example when exposed to vapor, Al3+ in the composition of clay can adhere to the surface of H4 SiO4 . It is formed in the process of exposure to steam. In this case, highly active hydroxyl groups are formed. The H+ ion from water can combine with oxygen bound to Si4+ . This results in the formation of a highly active hydroxyl group, where hydroxyl groups can release the H+ ion by analogy with Brønsted acids; meanwhile, visbreaking and hydrogen transfer occur in the presence of Brønsted acids. Similar works have been provided by Hyne et al. (1982a), who studied the catalytic effect of minerals on the hydrocarbon groups, viscosity, and element distribution of heavy oil during thermally enhanced oil recovery. It has been found that minerals have a catalytic effect in heavy oil aquathermolysis. It was shown that in the presence of 10 wt.% mineral, the saturate and aromatic content increased, while the resins and asphaltenes components decreased. Moreover, the obtained data found that the average molecular weight of heavy oil and asphaltene decreased after thermal treatment in addition to sulfur content. This study has reported a decrease in heavy oil viscosity by 23.4–25.6% in the reaction system after aquathermolysis. In fact, the term “aquathermolysis,” as described, is the process of chemical interaction of high-temperature, high-pressure water, and heavy oil and bitumen reactive components, which differ from “hydrothermolysis,” which has come to be associated with the interaction with hydrogen at elevated temperature and pressure. Model compounds studying showed that mineral components, in particular, trace metals such as vanadium and nickel, are able to accelerate heavy oil organ sulfur components cracking, which leads to producing carbon monoxide and other gases. What is more, aquathermolysis may lead to an increase in saturates and aromatics content and a decrease in resins and asphaltenes content. This may lower the average molecular weight and viscosity of oil and improve its properties. This has attracted the interest of many researchers to study how to apply reservoir minerals to accelerate the aquathermolysis of heavy oil. Belgrave et al. (1997) studied heavy oil aquathermolysis kinetics with different model compounds and found that the mineralogy plays an important role in the generation of CO2 and H2 S. In a similar work, Clark et al. investigated the steam–oil chemical reaction and found that some metal ions, minerals, and reservoir sands can change the heavy oil composition drastically (Clark and Hyne 1984). Monin and Audibert (1987) studied heavy oil/mineral matrix systems’ thermal cracking at temperatures and pressures encountered during thermal recovery. They found that the chemical reactions, which include oil, possibly water, and a mineral matrix, may change the heavy oil composition significantly. Four thermal crude oils with different geochemical compositions were used. They showed that a large number of light hydrocarbons, CO2 , and H2 S could be generated in the presence of minerals in the reaction system (Pahlavan and Rafiqul 1995; Fan 2003). Several studies have studied the influence of rock-forming minerals on the physical and chemical properties of heavy crude oil under steam-stimulation (Zhou et al. 2022). For example Petrov et al. (2018) performed an analysis of minerals’ effect on heavy oil conversion in the presence of rock-forming additives among which calcites, dolomite, kaolin clay, and manganese oxide were present in the studied medium. The experiments considered different temperatures and pressures.

4.5 The Role of Minerals as Natural Catalysts

The obtained results showed that the temperature and pressure have a significant influence on the occurred processes. In addition, the obtained samples after the aquathermolysis experiments were characterized by a lower structural and Newtonian flow behavior and by greater output of fuel and oil fractions than the heavy oil. On the other hand, resins were found to be converted into lighter components during the destruction. It was also reported that additives partially structure a monomolecular surface layer with a decrease in entropy of the observed molecules on the surface. This leads to a shift in the equilibrium toward a thermal decomposition unimolecular reaction of —C—C— bonds by radical chain mechanism. Therefore, there are two competing mechanisms. Besides, it has been found that a temperature increase raises the processes of cracking macromolecular compounds; meanwhile, a growing temperature background in the absence of high pressure has been found to reduce the adsorption probability on the additive surface. Kayukova et al. (2018) have studied the specific features of Ashal’cha heavy oil conversion at hydrothermal conditions in a CO2 atmosphere in the presence of different oil-soluble carboxylate metals based on Fe, Co, and Cu at 300 ∘ C. The obtained results have shown that the catalyst and reservoir rocks minerals impact significantly the transformation behavior of hydrocarbon compositions and the SARA fractions. Moreover, it has been found that the most significant changes in the composition of heavy oil were established after hydrothermal treatment in the presence of a rock-forming additive, clay mineral (kaolin), which are reflected in a significant increase in the content of saturated and aromatic hydrocarbons. Additionally, this was fixed by resins and asphaltenes’ amount decrease. What is more, asphaltenes’ structure changes have been manifested in an increase in aromaticity and oxidation degrees because the catalyst metals mainly concentrate on the asphaltene structures and on the rock-forming reservoir minerals (Figure 4.18). It has been found as well that the Fe concentration increases to 5.84% after the catalytic treatment of oil with kaolin, and the concentration of Cu increases as well from 0% to 0.25%. A similar change in asphaltene’s content was observed in rock extracts where iron and copper concentrations reached 1.38% and 0.31%, respectively. Meanwhile, cobalt was absent in asphaltenes content. In addition, active metals, such as Ti, Fe, Cr, Mn, Co, Ni, Cu, Zr, and Ba were found in the rock samples after organic matter extraction. Besides, the Fe concentration has been associated with a slight decrease from 2.27% to 1.93% in the catalytic complex. On the other hand, the Co and Cu concentrations have been found to increase to 0.036% and 0.038% compared to their content in the original rock in which they were practically absent. This study has shown as well that the studied metal complexes (Fe, Co, and Cu) with carboxylate ligands can penetrate not only into the rock pore space but also into asphaltenes. This may provide a possible catalytic effect on the heavy oil upgrading process under hydrothermal conditions. Figure 4.18 The amount of adsorbed metals from the catalyst on the components of the reservoir system.

Asphaltene

Co

Fe

Cu R + RCH3

R C C S Ö MexSy H H

CO2 H2

Content before an of catalyst metal s d after aq uatherm olysis

Fe

Co

Rock matrix

Cu

H2S CH4

Fe

Co

Kaolin

Cu

197

4 Fundamentals of In Situ Upgrading

Experimental studies have shown that the mineral matrix is actively involved in the process of kerogen cracking under hydrothermal conditions. An increase in contacts with process activators during the crushing of samples leads to an increase in the released synthetic oil. In this case, the amount of hydrocarbons obtained primarily depends directly on the mineral composition of the original rock and the contact of organic matter with process activators and inhibitors. Silica (primarily quartz) and clay minerals have a positive effect on the synthesis of oil from kerogen: an increase in their concentration in the rock leads to an increase in the yield of synthetic oil (Stennikov et al. 2018). Even such an inert component as calcium carbonate can contribute to the conversion of kerogen because it reacts with the resulting sulfur and leads to the formation of calcium sulfates (Al-Otoom 2005; Kazakov et al. 2017). More recent evidence (Dobrynkin et al. 2017) suggests that mineral matrices’ intrinsic catalytic properties of various kinds (basalts, clays, sandstones) are of interest for in situ heavy oil upgrading. In fact, this may result in advanced technologies for enhanced oil recovery. The elemental, surface, and phase composition in addition to matrix particle morphology, surface, and acidic properties using elemental analysis, X-ray diffraction, adsorption, and desorption of nitrogen and ammonia were studied. The obtained results have fixed for the first time the inorganic matrices’ catalytic activity of ammonium nitrate decomposition, oxidation of hydrocarbons, and carbon monoxide and hydrocracking of asphaltenes into maltenes. To check their applicability for the asphaltenes’ hydrocracking catalytic systems development, basalt and clay matrices were used as supports for iron/basalt, nickel/basalt, and iron/clay catalysts, and the obtained results confirmed the previously catalytic effect of the matrices in the reactions of ammonium nitrate, oxidation of hydrocarbons and carbon monoxide, and hydrocracking of asphaltenes decomposition. A significant effect of clay minerals on the efficiency of nickel-based aquathermolysis catalyst was demonstrated in our previous works (Aliev et al. 2021a, 2018; Vakhin et al. 2021). Significant changes in the composition of aromatic fractions have been established. Resins cracking products have been found to transform into aromatic hydrocarbons. This has increased the relative content of monoaromatic compounds in the aromatic fractions’ composition, which was studied in the range of m/z = 132.5–133.5. These results are shown in Figure 4.19 as confirmed by the results of oil viscosity measurement after hydrothermal exposure. 6000 5500 5000 4500

Viscosity (mPa.s)

198

4000 3500 3000 2500 2000 1500 1000 500 0 10

20

30

40

50

60

Temperature (˚C) Clay 250C Clay+Fe 250C

Figure 4.19 treatments.

Clay 300 Clay+Fe 300C

Clay+Ni 250C Initial crude oil

Clay+Ni 300C

Dynamic viscosity of initial crude oil before and after various catalytic hydrothermal

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance Reaction temperature is one of the main keys and parameters on the upgrading performance of unconventional resources (heavy, extra-heavy crude oils, and natural bitumen). It should be mentioned that based on the phase states of water (Figure 4.20) and the conditions of the process, the chemical reactions that occurred can be divided into aquathermolysis in sub-CW and an intensive thermal cracking of high molecular weight components like resins and asphaltenes in NC and SCW (Al-Muntaser et al. 2020; Eletskii et al. 2017; Hyne et al. 1982; Shiraishi et al. 2011). Many researchers accepted the possibility of a reaction (Eq. (4.1)) under the hydrothermal conditions, where the aquathermolysis reaction plays an important role (Hyne et al. 1982; Hart 2014; Kruse and Dinjus 2007). Production of hydrogen in situ with NCW and SCW has a great potential for their utilization to develop the process of in situ or ex situ upgrading of heavy oil (Ancheyta and Speight 2007). As shown in Eq. (4.1), the destruction of C—S bonds leads to the cleavage of individual fragments from resin and asphaltene molecules and produces methane, hydrogen sulfide, hydrogen, carbon dioxide, and low molecular weight hydrocarbons that are heavier than methane (Belgrave et al. 1994; Eletskii et al. 2017). By the destruction of carbon-heteroatom bonds, along with a change in the group and fractional composition, the structure of asphaltene molecules and their aggregates changes, thus causing a variation in the rheological characteristics of the feedstocks. Usually, there are two simultaneous reaction routes for the thermal cracking of asphaltenes: the decomposition of asphaltenes to form lighter fractions as light upgraded oils, and their condensation and aggregation to form coke. Then hydrocarbon radicals are generated by the cleavage of the aliphatic side chains and the weak bonds including (C–S, C–O, C–N) linking aromatic hydrocarbon groups. The main elementary reactions for the thermal cracking of high molecular weight hydrocarbons (resins and asphaltenes) include C–C cleavage, bond scission, hydrogen abstraction, isomerization, and addition to olefins as shown in Figure 4.21 (Clark et al. 1983; Liu et al. 2013). With the increase of temperature from subcritical point (100–320 ∘ C) to above the near-critical point (320–374 ∘ C) and supercritical point (T ≥ 374 ∘ C), the physical properties of water such as

I

Sub-CW

II

SCW

NCW

35 1.0

15

Critical point (374.3 °C, 22.1 MPa) Gas (steam)

300 °C

Dielectric constant

50

250 °C 200 °C

5

40 100

350 °C

10

10 200

Hydrocarbon solubility (wt.%)

20

0 100

Supercritical fluid

Density (g/cm3)

80

400 °C Saturated vapor pressure curve

0.5

Near-Critical water

Pressure (MPa)

25 20

Supercritical water region (SCW)

Liquid (water)

30

300

400

500

Temperature (°C)

600

700

°C

Inorganic solubility (wt.%)

100

Boiling point

200

300

400

500

600

Critical temperature

Figure 4.20 Phase diagram of water (I) and physical properties of water in subcritical, near-critical, and supercritical conditions (II). Source: Adapted from Al-Muntaser et al. (2020), Arcelus-Arrillaga et al. (2017), Can𝚤az and Erkey (2014), He et al. (2014).

199

200

4 Fundamentals of In Situ Upgrading .

. +

C–C cleavage

Mechanism 1 +H .

H-abstraction β-scission .

+ Olefins

Mechanism 2

β-scission . + Olefins Addition with Dehydrogenation

H2 +

Cyclization

.

Condensation

Figure 4.21 Main reactions related to the upgrading of heavy hydrocarbons (residual oil). Source: Yuan et al. (2011)/Elsevier.

density, dielectric constant, hydrocarbon solubility, and ion dissociation constant can be improved, which can greatly speed up the reaction rate (Figure 4.20) (Liu et al. 2013). In addition, due to the superior properties of hot pressurized water, water can play a role as a nonpolar solvent with high solubility and diffusivity. The gas and liquid phase becomes miscible, and it is difficult to distinguish gas and liquid state. It can effuse through porous media like a gas and dissolve materials like a liquid (Liu et al. 2013). Al-Muntaser et al. (2021) studied in detail the HTU of high sulfur-content heavy oil at subcritical (sub-CW), near-critical (NCW), and SCW conditions. In the mentioned work, all obtained products after HTU, including gases, liquid, and coke (if formed), were analyzed to understand the upgrading performance at different conditions. The results show that at sub-CW (200, 250, and 300 ∘ C), 250 ∘ C is the optimum temperature where a viscosity reduction from 2073 to 1758 mPa⋅s was achieved with a slight removal of sulfur (mainly sulfur) and the generation of a small amount of light and noncondensable hydrocarbons in gas phase (C1 –C4 , isoalkanes and alkenes, H2 S, CO2 , and H2 , etc.). At NCW (350 ∘ C) and SCW (400 ∘ C), heavy oil was upgraded into light oil with a significant removal of heteroatoms, an increase of saturates content, a reduction of aromatics, resins and asphaltenes contents, and a high yield of light gaseous hydrocarbons (mainly methane). Simultaneously, each SARA fraction was also greatly ameliorated: the content of light alkanes with low molecular weight in saturates was increased, diaromatics content in aromatics was increased with a reduction of polyaromatics content, and aromatics-type carbon atoms in resins was increased with a decrease in aliphatic hydrocarbons. Moreover, MALDI-TOF measurements of asphaltenes show that the molecular weights of asphaltenes were reduced. All these results indicated that HTU at sub-CW can be used for heavy oil pretreatment (in situ or ex situ upgrading) considering its main effect of viscosity reduction with a small removal of heteroatoms, while HTU at NCW and SWC has a great potential in in situ and ex situ upgrading and oil refining as it can convert heavy oil into light oil (Al-Muntaser et al. 2020).

4.6.1 Material Balance (Products Distribution) and Pressure Changes During the HTU Process The gas, liquid (i.e. produced upgraded oil), and coke yields at different reaction temperatures are presented in Table 4.8. The gas yield increased with temperature, while the liquid yield decreased with temperature. At 200 and 250 ∘ C, there is no coke (chloroform-insoluble products) being obtained. Increasing the temperature reaction to 300 ∘ C leads to a slight coke formation (0.73 wt.%).

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance

Table 4.8

Product yields in HTU process.

Reaction system

Heavy oil + N2 + steam

Temperature (∘ C)

Liquid yield (wt.%)

Coke yield (wt.%)

Gas yield (wt.%)

200

97.46

—

2.54

250

96.29

—

3.71

300

97.19

0.73

2.08

350

84.19

8.94

6.87

400

62.69

19.25

18.06

Heating period

250

Pressure (bar)

200

150

200 ˚C 250 ˚C 300 ˚C 350 ˚C 400 ˚C

100

Upgrading period

50

0

0

200

400

600

800

1000

1200

1400

Time (min) Figure 4.22

Pressure profile during HTU of studied heavy crude oil.

At NCW (350 ∘ C) and SCW (400 ∘ C) conditions, the coke was observed, and the yield was 8.94 and 19.25 wt.%, respectively. The liquid yield was significantly decreased with the great increase of gas yield and the formation of coke compared with that sub-CW condition, which means that more liquid was transformed into gases and coke. This can be attributed to the intensified thermal cracking reactions at higher temperatures such as the bond scission of the side chains of compounds present in crudes like alkylaromatics (Sanchez-Minero et al. 2013). Figure 4.22 shows the pressure change as a function of time during HTU process. At sub-CW condition 200, 250, and 300 ∘ C, there is a small increase in the pressures 0.12, 0.27, and 0.19, respectively. In addition, an obvious pressure increase was observed at NCW (350 ∘ C) and SCW (400 ∘ C) condition from 154.21 to 163.86 bar and from 233.2 to 258.52 bar, respectively, due to more released gases that resulted from intensified thermal cracking process.

4.6.2

Evolved Gas Components Analysis by Gas Chromatography

Clark et al. (1983), Belgrave et al. (1994), and Gai et al. (2016) reported that aquathermolysis of heavy oil can produce light hydrocarbons (C1 –C4 , isoalkanes, and alkenes etc.), hydrogen sulfide, carbon dioxide, small amount of hydrogen, etc. Table 4.9 shows the composition of evolved gases

201

202

4 Fundamentals of In Situ Upgrading

Table 4.9

Composition of the evolved gaseous products after the HTU.

Compounds

Gas yields (vol.%) 200 ∘ C

250 ∘ C

300 ∘ C

350 ∘ C

400 ∘ C

C1

0.0071

0.7815

0.7745

32.5590

48.3824

ΣC2

0.0117

0.0138

0.3398

23.1186

3.2077

ΣC3

0.0109

1.4338

0.1814

19.4117

15.5046

ΣC4

0.0187

0.0618

0.0462

9.0073

17.7519

H2

0.0010

0.0036

0.0472

1.0042

2.0145

CO2

0.0880

0.0801

0.3968

1.2291

1.8742

1.7468

1.9957

4.7004

7.9926

3.5509

H2 S

0.0550

0.9699

N2

98.5493

95.1737

O2

0.8084

0.9481

1.0794

1.2248

1.9425

Unidentified

0.4499

0.5337

0.1259

2.4570

1.0709

95.262

after HTU process. For light compounds (alkanes, isoalkanes, alkenes, etc.), the total content of C1 –C4 generally increased with temperature (0.05%, 2.29%, 1.34%, 84.10%, and 84.85%). The values at sub-CW condition were similar, with a slight advantage in the case of 250 ∘ C. At the NCW and SCW conditions, the volume percentages of the light and noncondensed alkanes C1 –C4 are much higher than that obtained at sub-CW condition. This means that more light components C1 –C4 were obtained as a result of the thermal cracking process at NCW and SCW conditions. The formation of light hydrocarbon gases is because of free radical reactions where the smallest hydrocarbon radicals such as methyl and ethyl were generated by the random cleavages of the hydrocarbon chains. Generally, the increase in the content of hydrocarbons gases is an indication of the oil upgrading degree. In this research, based on the content of C1 –C4 in evolved gas, it can be further inferred that the upgrading occurred in sub-CW or in NCW and SCW conditions. For other gases (H2 , H2 S, CO2 , and O2 ), as shown in Table 4.9 by increasing the temperature reaction from 200 to 400 ∘ C, their contents are increasing, and the maximum increase is observed at NCW and SCW conditions, respectively. Moreover, the increase in the content of carbon dioxide by increasing temperature reactions means that, water participated in the reaction to release its oxygen atom as an oxidant consequently, we got more carbon dioxide formation because the amount of oxygen in heavy oils is very small in general (Sato et al. 2010). In general, the production of light and noncondensable hydrocarbons and other gases by in situ upgrading could contribute toward viscosity reduction via miscible displacement and subsequently enhance oil recovery. Further, the presence of hydrogen in the gas especially at NCW and SCW conditions promotes hydroconversion (hydrogenation) of the heavy oil fragments into stable saturated molecules, if the partial pressure of hydrogen is high enough.

4.6.3

Liquid Products Analysis

4.6.3.1 Viscosity and API Gravity of Oil Before and After HTU

Table 4.10 shows API gravity and high viscosity of oil before and after HTU at different temperatures. Thermal conversion at sub-CW condition leads to the increase of API gravity and reduction of viscosity. However, the maximum viscosity reduction and increase of API for sub-CW were obtained at 250 ∘ C rather than 300 ∘ C, which is consistent with previous studies which confirm

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance

Table 4.10 API gravity and viscosity of oils before and after hydrothermal upgrading processes. Samples

Initial oil

Initial oil + water + N2

Temperature (∘ C)

API gravity (∘ )

Viscosity (mPa⋅s)

—

14.1

2073

200

14.5

1994

250

14.9

1758

300

14.3

2034

350

23.3

8.46

400

32.8

3.94

and demonstrate that the optimum temperature range for the aquathermolysis reactions of heavy oil at sub-CW condition is 220–280 ∘ C (Liu et al. 2004; Rogel et al. 2013; Song et al. 2009). At 300 ∘ C, the viscosity was slightly higher than 250 ∘ C. The condensation or recombination reaction of free radicals is the main pathway to form coke, as well as one of the main reasons that lead to an increase of upgraded oil viscosity (Guo et al. 2016). The increase of viscosity at 300 ∘ C compared with 250 ∘ C can be the signal that the condensation reaction has started to prevail. At NCW and SCW conditions, the viscosity was reduced from 2073 (initial oil) to 8.46 and 3.94 mPa⋅s, and the API gravity was increased from 14.1 (initial oil) to 23.3 and 32.8, respectively. The heavy oil has been thoroughly upgraded into light oil. However, such upgrading leads to a high coke yield. 4.6.3.2 Elemental Analysis and Desulfurization of Oil Samples During Thermal Conversion Process

The results of the elemental analysis of the oil samples before and after HTU are shown in Table 4.11. At sub-CW condition, no noticeable change was observed in the nitrogen content, but sulfur content decreased with the increase of temperature. At NCW and SCW conditions, a noticeable decrease in the content of both nitrogen and sulfur was observed, especially for sulfur: its content decreased from a 4.52% in initial oil to 2.51% at 400 ∘ C. A sulfur removal rate of 44.4% was achieved, which means that the HTU in NCW and SCW has a strong ability to remove heteroatoms (especially sulfur) without even an additional presence of catalysts and hydrogen donor. It is well known that the presence of heteroatoms (S, N, O) is one of the main reasons that lead to the high viscosity of heavy oil (Yi et al. 2018). Hypothetically, these heteroatoms usually exist in the resins and asphaltene molecules in the form of C—O, C—N, and C—S bonds (Rahimi and Gentzis 2006; Shokrlu and Babadagli 2013). In HTU process, the weakest bonds are always cleaved first. The bond dissociation energies of the main basic bonds C—H, C—C, C—N, and C—S are in turn 96–99, 83–85, 69–75, and 66 Kcal/mol, respectively. The C—S bond is the easiest to be cleaved (Shokrlu and Babadagli 2013; Yi et al. 2018). It has been reported that the C—S bond can be cleaved even at 180 ∘ C. Therefore, the decrease of sulfur content with temperature at 200, 250, and 300 ∘ C can be attributed to the cleavage of C—S bonds. However, as the nitrogen content was not decreased at sub-CW condition, it seems that the cleavage of C–N did not occur or it occurred, but the products containing N atoms still remained in the oil. At NCW and SCW conditions, the reduction in the content of N and S indicated that the cleavage of C—N and C—S bonds significantly occurred. In addition, it can be seen that the NH /NC ratio increased after HTU especially after upgrading at 250 ∘ C, NCW, and SCW condition, which supported the earlier notion of increased conversion at these conditions.

203

204

4 Fundamentals of In Situ Upgrading

Table 4.11 Elemental analysis of oil samples before and after hydrothermal upgrading processes. Temperature reaction (∘ C)

Content (wt.%)

NH /NC

Sulfur removal (%)

C

H

N

S

Initial oil

83.45

11.35

0.36

4.52

1.62

—

200

83.37

11.47

0.35

4.47

1.65

1.11

250

82.36

11.52

0.34

4.22

1.68

6.64

300

83.52

11.45

0.36

4.13

1.65

8.63

350

83.21

11.75

0.23

2.70

1.72

40.27

400

82.07

12.18

0.21

2.51

1.78

44.47

4.6.3.3 FTIR Spectroscopy of Oils Before and After Thermal Conversion

The strong absorption bands at 2923 and 2852 cm−1 can be ascribed to the stretching vibration of aliphatic groups (v(CH3 + CH2 )). The absorption bands at 1365, 1465, 1705, and 1605 cm−1 could be attributed to methyl and methylene groups (v-(CH3 + CH)2 ), carbonyl and/or carboxyl groups (v-(C = O)) and the stretching vibration of aromatic carbons (v(C = C)). In addition, several weak bands were observed from 700 to 900 cm−1 , which can be assigned to the out-of-plane deformation vibration of one isolated aromatic C—H bond (γ(CHar ), 870 cm−1 ), two or three adjacent aromatic C—H bonds (γ(CHar ), 814 cm−1 ), and four close aromatic C—H bonds (γ(CHar ), 750 cm−1 ), and the skeletal vibration of more than four methylene groups (r(CH2 )n ,720 cm−1 ). For a better comparison of the spectra among the samples before and after HTU at different temperatures, spectral indices (C1 –C4 ) were defined and calculated according to the intensity of bands (Permanyer et al. 2002; Rakhmatullin et al. 2018; Vakhin et al. 2018a, 2018b): C1 = D1600 /D720 (aromaticity index), C2 = D1380 /D1465 (branching index), C3 = (D720 +D1380 )/D1600 (aliphatic index), and C4 = D1030 /D1465 (index of sulfurization) shown in Table 4.12. After HTU, there was an increase in both the aromaticity index C1 and aliphatic index C3 in all cases of HTU (sub-CW, NCW, and SCW), which is believed to be the consequence of the formation of more small molecule compounds that resulted from the cleavage of macromolecules. Also, branching index C3 was decreased with the increasing of temperature due to the cleavage of branched alkanes. In addition, an obvious decrease in the index of sulfurization after HTU was observed, which further confirms the sulfur removal. Table 4.12 The values of spectral indices (C) calculated according to FTIR spectroscopy data. Spectrophotometric indices

Temperature (∘ C) C1

C2

C3

C4

Initial oil

0.4242

0.6088

5.8334

0.2144

200

0.4387

0.5642

7.0859

0.1648

250

0.4391

0.5684

7.0411

0.1586

300

0.4327

0.5747

6.8494

0.1783

350

0.4394

0.5541

7.1184

0.1721

400

0.4783

0.5480

7.5421

0.1839

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance

Table 4.13

SARA fraction of oil samples before and after thermal conversion.

Reaction system

Initial oil

Crude oil + steam

Temperature reaction (∘ C)

Mass content (%) Saturates

Aromatics

Resins

Asphaltenes

—

28.79

44.32

20.98

5.91

200

27.49

46.35

20.39

5.77

250

29.82

44.87

19.95

5.36

300

29.40

36.04

26.30

8.26

350

54.99

27.23

14.92

2.86

400

62.68

26.62

9.82

0.88

4.6.3.4 Changes in SARA Fractions

Table 4.13 shows the SARA fractions of oil samples before and after HTU. At 200 and 250 ∘ C, no significant change was observed for the SARA fractions. Only the content of resins and asphaltenes was slightly decreased accompanied with a small increase of either saturates or aromatics. As mentioned in Section 3.3.1, at 200 and 250 ∘ C, the viscosity was decreased. Combining with SARA fraction results, it can be inferred that the decrease in viscosity might result from the cleavage of some weak bonds (C—N and C—S, etc.) that break those macromolecules of resins and asphaltenes into small ones. Further, according to the analysis in elemental composition and evolved gas composition parts, the cleavage of C—S bonds might be the main reason that leads to the reduction in viscosity. It has been shown that some sulfur atoms connect two ring structures, such as cycloalkyl aryl sulfides, and have been evidenced to be responsible for the formation of abundant crosslink bridges in the macromolecular network in resins and asphaltenes that partially leads to the high viscosity of heavy oils. Therefore, once C—S bonds are cleaved, the decomposition of sulfur-containing fragments in resins and asphaltenes will cause a reduction in viscosity. It is noteworthy that there is no additional hydrogen donor being added in these experiments to react with the formed free radicals produced in the cleavage process, but the viscosity still decreased with the generation of some light hydrocarbons, which require more hydrogen. It can be inferred that the water somehow participated in the reactions acting as if it is hydrogen donor. Lysogorskiy et al. (2016) also indicated that the cleavage of C-heteroatom bonds in the presence of water can be an important mechanism for the viscosity reduction of heavy oils. However, at 300 ∘ C, the content of resins and asphaltenes was increased instead of their reduction, which explains the increase of viscosity at 300 ∘ C. This can be attributed to the condensation and polymerization reactions of free radicals. With the increase of temperature, more C—S bonds as well as some weak C—C bonds participated in the cracking reactions, like C—C bonds inside chains, C—C bonds in the alpha position of C6 H5 —CH2 —CH2 —CH3 , and C6 H5 —C6 H5 , C—C bonds in the beta position of C6 H5 —CH2 —CH2 —CH3 and C6 H5 —CH2 —CH2 —C6 H5 . Once the free radicals were formed, the condensation and polymerization can easily occur among them in the absence of enough hydrogen source, which will lead to the formation of macromolecules of resins and asphaltenes (Eletskii et al. 2017; Prakoso et al. 2017). At NCW and SCW (350 and 400 ∘ C) conditions, the content of saturates was greatly increased from 28.79% in initial oil to 54.99% and 62.68% at 350 and 400 ∘ C, respectively, with a great decrease in the content of aromatics, resins, and asphaltenes from 44.32%, 20.98%, and 5.91% in initial oil to 27.23%, 14.92%, and 2.86% at 350 ∘ C and 26.62%, 9.82%, and 0.88% at 400 ∘ C, respectively. The great variation in SARA fractions can explain why heavy oil has become light oil with a great decrease in viscosity (from 2073 to 8.46 and 3.94 mPa⋅s at 350 and 400 ∘ C, respectively). However, this

205

206

4 Fundamentals of In Situ Upgrading Temperature reaction

SARA fractions

Initial oil

200 ˚C

250 ˚C

300 ˚C

350 ˚C

400 ˚C

Saturates

Aromatics

Resins

Asphaltenes

Figure 4.23

SARA fractions in solvents separated from upgraded oils at Sub-CW, NCW, and SCW conditions.

significant improvement in oil quality is concurrent with the formation of coke and high yield of C1 –C4 as mentioned in Sections 3.1 and 3.2. This is the consequence of the intensification of cracking reactions. As is indicated that the HTU process is dominated by free-radical reactions (Yan et al. 2015). With the increase of temperature and pressure in NCW and SCW conditions, a large amount of C—N and C—S bonds were cleaved, which is partly evidenced by the significant reduction in nitrogen and sulfur content (Section 3.3.2). Simultaneously, weak C—C bonds like as mentioned before as well as presented in tert-butylbenzene, 2-heptyl-naphthalene, and heptylbenzene can be easily cleaved in SCW (Gu et al. 2012; Han et al. 2011). The cleavage of many C—C, C—N, and C—S bonds will lead to an enormous generation of free radicals. With the progression of these cracking reactions, the free radicals are inclined to recombine each other to form coke by polymerization and condensation reactions, especially in the absence of enough hydrogen. Figure 4.23 shows the appearance of collected SARA fractions in solvents (saturates in n-heptane, aromatics, and asphaltenes in toluene and resins in toluene + isopropyl alcohol mixture at a ratio of 1 : 1). Colors of the solutions of SARA fractions can reflect the upgrading degree of heavy oil at sub-CW, NCW, and SCW conditions. Further, the changes in these SARA fractions were characterized and analyzed in details. 4.6.3.5 Analysis of SARA Fractions 4.6.3.5.1 GC Analysis of Saturates

The carbon number distribution of saturates before and after the HTU is shown in Figure 4.24. At sub-CW condition, the variation of carbon number distribution was not obvious. At NCW and SCW conditions, the content of low molecular weight alkanes significantly increased due to the intensified cracking reactions mainly referring to the decomposition and dealkylation of the resins and asphaltenes by the cleavage of a lot of C—C, C—N, and C—S bonds. 4.6.3.5.2

GC-MS Analysis of Aromatics

Table 4.14 shows the results of GC-MS analysis for aromatics. Noteworthy changes occurred only at NCW and SCW conditions, where the content of diaromatics significantly increased from 48.07%

4.6 The Effect of the Reaction Temperature on the Hydrothermal Upgrading Performance 40

Initial 200 °C 250 °C 300 °C 350 °C 400 °C

35

Content (%)

30 25 20 15 10 5 0

C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 C33 C34 C35

n-alkane (carbon number) Figure 4.24 Table 4.14

Distributions of n-alkanes of saturated fractions before and after hydrothermal upgrading. Results of GC-MS analysis of aromatics. Content of different type of aromatics (%)

Types of aromatics

Initial oil

200 ∘ C

250 ∘ C

300 ∘ C

350 ∘ C

400 ∘ C

Monoaromatics

19.88

11.84

21.14

20.88

16.01

17.20

Diaromatics

48.07

44.33

51.07

49.98

79.30

79.24

Polyaromatics

28.41

39.75

25.69

27.07

3.66

2.44

Unidentified

3.64

4.08

2.10

2.07

1.03

1.12

in initial oil to 79.30% and 79.24% at 300 and 400 ∘ C, respectively, with a significant decrease in the content of polyaromatics (from 28.14% in initial oil to 3.66% and 2.44% at 300 and 400 ∘ C, respectively) and a small reduction in the content of monoaromatics. The formation of more diaromatics might be attributed to the destruction of polyaromatics as well as to the decomposition of resins and asphaltenes. 4.6.3.5.3 13 C

NMR Measurements of Resins

NMR spectroscopy was applied to further characterize the changes of resins before and after HTU. Figure 4.25 shows the 13 C NMR spectra of resins. The intensity of aromatic hydrocarbons obviously increased with temperature with a decrease in aliphatic hydrocarbons, which indicate that resins were decomposed mainly via the removal of the aliphatic part by cleaving C—C bonds (presented in the side chains or between two aromatic parts) together with the cleavage of C−N and C−S. It is obvious that the more upgraded the oil was (at 250, 350, and 400 ∘ C), the more significant these changes were. 13 C NMR spectra contain a large number of distinguishable signals that can be assigned to different typical fractions, such as primary (methyl groups CH3 ) – Cp , secondary (methylene groups CH2 ) – Cs , tertiary (methine groups CH) – Ct , quaternary (C) – Csq and aromatic types of carbon atoms, Car . The corresponding regions of these fractions (Cp , Csq , Ct , Car ) in 13 C NMR spectra can be integrated to estimate their molar content, thus providing a quantitative description as described in our previous work (Rakhmatullin et al. 2017, 2018). Table 4.15 shows

207

208

4 Fundamentals of In Situ Upgrading

400 ˚C

400 ˚C

350 ˚C

350 ˚C

300 ˚C

300 ˚C

250 ˚C

250 ˚C

200 ˚C

200 ˚C

Initial oil

160

150

Initial oil

140

130

120

110

100

90

70

60

50

40

30

20

10

Chemical shift (ppm) (b)

Chemical shift (ppm) (a)

Figure 4.25 13 C (175 MHz) NMR spectra of samples before and after HTU in CDCl3 at 30. (a) Aromatic hydrocarbons (b) Aliphatic hydrocarbons. Table 4.15 Molar fractions of Cp , Csq , Ct , and Car calculated according to the 13 C NMR spectra of resins. Molar fractions (%)

Group type Initial oil

200 ∘ C

250 ∘ C

300 ∘ C

350 ∘ C

400 ∘ C

Cp

9.5

10.8

9.6

10.2

6.3

5.4

Csq

35.1

40.1

36.0

38.5

23.5

12.5

Ct

12.8

17.4

13.8

18.9

15.2

7.7

Car

42.6

31.7

40.6

32.4

55.0

74.4

the obtained results. In general, the higher the upgrading degree of the oil is, the lower Cp , Csq , and Ct values, and the higher Car values are obtained. At 400 ∘ C, for upgraded light oil due to the intensified thermal cracking reactions, the values of Cp , Csq , and Ct (5.4, 12.5, and 7.7) were far lower than that of initial oil (9.5, 35.1, and 12.8), and the Car value (74.4) was much higher than that of initial oil (42.6). Car could be considered as an indication of upgrading degree. 4.6.3.5.4

MALDI-TOF Measurements of Asphaltenes

Figure 4.26 shows the results of MALDI-TOF/TOF mass spectrometry analyses for asphaltenes before and after the HTU at 200–400 ∘ C. All measurements were conducted by a Bruker ultraflex III MALDI-TOF/TOF mass spectrometer. The peak of the asphaltenes of initial oil was around 1650 amu. The upgraded heavy oil at sub-CW (200, 250, and 300 ∘ C) showed a slight decrease of the peak values 1624, 1608, and 1639 amu, respectively. At NCW 350 ∘ C and SCW 400 ∘ C, the peaks were significantly decreased to 1035 amu and 920 amu, respectively. This decrease means that the molecular weights of asphaltenes were decreased after HTU and illustrate that the structures of asphaltenes were destructed into smaller compounds by the cleavage of C—S, C—C, and C—N bonds (Ancheyta et al. 2010). In addition, according to the results of the MALDI mass spectra,

Asphaltenes of initial oil

1000

800

600

600

400

400

200

200

0

1000

2000

3000

4000

5000

1200

m/z

0

1000

2000

3000

4000

m/z

5000

200 ºC

350 ºC

1000

1000

800

800

600

600

400

400

200

200

0

1000

2000

3000

4000

5000

m/z

250 ºC

1200

0

500

1000

1500

2000

2500

3000

3500

4000

5000

1000

800

800

600

600

400

400

m/z

400 ºC

1200

1000

200

200

0

1000

Figure 4.26

300 ºC 1000

800

2000

3000

4000

5000

m/z

0

500

1000

1500

2000

MALDI mass spectra of asphaltenes of initial and upgraded oil thermally treated at 200–400 ∘ C.

2500

3000

3500

4000

5000

m/z

210

4 Fundamentals of In Situ Upgrading

the asphaltenes obtained after thermal treatment at near-critical 350 ∘ C and supercritical condition 400 ∘ C, supposedly contained lower molecular fragments than that obtained at subcritical condition.

4.6.4

Analysis of Coke Obtained at 300, 350, and 400 ∘ C using FTIR-Spectroscopy

Figure 4.27 shows the FTIR-spectra of obtained coke after HTU process at 300, 350, and 400 ∘ C. The spectra at 300, 350, and 400 ∘ C are different due to the difference in cracking severity. The adsorption of aliphatic bonds CH2 and CH3 identified at 2920–2850 cm−1 and the bending vibrations of the methyl (δCH3 ) and both methyl and methylene groups (δ-(CH3 + CH2 ) connected with aromatics at 1375 and 1460 cm−1 were significantly decreased by increasing the upgrading temperature from 300 to 400 ∘ C, respectively. This means that more aliphatic fragments were removed due to the cleavage of C-heteroatoms bonds (C—S, C—N, C—O and C—C) in the side chain with the intensification of cracking reactions. It should be noted that the addition of aliphatic hydrocarbons like olefins described in Figure 4.21 to form higher fused-ring aromatics and their further condensation, dimerization, or oligomerization to form coke (Al-Muntaser et al. 2018; Hart 2014; León et al. 2016; Rodriguez-DeVecchis et al. 2017). Simultaneously, the intensification of polymerization and condensation of aromatics at SCW can also be evidenced by the increased intensity of polyaromatics between 870 and 740 cm−1 and the decrease in the intensity of double bond C=C in aromatics at 1600 cm−1 . In general, as the temperature increases from 300, 350 to 400 ∘ C, coke becomes more condensed and characterized by a high content of polyaromatic hydrocarbons as shown in Figure 4.27. Generally, HTU of heavy oil was carried out at sub-CW, NCW, and SCW conditions. The results showed that the viscosity reduction can be achieved at sub-CW together with a slight removal of sulfur and formation of a small amount of light hydrocarbons in gas phase such as hydrocarbons with carbon number C1 –C4 (isoalkanes and alkenes, etc.), hydrogen sulfide, carbon dioxide, small amount of hydrogen, etc. The upgrading of heavy oil at sub-CW depends on temperature, and it requires an optimum temperature. In this study, at sub-CW the maximum viscosity reduction was obtained at 250 ∘ C. A higher temperature, such as 300 ∘ C, led to an increase of viscosity with an Polyaromatics γ CHar

300 ˚C 350 ˚C 400 ˚C

S=O st δas CH2 C=C aromatics

δs CH3

vas CH3

vs CH2

OH

1000

1500

2000

2500

Wavenumber (cm–1)

Figure 4.27

FTIR spectra of obtained coke after the HTU.

3000

3500

4.7 Hydrodesulfurization

increase of resins and asphaltenes content, which indicates that polymerization and condensation reactions started to prevail. In general, the upgrading at sub-CW at lower temperature (200 and 250 ∘ C) did not result in a significant change of SARA fractions, and the main reaction under these conditions was the cleavage of C-heteroatoms bonds (C—S and C—N, etc.), and some weak C—C bonds like C—C bonds in side chains such as C6 H5 —CH2 —CH2 —CH3 and connecting two aromatic fragments such as C6 H5 —CH2 —CH2 —C6 H5 . At NCW and SCW, the heavy oil was totally upgraded into light oil with a great decrease in viscosity and heteroatoms content (especially sulfur content). In terms of SARA fractions, the saturates were significantly increased with a great decrease in the content of aromatics, resins, and asphaltenes. What is more, the chemical composition of the SARA fractions was also greatly upgraded due to the intensified hydrothermal cracking reactions: (i) for saturates, the content of low molecular weight alkanes significantly increased; (ii) for aromatics, the content of diaromatics significantly increased with a noteworthy reduction in the polyaromatics content and with a small decrease in the monoaromatics content; (iii) for resins, the aromatic types of carbon atoms (Car ) significantly augmented with a decrease in aliphatic hydrocarbons, which indicate that resins decomposed mainly by the removal of the aliphatic part (a high Car value could be considered as an indicator of high upgrading performance); (iv) the decomposition of asphaltene was increased by increasing the temperature reactions from 200 to 400 ∘ C mainly forward to lower molecular weight side, and this is reflected in the peak of MALDI mass spectra of asphaltenes. However, at sub-CW condition coke formation was observed exclusively at 300 ∘ C, and on the contrary, a great amount of coke was formed at NCW and SCW conditions due to the occurrence of polymerization and condensation. In addition, the generations of light and noncondensable hydrocarbons like C1 –C4 (isoalkanes and alkenes, etc.), hydrogen sulfide, carbon dioxide, and small amount of hydrogen were increased. In general, HTU at sub-CW, NCW, and SCW can achieve an oil upgrading. At sub-CW under an optimum temperature range, the main effect of HTU is the reduction of viscosity with a small removal of heteroatoms, while at NCW and SCW, a total conversion from heavy oil to light oil can be achieved with a significant removal of heteroatoms without an additional presence of catalysts and hydrogen donors. For ex situ upgrading, HTU in sub-CW can be used as a pretreatment of heavy oil for the convenience of transportation. HTU at NCW and SCW can be applied for oil refining and processing. However, for an in situ upgrading process, the formation of coke at NCW and SCW means that a lot of crude oils will be lost and will remain in reservoirs at the form of immobile hydrocarbon deposits, which are a huge waste. Simultaneously, such a high temperature (350 or 400 ∘ C) can be only achieved at the expense of high cost for steam generation. Therefore, some cheaper methods to generate heat and low-cost methods to prevent the formation of coke like using catalysts and hydrogen donors are required for upgrading at NCW and SCW conditions. For HTU at sub-CW, the use of catalyst or hydrogen donor or their combination is also necessary to yield a better viscosity reduction at a lower temperature, thus reducing the injection volume of steam and achieving a higher performance of steam injection.

4.7

Hydrodesulfurization

The removal of heteroatoms, particularly sulfur-containing compounds, from the composition of crude oil and petroleum products is crucial because the evolvement of sulfur dioxide after combustion of fuels leads to acid rain and environmental pollution. Moreover, sulfur compounds tend to poison the catalysts in every step of oil processing and lead to the corrosion of pipelines, pumps, and installations. Therefore, requirements of petroleum refineries for the crude oil

211

212

4 Fundamentals of In Situ Upgrading

are getting higher day by day. The most widely applied method in the petroleum industry to remove sulfur compounds is hydrotreatment or hydrodesulfurization. It is defined as a chemical reaction between hydrogen and sulfur-containing compounds in the presence of the appropriate catalysts under high pressure and temperature conditions. Hydrogen sulfide is a product of such chemical interactions, which can be easily separated from the crude oil. Since the discovery of aquathermolysis reactions, the in situ upgrading of heavy oil is widely employed to remove sulfur from organosulfur compounds. According to the proposed mechanism of aquathermolysis, the hydrogen produced from the water because of its dissociation and WGSR is of little value, if it cannot be made to react further with the sulfur-containing compounds of crude oil. Therefore, introduction of catalysts into the reservoir formations can essentially promote the in situ hydrodesulfurization of the crude oil. Moreover, the catalytic hydrodesulfurization process occurs in the cheapest HP/HT reactor of all, the reservoir traps. In Section 4.5, we have provided the catalytic role of rock minerals on the aquathermolysis process. This suggests that the minerals in the sand have an important catalytic role in hydrodesulfurization when the whole core material is aquathermolyzed. According to Hyne et al. (1982b), there can be two possible sources of H2 S during steam flooding operations, one of which is for sure hydrodesulfurization of heavy oil. On the other hand, the H2 S can be generated from the decomposition of rock minerals. To avoid the latter source, laboratory experiments were carried out. It was shown by separate aquathermolysis of extracted sand that the contribution of H2 S from the porous medium is not a major factor in these cases. No H2 S whatever was generated from aquathermolysis at 240 ∘ C of the sand from the Venezuelan sample and a Canadian sample sand generated some 2% of the total H2 S generated when the whole core sample was aquathermolyzed. We believe this was due to small amounts of iron sulfide in the sand. Interestingly, when the separated bitumen from both samples was aquathermolyzed in the absence of sand and any added catalyst no H2 S was generated.

4.8

Evolved Noncondensable Gases

Belgrave et al. (1994), Clark et al. (1983), and Guo et al. (2016) reported that aquathermolysis of heavy oil could produce light hydrocarbons such as C1 –C4 , isoalkanes, and alkenes, hydrogen sulfide, carbon dioxide, and small amount of hydrogen. Sections 4.8.1 and 4.8.2 will discuss the individual noncondensable evolved gases.

4.8.1

CO2 and CO Production

Carbon dioxide is a major component of the gas phase produced during aquathermolysis. The amount and the rate at which it is produced is important not only as a measure of the extent of chemical reaction that produces it but also because it has been shown that when dissolved in the heavy oil, it can cause significant reduction in viscosity. Hyne (1986) was the first that studied the general pattern of the CO2 yield after aquathermolysis process in case of heavy crude oil samples produced from Venezuela and Canada. The Venezuelan sample produced more CO2 than the Canadian core material at all temperatures, but the general pattern of the rate of production is the same in all the cases. There is an initial rapid release of CO2 to the gas phase primarily due to release of CO2 from mineral carbonates, followed by a slower and declining rate of production over a much longer time period. X-ray diffraction indicated substantiality more carbonate material in the Venezuelan sample, although different carbonates will decarboxylate at different rates, and it requires relatively little reactive carbonate to produce the kind of CO2 concentrations measured in this work. The amount of CO2 produced

4.8 Evolved Noncondensable Gases

in the initial stage is relatively constant for each sample irrespective of temperature suggesting that the CO2 production from the mineral carbonate occurs quickly and more or less completely at all three temperatures. The decarboxylation of the mineral carbonates, rapid though it may be in all cases, is still faster at 300 ∘ C than at 200 ∘ C, as it should be. After mineral decarboxylation is essentially complete, a much slower CO2 production ensues, which they attribute to aquathermolysis of the organic constituents of the oil sand, in particular the organosulfur compounds and the carboxylic acids. Both the amounts and rates of CO2 production are highly temperature-dependent as would be expected from the kinetic theory. To prove the mineral source of CO2 , the debitumenized Venezuelan sample was aquathermolyzed. A sample of the same homogenized core was extracted with methylene chloride and alcoholic sodium hydroxide to generate an organic-free sand. An amount of this clean sand equivalent to that present with 1000 g of bitumen in the whole core was then aquathermolyzed under precisely the same conditions as the whole core. The amount of CO2 is thus likely associated with hydrolysis of the mineral carbonates in the sand as postulated. We have observed the same behavior (Figure 4.28) during the hydrothermal treatment of extra-heavy oil and oil-saturated crushed core samples extracted from Boca de Jaruco Oil Field, Republic of Cuba. The temperature and pressure of the steam were 300 ∘ C and 90 bar, respectively. The gas factor after the aquathermolysis of oil-saturated crushed rocks is more than two times greater than aquathermolysis of extra-heavy oil. Moreover, the added metal ions further increase the yield of gases. It has been argued that decarboxylation of carboxylic acids is a significant source of CO2 in the steam treatment of oil sands. We have examined this possibility by extracting a core sample with CH2 Cl2 – a relatively nonpolar solvent in which most of the organic phase is soluble. Polar carboxylic acids, strongly absorbed onto the sand phase, however, may not be removed. Thus, such sand extracted with CH2 Cl2 may indeed still contain significant quantities of carboxylic acids. When subjected to aquathermolysis at 240 ∘ C for 14 days, it yielded 350 ml of CO2 per kilogram of sand. Another sample was then washed with methylene chloride and 1 m alcoholic NaOH which removed any absorbed carboxylic acids. Infrared analysis of the recovered extract showed evidence of carboxylate groups (1710, 1440, 1360, and 1020 cm−1 ). Sand so extracted was again subjected to aquathermolysis and the CO2 production per kilogram of this sand was 160 mls compared with 60 Experiments with oilsaturated crushed core samples in the presence of Ni tallate

50

Experiments with oilsaturated crushed core

Gas yield (m3/ton)

40

30

20 Experiments with crude oil

10

0 0

5

10

15

20

25

30

Exposure time (h) Figure 4.28

Comparison of gas yields after steam treatment of crude oil and oil-saturated crushed rocks.

213

214

4 Fundamentals of In Situ Upgrading

350 mls without caustic ethanol wash. The CO2 production from CH2 Cl2 extracted sand thus drops by more than 50% upon further extraction with NaOH. This suggests that the carboxylic acid present in sand are retained on the surface of the CH2 Cl2 extracted sand and are a significant source of CO2 as a result of decarboxylation under aquathermolysis conditions. If CO2 is generated from both the mineral and organic components of the whole oil sand, it should be possible to perform three aquathermolysis treatments on – (i) the whole core, (ii) the separated bitumen, and (iii) the separated sand and to obtain a mass balance with respect to CO2 produced in the three aquathermolysis. Our results from this experiment are shown in Table 4.16. Two separated oil sand samples were examined with substantially different CO2 production, especially from the mineral phase. It is seen that in both cases, the total CO2 produced from the separated sand plus that from the bitumen exceeds that from the whole core by 33% in sample 1 and 19% in sample 2. It is suggested that this apparent lack of mass balance is due to the fact that in the whole core, the mineral carbonates in the sand are “protected” from attack by high temperature water to some extent by the surface layer of bitumen and the absorbed carboxylic acids. Once these are stripped off by solvent extraction, and the clean sand is aquathermolyzed, the mineral carbonates are subject to more direct aquathermolysis and produce more CO2 . The CO data in Table 4.16 provide a good illustration of the important catalytic role played by the minerals in the sand in the WGSR conversion of CO to CO2 . When the separated bitumen is aquathermolyzed, substantial quantities of CO are detected. This suggests that while the minerals in the sand may not be critical catalysts for the initial cleavage of the organosulfur species, which eventually results in CO production, they are essential in the subsequent conversion of CO to CO2 . The very much lower CO levels detected in the gas phase produced on aquathermolyzing of the whole core (with sand present) demonstrates the effective WGSR catalytic role of the minerals in the sand. The mineral sand particles are believed to be surrounded by a thin layer of absorbed water in which we suggest the polar – COOH end of the high molecular weight carboxylic acids (humics, fulvics, etc.) that may be absorbed and thus linked to the polar sand minerals. When this sand, stripped of its oil/bitumen cover, is aquathermolyzed, the carboxylic acids (RCOOH) and mineral carbonates (MCO3 ) will be the main source of CO2 through decarboxylation. When the oil sand is “whole” and the oil/bitumen is in place, the low polarity hydrocarbon “tails” of the adsorbed carboxylic acids (and other polar organics) can act as a transition medium between the oil and the sand particle by interacting with the low polarity oil phase. The adsorbed carboxylic acids thus form from the bridge between the polar mineral phase and the nonpolar oil acting as a sort of a “reverse soap” enabling the polar sand to “dissolve” in the nonpolar oil. In such a situation, aquathermolysis acts on both the oil and the sand producing CO2 from both sources Table 4.16 A mass balance with respect to CO2 produced after aquathermolysis process. Gas components

Whole core

Bitumen

7980

1050

82

3005

Sand

Sample 1 CO2 (mls) CO (mls)

9453

Sample 2 CO2 (mls) CO (mls)

5075

2076

29

2241

3958

4.8 Evolved Noncondensable Gases

but with reduced efficiency from the sand because of the adsorbed oil layer. The presence of the sand minerals in close proximity to the aquathermolyzing oil enables the minerals to act catalytically in the WGSR conversion of the CO produced by aquathermolysis of the oil to yield CO2 and hydrogen. Carbon monoxide (CO) has been proposed as an important intermediate in the aquathermolysis process (Figure 4.2). The produced CO, however, is the feed for the WGSR and thus subsequent depletion of CO in the gas phase is observed. Although we do not wish to over interpret the quantitative aspects of this data, it is noteworthy that the amount of CO generated is greater at the higher temperature as might be expected from the enhanced aquathermolysis occurring at 300 ∘ C compared with 200 ∘ C. The WGSR has recently been shown to occur at relatively low temperatures if the appropriate catalysts are present. Thermodynamically, the WGSR equilibrium predicts that significant quantities of hydrogen and carbon dioxide will be present in the 200–300 ∘ C aquathermolysis window temperature range. This equilibrium will continue to be displaced to the right if the hydrogen is consumed in other “upgrading” reactions. As this occurs, more CO2 will be produced and thus the ratio of CO2 /H2 in the gas phase resulting from the WGSR will not necessarily be 1 : 1. In the reverse direction, of course, any other source of CO2 such as decarboxylation of mineral carbonates will tend to suppress the WGSR, minimize the production of hydrogen in situ, and reduce the beneficial upgrading reactions resulting from hydrogen uptake. These factors thus suggest that the nature of the porous medium and in particular its mineral carbonates may have a significant effect on the role played by aquathermolysis during in situ steam stimulation. The field operating significance of this observation is that the extent to which desirable hydrogen may be produced in situ as a result of WGSR conversion of added or aquathermolytically generated CO will depend to some extent on the catalytic activity of the minerals present in the porous medium of the reservoir or the metals incorporated in the organometallic structures of the heavy oil and bitumen.

4.8.2

Methane and C2+ Generation

The generation of methane is due to the homolytic rupture of the C—C bonds that serve as an indicator to evaluate the overall aquathermolytic reactions. It was proposed that methane can interact with water to produce hydrogen under the reservoir conditions, and hence, considering methane gas as a hydrogen donor in heavy oil upgrading processes. Fan et al. (2001) noticed the decrease in the amount of evolved methane gas with increasing the amount of water to the reaction medium. The authors also assumed that methane could react with water and form hydrogen, carbon monoxide, and carbon dioxide in the presence of the catalyst. Maity et al. (2010) reported the production of light hydrocarbon gases (C2 –C4 ) due to the pyrolysis of the unstable alkyl side chains during catalytic aquathermolysis reactions. We have evaluated the yield of methane and CO2 after aquathermolysis of heavy oil in the presence of iron carboxylate catalyst precursor in case of Ashal’cha heavy oil sample (Table 4.17). The methane formation was observed only at 350 ∘ C, which is probably due to the cleavage of C—C bonds. In the lower temperature ranges, the weakest bonds such as C—S, C—N, and C—O are cracked. Therefore, the most structural changes in the composition of heavy oil and the formation of new light hydrocarbons, as well the most viscosity reduction degree were specific for the oil sample catalytically treated at 350 ∘ C. Clark and Hyne (1990) reported that methane gas was among the main gas products after all series of hydrothermal experiments were carried out on the oil-saturated reservoir rock samples from Cold Lake, Peace River, Ethel Lake, and Athabasca fields. Katritzky et al. (1996) explained the formation of methane and carbon dioxide after steam treatment of heavy oil as the destructive products of oxygen-containing compounds in the oil composition, which is formed during hydrolysis.

215

4 Fundamentals of In Situ Upgrading

Table 4.17 Composition of the gaseous products of catalytic aquathermolysis of heavy oil sample. Aquathermolysis temperature (∘ C)

Content (%)

250

300

350

N2 +O2

96.30

90.64

93.75

CO2

3.70

9.36

3.52

Methane

None

Traces

2.73

Ovalles et al. (1995) demonstrated the hydrogen-donating performance of methane in the presence of a dispersed molybdenum catalyst at 410 ∘ C. The hydrogenation of the cracked fragments of heavy oil leads to an increase of 7 ∘ in the API gravity, 16% of reduction in sulfur content, and 55% conversion of the >500 ∘ C fraction with respect to the original crude oil. The aquathermolysis gases were generally composed of C2 + alkanes, C9 –C13 ketones, fatty acids, and aldehydes. Kayukova et al. (2016) investigated the influence of water on the quality of heavy oil under hydrothermal-catalytic processes at 360 ∘ C. She reported that the more intense gas was generated in the presence of an aqueous phase. Particularly, the content of ethane and butane were increased in the gaseous products. Sitnov et al. (2022) evaluated the composition of “gas cap” after aquathermolysis of oil-saturated sandstone rocks in the presence and the absence of iron oxide nanoparticles. The GC analysis showed an increase in the total amount of newly formed gases from 0.08 g/100 g of rock at 200 ∘ C to 1.45 g/100 g of rock at 300 ∘ C with increasing temperature and catalyst’s presence. As a result, the composition was found to adopt more complicated structure, characterized by both normal and isomeric alkanes, including methane, ethane, propane, and butane (Figure 4.29). Broadly speaking, methane, ethane, and propane started to form at 250 ∘ C under noncatalytic steam treatment due to the presence of catalytically active rock minerals (Table 4.18). However, strong evidence of increasing the yield of these gases more than twice (from 0.2 to 0.43 g/100 g rock) was found by adding a catalyst based on iron oxide nanoparticles, which are comparable to the experiment performed at 300 ∘ C in the absence of the catalyst. This also indicates Without catalyst

Amount of gases (g/100 g rock)

216

Without catalyst

With catalyst

With catalyst

Without catalyst

1.4

CO2

ethane i-butane n-butane n-butane i-pentane n-pentane n-hexane n-heptane Methylcyclohexane

1.2 1.0 0.8 0.6

1.48 g

Other CH 4 gases

H2 S

0.4

0.0

Other gases CH4

H2S

1.6

0.2

With catalyst

Other gases

H2S

0.08 g

CH4

H 2S

CO2

CO2

0.51 g 0.43 g

CO2

CO2

CO2 200

H2S

Other gases

Other gases

0.20 g

0.11 g

200

250

250

300

300

Temperature experiments (˚C)

Figure 4.29 Mass content and gases’ composition (“gas cap”) formed from the aquathermolysis of reservoir rock. Source: Adapted from Sitnov et al. (2022).

4.8 Evolved Noncondensable Gases

Table 4.18 rock.

Mass content and gases’ composition (“gas cap”) formed from the aquathermolysis of reservoir

Without catalyst at 200 ∘ C

With catalyst at 200 ∘ C

Without catalyst at 250 ∘ C

With catalyst at 250 ∘ C

Without catalyst at 300 ∘ C

With catalyst at 300 ∘ C

Amount of gases (g/100 g rock)

0.08

0.11

0.2

0.43

0.51

1.45

Methane

0.11

0.09

2.05

9.38

9.83

13.10

Ethane

0.1

0.05

1.51

8.99

8.86

16.11

Propane

0.12

0.06

1.57

8.26

8.43

19.35

i-butane

0.03

0.04

0.18

1.29

1.23

2.60

n-butane

0.1

0.15

0.79

3.58

3.47

6.42

i-pentane

0.15

0.55

0.26

0.95

1.05

1.43

Samples

n-pentane

0.23

0.92

0.45

1.42

1.56

1.92

n-hexane

0.97

1.83

0.24

0.57

0.64

0.54

n-heptane

1.09

0.97

0.06

0.17

0.16

0.08

Methylcyclohexane

0.84

0.74

0.03

0.09

0.04

0.03

CO2

84.00

82.65

80.09

40.12

35.26

18.00

H2 S

0.16

0.19

9.19

16.54

18.71

9.70

Other gases Total (%)

12.09

11.76

3.58

8.64

10.75

10.72

100.00

100.00

100.00

100.00

100.00

100.00

an increase in the catalytic ability of the reservoir minerals which result in a synergistic effect during catalytic steam treatment. The influence of iron oxide concentration on the evolved gases after aquathermolysis process was illustrated by Mukhamatdinov et al. (2022). The yield of individual gases, alcohols, and other gases, and mixtures of C4 –C10 isomers, olefins + dienes, and cycloalkanes + aromatic hydrocarbons per 1 ton of oil, depending on the concentration of magnetite suspension in the oil during aquathermolysis at 300 ∘ C is illustrated in Figures 4.30 and 4.31, respectively.

Gss content (g/t oil)

300 250 200 150 100 50

Without magnetite

+0.2% Fe3O4

ol s O th er ga se s

lc oh

e

+0.6% Fe3O4

A

an

e an

pe nt n-

nbu t

3H 8

C

2H 6

C

4

CH

2

H

2

CO

H

2S

0

+1.0% Fe3O4

Figure 4.30 Yield of individual gases, alcohols, and other gases depending on the concentration of magnetite suspension. Source: Adapted from Mukhamatdinov et al. (2022).

217

4 Fundamentals of In Situ Upgrading

700 600 Gss content (g/t oil)

218

500 400 300 200 100 0 Isomers C4–C10 Without magnetite .

Olefins and dienes +0.2% Fe3O4

+0.6% Fe3O4

Cycloalkanes and aromatics +1.0% Fe3O4

Figure 4.31 The content of C4 –C10 isomers, olefins + dienes, and cycloalkanes + aromatic hydrocarbons depending on the concentration of the magnetite suspension. Source: Adapted from Mukhamatdinov et al. (2022).

Data from catalytic aquathermolysis show that the increase in gas phase content at 300 ∘ C is due to the cleavage of the longer hydrocarbon chains into hydrocarbons with a smaller number of carbon atoms. The presence of a catalyst promotes decarboxylation reactions, as indicated by the significantly greater amount of carbon dioxide produced as the catalyst concentration increases. In the presence of a catalyst, the content of hydrogen sulfide decreased, probably marking the process of transition of the oxide form of the catalyst to the sulfide form. An increase in the content of n-C3 –C5 as well as C4 –C10 isomers confirms the elimination of alkyl radicals in the side chains of the cyclic structures of resins and asphaltenes. In addition, the decrease in olefins and dienes compared to the control experiment probably indicates the hydrogenation of the obtained compounds during cracking as a result of hydrogen transfer with the participation of a hydrogen donor solvent.

4.9

Field Tests

Field tests of the catalytic aquathermolysis of aromatic sulfonic iron were conducted in the Henan oilfield, China. G61012 and G6606 oil wells were chosen for the tests. The mass of oil was calculated according to the exploitation radius and some parameters of the oil layers such as infiltration ratio, pole ratio. The water was added by injecting the high temperature and pressure steam, which was adjusted to be basic by adding NaOH to make the oil floor conditions accord to the laboratory conditions. The catalyst was added into the oil floor which is used as a nature reactor at an interval when the steam was injected. They were added in the laboratory optimum ratios. The nitrogen was injected if the oil floor did not have enough pressure. The parameters of field tests were as follows: for the G61012 oil well, steam of 998 m3 , and no nitrogen, was injected into the oil floor; for the G6606 oil well, steam of 1120 m3 and nitrogen of 23 000 m3 were injected. After the well was closed for 3 days, 14 days after reopening, the well was determined to be a period to evaluate the field tests, the yield, and the viscosity reduction ratio for a period were the most important parameters to evaluate the field tests (Chen et al. 2008). The production parameters of the G61012 and G6606 oil wells can be seen in Figure 4.32. It is not difficult to see that for G61012, the addition of production was 188.7 tons in a period, the viscosity reduction was 79.66%, the ratio of oil to steam increased by 0.12, and the ratio of production to injection increased by 0.26 and that, for G6606, the addition of production was 217 tons, the viscosity

4.9 Field Tests

Production parameter of G61012 before and after aqathermolysis technology 115 557 Parameter in a period before aquathermolysis technology Parameter in a period after aquathermolysis technology 23 000

188.7

0.43 0.19

0.185

0.105

900 998

68.1 0.07

0.17 23 500

0 Steam Nitrogen 3 3 injection (m ) injection (m ) Production (t)

Oil/Steam

Production /injection

Average Oil (mPa•S) pressure(MPa) viscosity

(a) Production parameter of G6606 before and after aquathermolysis technology Parameter in a period before aquathermolysis technology Parameter in a period after aquathermolysis technology 53579.1 0.67 500.4 23 000 23 000 0.45 1120 915

283.4

0.30

0.39 0.074

0.033

Steam Nitrogen 3 3 injection (m ) injection (m ) Production (t)

Oil/Steam

Production /injection

9510.7

Average Oil (mPa•S) pressure (MPa) viscosity

(b) Figure 4.32

Production parameters of (a) G61012 and (b) G6606.

reduction was 82.25%, the ratio of oil to steam increased by 0.15, and the ratio of production to injection increased by 0.28, respectively (Chen et al. 2008). The results showed the evident effect of aquathermolysis technology, and the aromatic sulfonic iron was a good aquathermolytic catalyst for exploitation of heavy oil. A field test for the catalytic aquathermolysis of a kind of super heavy oil was conducted in the Xinjiang Oil field of China. The F10223 oil well was chosen for the test, whose basic petro physical properties of its oil-bearing formation are as follows: the average porosity is about 30.0%, the average oil saturation is about 73.3%, the average permeability is about 1052 mD, the average vertical permeability is about 834 mD, the reservoir depth range is about 159.5−189.5 m, and the thickness of the pay zone is about 10.0 m, and the viscosity is about 8.5 × 104 mPa⋅s at 50 ∘ C. In this technology, the mass of oil was calculated according to the exploitation radius and some parameters of the oil layers such as infiltration ratio, pole ratio. The water was added by injecting the high temperature and pressure steam to make the oil floor conditions the same as laboratory conditions. The catalyst

219

220

4 Fundamentals of In Situ Upgrading

Table 4.19

The group composition of produced oil before and after the field test.

Oil sample

Viscosity (mPa. s)

Saturated HC (%)

Aromatic HC (%)

Resins (%)

Asphaltenes (%)

Before field test

85 000

31.21

20.85

34.33

13.61

After field test

12 900

49.29

24.50

18.82

7.39

was injected with superheated steam into the oil floor, which was used as a natural reactor when the steam was injected. The parameters of field tests were as follows: 1200 m3 steam was injected into the oil floor; then the well was closed for six days; after reopening, the viscosity reduction and four group compositions of the produced oil were analyzed to evaluate the efficiency of field test. Before the field test, the F10223 well had already been exploited for two CSS periods, and its production conditions have declined obviously. To postpone the declining trend, the new catalyst has added in the third CSS period. During the first month (31 days), it is found that the viscosity of the produced oil was reduced by 84.82% on average after the catalytic aquathermolysis reaction, and the oil production is about 136 t. Table 4.19 lists the group compositions of the produced oil before and after the field test. It can be seen from it that the resin and asphaltene have decreased by 15.51% and 6.22%, while the saturated HC and aromatic HC have increased by 18.08% and 3.65% after the field test. These results proved that the catalytic aquathermolysis method has an evident effect for exploitation of F10223 heavy oil, and the new catalyst has excellent efficiency in the field test. The field test that applied aquathermolytic technology in puff-and-huff operation was carried out in Qi-40 and Qi-108 blocks of Liaohe Oil field, China (Wen et al. 2007). The reservoir lithology is a mixture of sand, clay, and gravel that is mostly pebbly sandstone. The rock has a high percentage of matrixes with high montmorillonite content. The reservoir has good connectivity, and its connective coefficient is 0.85. Reservoir porosity is 23–30% with 15–20 μm pore throat radii. Reservoir permeabilities are between 1.2 and 2.0 Darcies. First, a little part of steam was injected into the well to preheat the reservoir. Then the catalytic solution (molybdenum oleate) was injected into the well with the concentration of 0.4 wt.% from the oil in-place. Surplus steam was injected according to design parameter. Finally, start producing after a 7–10 days shut-in (Wen et al. 2007). The production from the test wells before and after the catalyst injection were shown in Table 4.20. Table 4.20 shows that the production of heavy oil was obviously improved after the catalyst injection. From Table 4.21, it is clear that the oil compositions were changed after the aquathermolytic treatment. It shows that the viscosity decreases by 78.2% (measured at 50 ∘ C) in field test, at the same time, hydrogen in the heavy oil increased significantly. The results indicate that the oxygen, sulfur, and nitrogen content decreased. This is because the aquathermolysis usually occurs in the heteroatom compounds in heavy oil. The oil has more saturate and aromatic components, which are lighter, and less resin and asphaltene components, which are heavier. The comparison of heavy oil production in Qi40-23-35 oil well before and after catalyst injection was shown in Figure 4.33. It indicates that aquathermolytic technology is effective to improve heavy oil production. The cycle decline rate of heavy oil production in puff-and-huff operation was improved obviously (Wen et al. 2007). Xu and Pu (2018) compared the results of laboratory and field studies of catalytic aquathermolysis of heavy oil in a number of wells in the Shengli Oil field of China. The experiments were carried out at the Kendong wells of the Shengli Oil field. Laboratory studies were allowed to determine the optimal composition of the catalytic solution: an oil-soluble catalyst produced in the laboratory,

4.9 Field Tests

Table 4.20

Field test results of aquathermolytic technology.

Depth/thickness of oil belt (m)

Well no.

Steam injected (tons)

Steam injection cycles

Prior cycle production Time Oil (days) (tons)

Catalytic cycle production Time Oil (days) (tons)

Improved recovery (tons)

Qi40-23-35

819.8/14.2

1700

Seventh

204

758.4 189

1023

264.6

Qi40-17-24

767.9/22.0

2200

Ninth

165

499.7 173

834

334.3

986.9

351.0

Qi40-3-211

1023.4/8.1

1940

Fifth

199

635.9 207

Qi40-15-23

712.7/20.4

2109

Seventh

397

1011.2 360

Qi108-23-15

1277/16.8

2100

Sixth

141

788

144

1022

Qi108-3-17

1146/26.9

2821

Second

122

1184

128

969

−115

Qi108-22-04

1094/20.2

1975

Sixth

210

612.7 191

870.9

258.2

Qi108-8-20

1172/12.0

2570

Ninth

165

668.4 169

857.9

189.5

1123.6 1112.4 234

Source: Adapted from Wen et al. (2007).

Table 4.21

Analysis of oil sampled from Qi40-23-35 well.

Analysis item

Untreated

Viscosity/Pa⋅s, at 50 ∘ C

12.4

Viscosity reduction ratio (%)

78.2

Elemental composition (wt.%)

2.7

C

85.80

83.90

H

11.80

13.31

O

1.26

1.12

S

0.40

0.14

N

0.72

Saturates SARA composition (wt.%)

Treated

0.64

27.4

32.0

Aromatics

31.5

34.1

Resins

34.3

28.4

6.8

5.5

Asphaltenes Source: Adapted from Wen et al. (2007).

Cycle production (tons)

1600 1200 800 400 0

1

2

3

4

5 6 7 8 Cycles Cycle 1-6: no catalyst. Cycle 7-9: have catalyst.

9

Figure 4.33 Cycle production of Qi40-23-35 oil well in puff-and-huff operation. Source: Adapted from Wen et al. (2007).

221

4 Fundamentals of In Situ Upgrading

0.3%; hydrogen donor 0.05%; dispersant 0.5%. The catalyst solution was pumped into the steam supply line after the check valve. Optimal steam parameters were calculated based on geological data: injection volume of 2000 tons per well and a steam feed rate of 7.5 tons per hour. The steam injection process was carried out as follows: preparation of the catalyst solution, injection of 1000 tons of steam to preheat the formation, then injection of 600 tons of steam in a mixture with 9.5 tons of catalyst solution (catalyst feed rate of 0.06 ton/h), and injection of 400 tons of steam as displacement fluid. The well was opened for extraction after aging for 48 hours. Table 4.22 shows the parameters of the extracted oil before and after aquathermolysis for five wells. As a result of underground catalytic aquathermolysis, the cyclic increment of production varied from 376 to 1138 tons, depending on the well, on average – up to 653 tons. The daily production volume at wells GD827-12 and KD192P2 increased significantly: from 1.1 and 4.7 tons to 6.2 and 8.2 tons, respectively. In addition, daily fluid production of two wells improved. This can be explained by the fact that in the process of catalytic aquathermolysis, the bottomhole formation zone was cleared from asphaltene deposits. It can be noted that in the previously closed KD52N5 well, cyclic oil production after aquathermolysis was 489.6 tons (Xu and Pu 2018). Thus, field tests proved that the economic benefit of this technology is remarkable. The viscosity of heavy oil significantly decreased (Table 4.23). The highest preliminary reduction ratio (79.8%) was recorded for the KD192P2 well. After some time, the viscosity of the oil increased somewhat in all the wells, which may be due to the following reasons: first, small molecules formed during the reaction polymerize to form macromolecules; second, the higher temperature in the near-wellbore region of the formation leads to more intensive chemical reaction, thus the viscosity of oil extracted from the near-wellbore region immediately after the reaction of aquathermolysis is lower than that of the deeper layers. On the other hand, the decrease in reservoir temperature over time also causes an increase in the viscosity of heavy oil. Nevertheless, two weeks after the process of The parameters of the extracted oil before and after aquathermolysis.

Increase in production (tons)

Cycle crude outputs (tons)

Water cut (%)

Daily oil production (tons)

Production volume

Cyclon duration (days)

Dryness index (%)

Temperature (∘ C)

Pressure (MPa)

Well

Steam injection quantity per cycle (tons)

Steam injection cycle parameters

Daily fluid production (tons)

Table 4.22

Steam injection cycle

222

KD642P41 Preliminary 2062 19.5 324 707.5 124 12.2 7.4 39.3 917.6 1137.7 work 2103 19.8 345 72.6 221 15.7 9.3 40.8 2055.3 K92N6

Preliminary 2203 15.7 344 65.4 work 2017 15.1 341 65.6

291 14 4.9 65.0 1425.9 376.4 269 15.3 6.7 56.2 1802.3

GD827-12

Preliminary 2082 19.5 343 68.1 work 2148 19.5 348 67.9

453 3.7 1.1 70.3 498.3 172 27.2 6.2 77.2 1066

KD52N5

Preliminary Well shut down work

KD192P2

2253 13.3 331 70.1

Preliminary 2193 14.7 325 69.2 work 2125 14.9 329 68.5

Total Source: Adapted from Xu and Pu (2018).

568.1 489.6

153 22.6 3.2 85.8 489.6 128 9.3 4.7 49.5 601.6 694.0 158 22.6 8.2 63.7 1295.6 3265.8

4.9 Field Tests

Table 4.23 The viscosity reduction degree of heavy oil samples produced before and after the field test. Well

12058 —

10399 —

13573 —

Decrease in viscosity (%)

KD192P2

Viscosity (mPa⋅s)

Decrease in viscosity (%)

KD52N5

Viscosity (mPa⋅s)

Decrease in viscosity (%)

GD827-12

Viscosity (mPa⋅s)

Decrease in viscosity (%)

K92N6

Viscosity (mPa⋅s)

Decrease in viscosity (%)

Time (weeks)

Viscosity (mPa⋅s)

KD642P41

Before reaction

19600 —

9657 —

1

5919

69.8 3423

71.6 2692

74.1 3651

73.1 1951 79.8

2

5780

70.5 3260

72.9 2715

73.9 3366

75.2 2260 76.6

3

5650

71.2 3496

71

2797

73.1 3868

71.5 2385 75.3

4

5710

70.9 3470

71.2 2860

72.5 4099

69.8 2588 73.2

5

5836

70.2 3555

70.5 2954

71.6 4465

67.1 2675 72.3

6

5785

70.5 3536

70.7 3088

70.3 4710

65.3 2714 71.9

8

6155

68.6 3681

69.5 3213

69.1 4764

64.9 3167 67.2

10

6438

67.2 3759

68.8 3348

67.8 4872

64.1 3390 64.9

12

6372

67.5 3846

68.1 3484

66.5 4965

63.4 3592 62.8

14

6453

67.1 3912

67.6 3587

65.5 5013

63.1 3659 62.1

aquathermolysis, the relative decrease in viscosity was greater than 62%. The results prove that the use of catalytic aquathermolysis in field conditions has a long-term effect of reducing the viscosity of heavy oil, which leads to technological advantages and an increase in the profitability of the well. Recently, we have successfully employed the first pilot test of nickel tallate catalyst precursor in Boca de Jaruco, Cuba, with the total mass of 12 ton (Vakhin et al. 2021). The nickel-based catalytic composition in laboratory-scale hydrothermal conditions at 300 ∘ C and 90 bars demonstrated a high performance; the content of asphaltenes was reduced from 22% to 7 wt.%. The viscosity of crude oil was also reduced by three times. The pilot test of catalyst injection was carried out in bituminous carbonate formation M, in Boca de Jaruco Reservoir. The application of catalytic composition provided increase in cumulative oil production and incremental oil recovery in contrast to the previous cycle (without catalyst) is 170% up to date (the effect is in progress). After injection of catalysts, more than 200 samples from production well were analyzed in the laboratory. Based on the physical and chemical properties of investigated samples and considering the excellent oil recovery coefficient, it is decided to expand the industrial application of catalysts in the given reservoir. A schematic diagram of the injection of a catalytic composition into a well is illustrated in Figure 4.34. The catalytic composition is pumped into the productive horizon between the cycles of CSS. The formation of the active form of the catalyst, which is nanodispersed mixed nickel sulfides, occurs as a result of the decomposition of nickel tallate directly in the pay zone of the reservoir during subsequent steam treatment. A column of pipes with a plugged end and holes made at the end section for pumping the precursor solution into the perforation interval of the productive formation is lowered into the well; through the holes of the pipe, the calculated volume of the catalytic

223

4 Fundamentals of In Situ Upgrading

Figure 4.34 Schematic diagram of the injection of a catalytic composition into a well. Source: Adapted from Vakhin et al. (2021).

composition of oil is injected by means of a cementing unit using pumps at a rate of injection into the productive reservoir of 1–5 m3 /h. Through the same holes of the pipe, an organic solvent is injected, with the possibility of washing out the remnants of the previously injected catalyst from the borehole and delivering it to the oil-saturated zone of the formation. Then the well is closed for a period of at least two days in order to ensure maximum effective soaking of the productive reservoir with the previously injected catalyst and solvent. Next, steam is pumped at a temperature from 200 to 350 ∘ C, at a pressure in the reservoir from 3 to 15 MPa. During the heating of the formation, the temperature reaches values from 200 to 300 ∘ C with continued steam injection for at least five days with the possibility of destructive hydrogenation of resins and asphaltenes in the productive formation. Next, the well is stopped for soaking period, after which the fluids are produced from the well. The results of pilot field tests of the catalytic composition show a significant increase in accumulated oil production (Cycle 5) compared to the previous steam injection cycle (Cycle 4) conducted without the use of a catalyst (Figure 4.35). During the initial production period, a 3000

Cumulative oil production (ton)

224

Cycle 4

2500

Cycle 5

2000 1500 1000 500 0

0

50

100

150

200

250

300

350

400

450

Time (Days) Figure 4.35 Partial results of oil production after injection of a nickel-based catalytic composition into the M formation of the Boca de Jaruco reservoir.

1550 300

03.07.19

Viscosity (Pa·s) (at 36˚C)

250

1600

200

15.08.19 150

1550

Catalyst injection

26.11.19

100

50

0 15.05.19 04.07.19

Catalyst injection 26.11.19

09.01.20 1300

23.08.19

12.10.19 01.12.19

20.01.20

10.03.20

13.02.20

500 1000

Figure 4.36

1500

2000

2500

3000

3500

4000

4500 m/z

The viscosity of the produced extra-heavy oil and the average molecular weight of asphaltenes before and after injection of the catalytic composition.

226

4 Fundamentals of In Situ Upgrading

significant decrease in the average molecular weight of asphaltenes (1300 amu) was achieved. During this period, the formation of the active form of the catalyst into the sulfide form and its active participation in the processes of chemical conversion of oil is proposed. The low production rates at the beginning of the cycle are primarily due to the decomposition of the precursor in the induction period, during which the active form of the catalyst is formed. As a result of the catalytic intensification of the chemical conversion of resins and asphaltenes, the viscosity of oil is irreversibly reduced. This contributes to the displacement of additional oil from the pore space, which is reflected in an increase in accumulated oil production and a decrease in the steam-oil-ratio (SOR). For the oil samples extracted before the injection of the catalytic composition, the values of the average molecular weight of asphaltenes vary in the range of 1500–1600 AU (Figure 4.36). The production was in progress at 1 July 2021. The increase in the oil production was 110%. The cycle duration has increased to 550 days. A SOR of less than 3 t/t has been achieved, additional oil production is 1340 tons. In 2020, JSC “VNIIneft” conducted a set of studies on the dynamics of changes in the properties of oil, confirming the qualitative and quantitative change in the properties of oil of cycle 5 in comparison with cycle 4 (Base cycle). According to the results of the analysis, the oil of Cycle 5 (Aquathermolysis) in comparison with Cycle 4 (Base cycle) had lower viscosity by 44% and 38%, respectively, at 36 and 50 ∘ C; reduction of the boiling point from 200 to 167 ∘ C, increase in the proportion of light and medium fractions (up to 300 C) from 30% to 40%. The metal tallates dissolved in a hydrogen donor solvent (Nefras) were injected in Ashal’cha heavy oil reservoir. The pilot well went on through four cycles of alternated steam injection and production. After the third cycle (before catalyst injection), the total oil production was 4849 tons with a total steam injection of 20610 m3 (water equivalent), resulting in a SOR of 4.25. Moreover, the well had a water cut of approximately 90%. These rates were obtained within a total of 967 net days of production and 309 net days of steam injection. Later on, the well was set back to injection mode and then put back down for three weeks. Steam injection was restarted under the planned catalyst injection. A total of 7.5 tons of mixture of catalyst and solvent were injected into the formation, followed by steam injection according the steps presented in Table 4.24. After the catalyst injection and respective well production, updates of field results corresponding to the fourth cycle showed several remarks (Figure 4.37 and Table 4.25): (i) An observed jump from an average of 1.2 to 11.8 T/d of oil production. Higher oil rates were possible to be obtained after the well shut down due to low oil production rates at the end of the third cycle. (ii) Represents the 28% oil production within the 25% of time in comparison to that of the third cycle. The fourth cycle, up to 8 August 2019, produced 544 tons of oil in 143 days, meanwhile, the third Table 4.24

Catalyst injection and production strategy from well 15075.

Data

Base case

25.12.2018 26.12.2018

Steam injection stopped Steam injection

27.12.2018 26.01.2019

Catalyst injection case

Injection of catalyst mixture (7.5 tons of catalyst + hydrogen donor mixture Displacement with steam

End of steam injection

End of steam injection

Soaking

Soaking

Production

Production

27.01.2019 to 10.02.2019 11.02.2019 to 31.12.2019

0

Figure 4.37 11.11.2013 03.12.2013 22.12.2013 10.01.2014 27.01.2014 15.02.2014 06.03.2014 22.03.2014 08.04.2014 27.04.2014 16.05.2014 04.06.2014 23.06.2014 12.07.2014 31.07.2014 19.08.2014 07.09.2014 26.09.2014 15.10.2014 03.11.2014 22.11.2014 10.12.2014 28.12.2014 11.01.2015 25.01.2015 06.02.2015 19.02.2015 04.03.2015 19.03.2015 03.04.2015 22.04.2015 11.05.2015 30.05.2015 18.06.2015 07.07.2015 26.07.2015 14.08.2015 02.09.2015 21.09.2015 10.10.2015 29.10.2015 17.11.2015 06.12.2015 25.12.2015 13.01.2016 01.02.2016 20.02.2016 10.03.2016 30.03.2016 18.04.2016 07.05.2016 26.05.2016 14.06.2016 03.07.2016 22.07.2016 10.08.2016 29.08.2016 17.09.2016 06.10.2016 25.10.2016 13.11.2016 02.12.2016 21.12.2016 12.01.2017 31.01.2017 19.02.2017 14.03.2017 02.04.2017 22.04.2017 12.05.2017 31.05.2017 19.06.2017 08.07.2017 27.07.2017 15.08.2017 04.09.2017 23.09.2017 12.10.2017 02.11.2017 21.11.2017 10.12.2017 31.12.2017 19.01.2018 20.02.2018 11.03.2018 03.06.2018 23.06.2018 12.07.2018 31.07.2018 19.08.2018 07.09.2018 26.09.2018 15.10.2018 03.11.2018 22.11.2018 11.12.2018 30.12.2018 18.01.2019 06.02.2019 25.02.2019 16.03.2019 04.04.2019 23.04.2019 12.05.2019 31.05.2019 19.06.2019 08.07.2019 27.07.2019 15.08.2019 03.09.2019 22.09.2019 11.10.2019 30.10.2019 18.11.2019 07.12.2019 26.12.2019

100

90

80

70

60

50

40

30

20

10

Field history rates, before and after catalyst injection. Catalyst injection, T Water Injection Rate (H), sm3/day Water Rate (H), sm3/day Oil Rate (H), T/d oil rate, T/d, Field results

228

4 Fundamentals of In Situ Upgrading

Table 4.25

Field history of pilot well 15075, before and after catalyst injection. Third cycle

Fourth cycle

Before catalyst injection

After catalyst injection

Parameter

Total production time (days)

571

143

Total oil production (tons)

1901

544

cycle in total produced: 1901 tons of oil within 571 days. It shows that after catalyst injection, oil mobility was improved and recovered, taking into account that the well is under the fourth cycle. (iii) Water cut was reduced from approximately 98% (end of third cycle) to 76% by first week of August 2019. These facts indicate that injection strategy and the in situ oil upgrading were successfully achieved. For the investigated samples, in contrast to the oil produced in 2017, the significant changes are observed in the contents of saturates and aromatic hydrocarbons. The increase in resins content corresponds to the period of oil sample with high viscosity. Then, the content of resins and thereby viscosity of oil samples are decreasing due to activation of catalyst particles. The evidence of chemical processes under the catalyst influence is the increase in the content of saturated hydrocarbons (Table 4.26). This correlates with the viscosity parameter. A complex time dependency was observed according to produced oil viscosity monitoring (Figure 4.38). It is assumed that in the first stage, the injected solvent is of primary importance. After the formation of the active form, oil conversion processes begin. This ensures that previously unaffected zones of the formation with more viscous oil are involved in production. The nanoparticles provide a decrease in the viscosity of the produced oil and an increase in oil recovery factor. The various field studies indicated that certain adjustments of the laboratory conditions are required to achieve a high degree of viscosity reduction under field operations, at usually longer reaction time due to the volume and composition of the heavy oil reservoirs. It could also be established that, most comparative results reported higher viscosity reduction values for the laboratory trials. Therefore, an extensive research is required toward improving the efficiency of the catalytic aquathermolysis process during oilfield applications.

Table 4.26 SARA analysis of the oil after field experiments with catalysts. Objects

Produced oil samples

Group composition (wt.%) S

A

R

Asp

26.3

37.1

28.9

6.0

29.2

34.1

29.4

6.1

30.0

40.2

23.7

5.4

24.6

34.8

34.8

5.3

24.0

32.4

37.1

5.4

36.0

39.1

20.8

5.1

36.1

39.2

20.9

5.1

4.10 Conclusions and Recommendations

70 000

Involving zones with heavier crude oil at 10 ˚C Injecting of catalyst

60 000

Dyamic viscosity (mPa·s)

50 000

40 000

Influence of catalyst Influence of solvent

30 000

at 20 ˚C 20 000

10 000

0

07.09.18

Figure 4.38

27.10.18

28.12.18

04.02.19

26.03.19

15.05.19

04.07.19

Viscosity of the extracted oil samples measured at various date, at 10 and 20 ∘ C.

4.10 Conclusions and Recommendations In this chapter, the main chemical reactions between oil sands and steam have been discussed. The primary sand reaction appears to involve reactive mineral carbonates yielding varying amounts of CO2 which can affect the acidity of the system and the viscosity of the oil by dissolving in it. The most reactive species in the oil/bitumen phase appear to be the organosulfur compounds, the carbon–sulfur bonds which may cleave during aquathermolysis especially if suitable metal ion catalysts are present to coordinate with the sulfur as a ligand. Hydrogen sulfide is produced as a result of hydrodesulfurization of heavy oil and decomposition of some minerals. Addition of metal salts that can form stable sulfides by reaction with H2 S may be a method of controlling H2 S production, but removal of sulfur from the oil by this mechanism, in situ, may be viewed as a useful upgrading step lessening the load on surface desulfurization facilities. Heavy oils/bitumen appears to undergo competitive chemical transformations that affect viscosity in opposite ways during aquathermolysis. Once reactive species have been generated by aquathermolytic cleavage of sensitive sites such as the carbon–sulfur bond, they can either polymerize and thus raise the viscosity or they can initiate a series of further reactions that result in the generation of products (e.g. saturates) which can reduce viscosity even when present in relatively small amounts. The balance between these two opposing effects appears to be dependent on a number of factors including type of oil/bitumen, time of aquathermolysis (steaming), and acidity. One of the most important component reactions of the aquathermolysis process is the WGSR. Initial cleavage of the reactive species (organosulfurs) in the oil or bitumen can trigger a series of reactions that yield carbon monoxide as a result of decarbonylation. This carbon monoxide can react with the high temperature water to yield carbon dioxide and hydrogen and the hydrogen, in turn, can participate in upgrading type hydrocracking and hydrodesulfurization reactions with the oil/bitumen phase. Thus, hydrogen from water is used to upgrade the oil in situ.

229

230

4 Fundamentals of In Situ Upgrading

Emphasis should thus be placed on examining the aquathermolytic behavior of a wider range of heavy oil sand samples and establishing a protocol for predicting the behavior based on laboratory examination of cores. The results of field tests of catalytic aquathermolysis are attractive and promising. Some of the concepts could be tested to some extent on existing pilots by gathering relevant production data over time. It is hoped that this chapter may encourage scholar and petroleum engineers to examine their pilots in a manner that might assist in checking the validity of some of the postulates. Much remains to be done at the molecular level to characterize and understand the chemical processes that occur during aquathermolysis.

References Alemán-Vázquez, L.O., Torres-Mancera, P., Ancheyta, J., and Ramírez-Salgado, J. (2016). Use of hydrogen donors for partial upgrading of heavy petroleum. Energy and Fuels 30: 9050–9060. Aliev, F.A., Mukhamatdinov, I.I., Sitnov, S.A. et al. (2021a). In-situ heavy oil aquathermolysis in the presence of nanodispersed catalysts based on transition metals. Processes 9: 127. Aliev, F.A., Kiekbaev, A.A., Andreev, D.V. et al. (2021b). The effect of clay minerals on conversion of yarega heavy oil during catalytic aquathermolysis processes. SOCAR Proc 2: 41–47;Al-Muntaser, A.A., Yuan, C., Suwaid, M.A., and Varfolomeev, M.A. (2018). Coke formation through oil upgrading of heavy oil by superheated steam. International Multidisciplinary Scientific GeoConference. SGEM Surveying Geology and Mining Ecology Management 18: 335–341. Al-Muntaser, A.A., Varfolomeev, M.A., Suwaid, M.A. et al. (2020). Hydrothermal upgrading of heavy oil in the presence of water at sub-critical, near-critical and supercritical conditions. Journal of Petroleum Science and Engineering 184: 106592. Al-Muntaser, A.A., Varfolomeev, M.A., Suwaid, M.A. et al. (2021). Hydrogen donating capacity of water in catalytic and non-catalytic aquathermolysis of extra-heavy oil: deuterium tracing study. Fuel 283: 118957. Al-Otoom, A.Y. (2005). A thermodynamic study on the utilization of Jordanian oil shale in the cement industry. Oil shale 22 (1): 37–49. Ancheyta, J. and Speight, J.G. (2007). Hydroprocessing of Heavy Oils and Residua. CRC Press. Ancheyta, J., Trejo, F., and Rana, M.S. (2010). Asphaltenes: Chemical Transformation During Hydroprocessing of Heavy Oils. CRC Press. Arcelus-Arrillaga, P., Pinilla, J.L., Hellgardt, K., and Millan, M. (2017). Application of water in hydrothermal conditions for upgrading heavy oils: a review. Energy and Fuels 31: 4571–4587. Armas, J.D., Garcia-Vila, A., Ortega, L.C. et al. (2021). In-situ upgrading in a dolomite porous medium: a kinetic model comparison and nanocatalyst deposition study. Journal of Petroleum Science and Engineering 205: 108799. Belgrave, J.D.M., Moore, R.G., and Ursenbach, M.G. (1994). Gas evolution from the aquathermolysis of heavy oils. Canadian Journal of Chemical Engineering 72: 511–516. Belgrave, J.D.M., Moore, R.G., and Ursenbach, M.G. (1997). Comprehensive kinetic models for the aquathermolysis of heavy oils. Journal of Canadian Petroleum Technology 36: 38–44. Belov, N.I., Gorbunov, Y.I., Ivanov, V.I. et al. (1987). Guidelines for the study of the stress-strain state of rocks at various stages of the exploration process. Advanced Materials Research 236: 839–843. Can𝚤az, R.O. and Erkey, C. (2014). Process intensification for heavy oil upgrading using supercritical water. Chemical Engineering Research and Design 92: 1845–1863. Chen, W.-H. and Chen, C.-Y. (2020). Water gas shift reaction for hydrogen production and carbon dioxide capture: a review. Applied Energy 258: 114078.

References

Chen, Y., Wang, Y., Wu, C., and Xia, F. (2008). Laboratory experiments and field tests of an amphiphilic metallic chelate for catalytic aquathermolysis of heavy oil. Energy & Fuels 22 (3): 1502–1508. Chen, Y., Wang, Y., Lu, J., and Wu, C. (2009). The viscosity reduction of nano-keggin-K3PMo12O40 in catalytic aquathermolysis of heavy oil. Fuel 88: 1426–1434. Chen, Q.Y., Liu, Y.J., and Zhao, J. (2011). Intensified viscosity reduction of heavy oil by using reservoir minerals and chemical agents in aquathermolysis. Advanced Materials Research. Trans Tech Publications 236: 839–843. Cheng, Z.M., Ding, Y., Zhao, L.Q. et al. (2009). Effects of supercritical water in vacuum residue upgrading. Energy and Fuels 23: 3178–3183. Clark, P.D. and Hyne, J.B. (1984). Steam-oil chemical reactions: mechanisms for the aquathermolysis of heavy oils. AOSTRA Journal of Research 1: 15–20. Clark, P.D. and Hyne, J.B. (1990). Studies on the chemical reactions of heavy oils under steam stimulation condition. AOSTRA Journal of Research 29: 29–39. Clark, P.D., Hyne, J.B., and Tyrer, J.D. (1983). Chemistry of organosulphur compound types occurring in heavy oil sands: 1. High temperature hydrolysis and thermolysis of tetrahydrothiophene in relation to steam stimulation processes. Fuel 62: 959–962. Cuijpers, M.C.M., Boot, M.D., Deen, N.G., and Golombok, M. (2019). Sulphur and viscosity reductions in heavy hydrocarbons by subcritical water processing. Journal of Petroleum Science and Engineering 172: 1069–1076. Dmitrievsky, A.N., Balanyuk, I.E., Karakin, A.V. et al. (2003). New fluid dynamic model for formation of giant hydrocarbon deposits. Journal of Geochemical Exploration 78: 345–347. Dobrynkin, N.M., Batygina, M.V., and Noskov, A.S. (2017). Studies of catalytic properties of inorganic rock matrices in redox reactions. Journal of Sustainable Development of Energy, Water and Environment Systems 5 (3): 408–416. Dobrynkin, N.M., Batygina, M.V., and Noskov, A.S. (2018). Catalytic redox transformations in rock matrices. Catalysis in Industry 10: 91–96. Eletskii, P.M., Sosnina, G.A., Zaikinaa, O.O. et al. (2017). Heavy oil upgrading in the presence of water. Journal of Siberian Federal University “Chemistry” 4: 545–572. Elliott, D.C., Hart, T.R., Schmidt, A.J. et al. (2013). Process development for hydrothermal liquefaction of algae feedstocks in a continuous-flow reactor. Algal Research 2: 445–454. Fan, H. (2003). The effects of reservoir minerals on the composition changes of heavy oil during steam stimulation. Journal of Canadian Petroleum Technology 42 (3). Fan, H.-F., Liu, Y.-J., and Zhong, L.-G. (2001). Studies on the synergetic effects of mineral and steam on the composition changes of heavy oils. Energy and Fuels 15: 1475–1479. Frolov, S.V., Akhmanov, G.G., Bakay, E.A. et al. (2015). Meso-neoproterozoic petroleum systems of the Eastern Siberian sedimentary basins. Precambrian Research 259: 95–113. Gai, X.-K., Arano, H., Lu, P. et al. (2016). Catalytic bitumen cracking in sub-and supercritical water. Fuel Processing Technology 142: 315–318. Giles, R., Ahn, G., and Jung, K.W. (2015). H-D exchange in deuterated trifluoroacetic acid via ligand-directed NHC-palladium catalysis: a powerful method for deuteration of aromatic ketones, amides, and amino acids. Tetrahedron Letters 56: 6231–6235. Gore, R.C. (1954). Infrared spectroscopy. Analytical Chemistry 26: 11–19. Gu, Z., Chang, N., Hou, X. et al. (2012). Experimental study on the coal tar hydrocracking process in supercritical solvents. Fuel 91: 33–39. Guo, K., Li, H., and Yu, Z. (2016). In-situ heavy and extra-heavy oil recovery: a review. Fuel 185: 886–902. Gushchin, V.A. and Ryl’kov, A.V. (1987). Geochemical features of two-phase hydrocarbon-pool formation in northern west Siberia. International Geology Review 29: 1113–1116.

231

232

4 Fundamentals of In Situ Upgrading

Hamedi Shokrlu, Y. and Babadagli, T. (2014). Kinetics of the in-situ upgrading of heavy oil by nickel nanoparticle catalysts and its effect on cyclic-steam-stimulation recovery factor. SPE Reservoir Evaluation and Engineering 17: 355–364. Han, L., Zhang, R., Bi, J., and Cheng, L. (2011). Pyrolysis of coal-tar asphaltene in supercritical water. Journal of Analytical and Applied Pyrolysis 91: 281–287. Hart, A. (2014). Advanced studies of catalytic upgrading of heavy oils (Doctoral dissertation, University of Birmingham). Hart, A., Adam, M., Robinson, J.P. et al. (2020). Tetralin and decalin H-donor effect on catalytic upgrading of heavy oil inductively heated with steel balls. Catalysts 10: 393. Hasan, M.R. and Reza, M.T. (2019). Hydrothermal deformation of Marcellus shale: effects of subcritical water temperature and holding time on shale porosity and surface morphology. Journal of Petroleum Science and Engineering 172: 383–390. Hasan, S.W., Ghannam, M.T., and Esmail, N. (2010). Heavy crude oil viscosity reduction and rheology for pipeline transportation. Fuel 89: 1095–1100. Hashemi, R., Nassar, N.N., and Pereira Almao, P. (2014). Nanoparticle technology for heavy oil in-situ upgrading and recovery enhancement: opportunities and challenges. Applied Energy 133: 374–387. He, C., Chen, C.-L., Giannis, A. et al. (2014). Hydrothermal gasification of sewage sludge and model compounds for renewable hydrogen production: a review. Renewable and Sustainable Energy Reviews 39: 1127–1142. Hoering, T.C. (1982). Thermal reactions of kerogen with added water, heavy water, and pure organic substances. Carnegie Institution of Washington Year B 81: 397–402. Hongfu, F., Yongjian, L., Liying, Z., and Xiaofei, Z. (2002). The study on composition changes of heavy oils during steam stimulation processes. Fuel 81: 1733–1738. Hosseinpour, M., Fatemi, S., and Ahmadi, S.J. (2016). Deuterium tracing study of unsaturated aliphatics hydrogenation by supercritical water in upgrading heavy oil. Part II: hydrogen donating capacity of water in the presence of iron(III) oxide nanocatalyst. Journal of Supercritical Fluids 110: 75–82. Hosseinpour, M., Fatemi, S., Ahmadi, S.J. et al. (2018a). The synergistic effect between supercritical water and redox properties of iron oxide nanoparticles for in-situ catalytic upgrading heavy oil with formic acid. Isotopic study. Applied Catalysis B: Environmental 230: 91–101. Hosseinpour, M., Fatemi, S., Ahmadi, S.J. et al. (2018b). Isotope tracing study on hydrogen donating capability of supercritical water assisted by formic acid to upgrade heavy oil: computer simulation vs. experiment. Fuel 225: 161–173. Hyne, J.B. (1986). Aquathermolysis: A Synopsis of Work on the Chemical Reaction Between Water (Steam) and Heavy Oil Sands During Simulated Steam Stimulation. AOSTRA Publication Series 55, [27] leaves. Hyne, J.B., Clark, P., Clarke, J.K. et al. (1982). Aquathermolysis heavy oils. In: Second International Conference on Heavy Crudes and Tar Sands, Caracas, 25. Hyne, J.B., Greidanus, J.W., Tyrer, J.D. et al. (1982a). Aquathermolysis of heavy oils. Revista Tecnica INTEVEP 2: 87–94. Hyne, J.B., Greidanus, J.W., Tyrer, J.D. et al. (1982b). The Second International Conference on Heavy Crude and Tar Sands, 1. Venez: Caracas. Ilyina, V.I. (1986). Subdivision and correlation of the marine and non-marine Jurassic sediments in Siberia based on palynological evidence. Review of Palaeobotany and Palynology 48: 357–364. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2015). A review of pyrolysis, aquathermolysis, and oxidation of Athabasca bitumen. Fuel Processing Technology 131: 270–289. Katritzky, A.R., Allin, S.M., and Siskin, M. (1996). Aquathermolysis: reactions of organic compounds with superheated water. Accounts of Chemical Research 29 (8): 399–406.

References

Kayukova, G.P., Uspensky, B.V., Abdrafikova, I.M., and Musin, R.Z. (2016). Characteristic features of the hydrocarbon composition of Spiridonovskoe (Tatarstan) and Pitch Lake (Trinidad and Tobago) asphaltites. Petroleum Chemistry 56: 572–579. Kayukova, G.P., Mikhailova, A.N., Kosachev, I.P. et al. (2018). Conversion of heavy oil with different chemical compositions under catalytic aquathermolysis with an amphiphilic Fe–Co–Cu catalyst and kaolin. Energy and Fuels 32: 6488–6497. Kazakov, M.O., Klimov, O.V., Dik, P.P. et al. (2017). Hydroconversion of oil shale on natural mineral matrices. Petroleum Chemistry 57: 1169–1172. Khelkhal, M.A., Eskin, A.A., Nurgaliev, D.K., and Vakhin, A.V. (2019). Thermal study on stabilizing the combustion front via bimetallic Mn@Cu tallates during heavy oil oxidation. Energy and Fuels 34: 5121–5127. Kostyukevich, Y., Zherebker, A., Vlaskin, M.S. et al. (2019). Speciation of structural fragments in crude oil by means of isotope exchange in near-critical water and Fourier transform mass spectrometry. Analytical and Bioanalytical Chemistry 411: 3331–3339. Kruse, A. (2008). Supercritical water gasification. Biofuels, Bioproducts and Biorefining Innovation a Sustainable Economy 2: 415–437. Kruse, A. and Dinjus, E. (2007). Hot compressed water as reaction medium and reactant: properties and synthesis reactions. Journal of Supercritical Fluids 39: 362–380. Leif, R.N. and Simoneit, B.R.T. (2000). The role of alkenes produced during hydrous pyrolysis of a shale. Organic Geochemistry 31: 1189–1208. León, A.Y., Rodríguez, N.A., Mejía, E., and Cabanzo, R. (2016). Prediction of the quality of coke obtained from vacuum residues by using spectroscopy infrared FTIR-ART. In: Journal of Physics: Conference Series. IOP Publishing, 12092. Li, G.R., Chen, Y., An, Y., and Chen, Y.L. (2016). Catalytic aquathermolysis of super-heavy oil: cleavage of [Formula presented] bonds and separation of light organosulfurs. Fuel Processing Technology 153: 94–100. Liu, Y. and Fan, H. (2002). The effect of hydrogen donor additive on the viscosity of heavy oil during steam stimulation. Energy and Fuels 16: 842–846. Liu, Y., Zhong, L., Jiang, S. et al. (2004). Research progress of recovering heavy oil by aquathermolysis. Journal of Fuel Chemistry and Technology 32: 117–122. Liu, Y., Bai, F., Zhu, C.-C. et al. (2013). Upgrading of residual oil in sub-and supercritical water: an experimental study. Fuel Processing Technology 106: 281–288. Long, M.A., Garnett, J.L., Vining, R.F.W., and Mole, T. (1972). A new simple method for rapid tritium labeling of organics using organoaluminum dihalide catalysts. Journal of the American Chemical Society 94: 8632–8633. Lysogorskiy, Y.V., Aminova, R.M., and Tayurskii, D.A. (2016). Initial steps in reactions of aquathermolysis of cyclohexyl phenyl sulfide by means of ab initio calculations. Computational & Theoretical Chemistry 1078: 138–145. Maity, S.K., Ancheyta, J., and Marroquín, G. (2010). Catalytic aquathermolysis used for viscosity reduction of heavy crude oils: a review. Energy and Fuels 24: 2809–2816. Monin, J.C. and Audibert, A. (1987). Thermal cracking of heavy oil/mineral matrix systems. In: SPE International Symposium on Oilfield Chemistry. OnePetro. Mukhamatdinov, I.I., Salih, I.S.S., Khelkhal, M.A., and Vakhin, A.V. (2020). Application of aromatic and industrial solvents for enhancing heavy oil recovery from the Ashalcha field. Energy and Fuels 35: 374–385. Mukhamatdinov, I.I., Khaidarova, A.R., Mukhamatdinova, R.E. et al. (2022). Development of a catalyst based on mixed iron oxides for intensification the production of heavy hydrocarbon feedstocks. Fuel 312: 123005.

233

234

4 Fundamentals of In Situ Upgrading

Muraza, O. and Galadima, A. (2015). Aquathermolysis of heavy oil: a review and perspective on catalyst development. Fuel 157: 219–231. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2011). Application of nanotechnology for heavy oil upgrading: catalytic steam gasification/cracking of asphaltenes. Energy and Fuels 25: 1566–1570. Ovalles, C., Hamana, A., Rojas, I., and Bolívar, R.A. (1995). Upgrading of extra-heavy crude oil by direct use of methane in the presence of water: deuterium-labelled experiments and mechanistic considerations. Fuel 74 (8): 1162–1168. Pahlavan, H. and Rafiqul, I. (1995). Laboratory simulation of geochemical changes of heavy crude oils during thermal oil recovery. Journal of Petroleum Science and Engineering 12: 219–231. Permanyer, A., Douifi, L., Lahcini, A. et al. (2002). FTIR and SUVF spectroscopy applied to reservoir compartmentalization: a comparative study with gas chromatography fingerprints results. Fuel 81: 861–866. Petrov, S.M., Abdelsalam, Y.I.I., Vakhin, A.V. et al. (2015). Study of the rheological properties of heat-treatment products of asphaltic oils in the presence of rock-forming minerals. Chemistry and Technology of Fuels and Oils 51: 133–139. Petrov, S.M., Ibragimova, D.A., Vakhin, A.V. et al. (2018). Investigating the structure and composition of heavy oil under thermal-catalytic treatment in presence of carbonaceous minerals (Russian). Oil Industry Journal 2018 (02): 44–47. Prakoso, A.A., Punase, A.D., and Hascakir, B. (2017). A mechanistic understanding of asphaltenes precipitation from varying-saturate-concentration perspectives. SPE Production & Operations 32: 86–98. Rahimi, P.M. and Gentzis, T. (2006). The chemistry of bitumen and heavy oil processing. In: Practical Advances in Petroleum Processing, 597–634. Alberta, Canada: Springer. Rakhmatullin, I.Z., Efimov, S.V., Margulis, B.Y., and Klochkov, V.V. (2017). Qualitative and quantitative analysis of oil samples extracted from some Bashkortostan and Tatarstan oilfields based on NMR spectroscopy data. Journal of Petroleum Science and Engineering 156: 12–18. Rakhmatullin, I.Z., Efimov, S.V., Tyurin, V.A. et al. (2018). Application of high resolution NMR (1 H and 13 C) and FTIR spectroscopy for characterization of light and heavy crude oils. Journal of Petroleum Science and Engineering 168: 256–262. Rana, M.S., Sámano, V., Ancheyta, J., and Diaz, J.A.I. (2007). A review of recent advances on process technologies for upgrading of heavy oils and residua. Fuel 86: 1216–1231. Remón, J., Arauzo, J., García, L. et al. (2016). Bio-oil upgrading in supercritical water using Ni–Co catalysts supported on carbon nanofibres. Fuel Processing Technology 154: 178–187. Rodriguez-DeVecchis, V.M., Ortega, L.C., Scott, C.E., and Pereira-Almao, P. (2017). Thermal upgrading of Athabasca bitumen in porous media: limitations and kinetic modelling. Fuel 208: 566–575. Rogel, E., Ovalles, C., Pradhan, A. et al. (2013). Sediment formation in residue hydroconversion processes and its correlation to asphaltene behavior. Energy and Fuels 27: 6587–6593. Rudmin, M., Wilson, M.J., Wilson, L. et al. (2020). Geochemical and mineralogical features of the substrates of the Vasyugan Mire, Western Siberia, Russia. Catena 194: 104781. Sanchez-Minero, F., Ancheyta, J., Silva-Oliver, G., and Flores-Valle, S. (2013). Predicting SARA composition of crude oil by means of NMR. Fuel 110: 318–321. Sato, T., Mori, S., Watanabe, M. et al. (2010). Upgrading of bitumen with formic acid in supercritical water. Journal of Supercritical Fluids 55: 232–240. Shah, A.A., Fishwick, R.P., Leeke, G.A. et al. (2011). Experimental optimization of catalytic process in situ for heavy-oil and bitumen upgrading. Journal of Canadian Petroleum Technology 50: 33–47. Shiraishi, T., Sasaki, M., and Goto, M. (2011). Non-catalytic liquefaction of bitumen with hydrothermal/solvothermal process. Journal of Supercritical Fluids 60: 127–136.

References

Shokrlu, Y.H. and Babadagli, T. (2013). In-situ upgrading of heavy oil/bitumen during steam injection by use of metal nanoparticles: a study on in-situ catalysis and catalyst transportation. SPE Reservoir Evaluation and Engineering 16: 333–344. Shvartsev, S.L., Silkina, T.N., Zhukovskaya, E.A., and Trushkin, V.V. (2003). Underground waters of petroliferous deposits of the nyurol’ka sedimentary basin (Tomsk Region). Geologiya i Geofizika 44: 451–464. Sitnov, S.A., Vakhin, A.V., Mukhamatdinov, I.I. et al. (2019). Effects of calcite and dolomite on conversion of heavy oil under subcritical condition. Petroleum Science and Technology 37: 687–693. Sitnov, S., Mukhamatdinov, I., Aliev, F. et al. (2020). Heavy oil aquathermolysis in the presence of rock-forming minerals and iron oxide(II, III) nanoparticles. Petroleum Science and Technology 38 (6): 574–579. Sitnov, S.A., Mukhamatdinov, I.I., Feoktistov, D.A. et al. (2021). Underground upgrading of the heavy crude oil in content-saturated sandstone with aquathermolysis in the presence of an iron based catalyst. Catalysts 11: 1255. Sitnov, S.A., Khelkhal, M.A., Mukhamatdinov, I.I. et al. (2022). Iron oxide nanoparticles impact on improving reservoir rock minerals catalytic effect on heavy oil aquathermolysis. Fuel 327: 124956. Song, G., Zhou, T., Cheng, L. et al. (2009). Aquathermolysis of conventional heavy oil with superheated steam. Petroleum Science 6: 289–293. Stennikov, A.V., Bugaev, I.A., Kalmykov, A.G. et al. (2018). An experimental study of the oil production from Domanik formation rocks under hydrothermal conditions. Moscow University Geology Bulletin 73: 60–65. Stuart, B.H. (2004). Polymers. In: Infrared Spectroscopy. Fundamentals and Applications, 11. England: John Wiley and sons Ltd. Stupakova, A.V., Khvedchuk, I.I., and Sautkin, R.S. (2019). Reformimng of deposits in ancient oil and gas basins (on the example of the Baikit anteclise eastern slope of the Siberian Platform). Georesources 21: 31–41. Suwaid, M.A., Varfolomeev, M.A., Al-muntaser, A.A. et al. (2020). In-situ catalytic upgrading of heavy oil using oil-soluble transition metal-based catalysts. Fuel 281: 118753. Tan, L., Erdenebaatar, O., Liu, G. et al. (2017). Catalytic cracking of 4-(1-naphthylmethyl) bibenzyl in sub-and supercritical water. Fuel Processing Technology 160: 34–38. Tavakkoli Osgouei, Y. and Parlaktuna, M. (2018). Effects of minerals on steam distillation during thermal heavy-oil recovery: an experimental investigation. Energy Sources Part A Recovery Utilization and Environmental Effects 40: 662–672. Thimm, H.F. (2007). Prediction of hydrogen sulphide production in SAGD projects. In: Canadian International Petroleum Conference 2007, CIPC 2007, 47. https://doi.org/10.2118/2007-094. Timko, M.T., Ghoniem, A.F., and Green, W.H. (2015). Upgrading and desulfurization of heavy oils by supercritical water. Journal of Supercritical Fluids 96: 114–123. Trujillo Portillo, M.L. (2017). Influence of rock matrix in kinetics oxidation for heavy oils. CT&F-Ciencia Tecnologia y Futuro 7: 43–58. Ustinova, V.N., Vyltsan, I.A., Zhilina, E.N., and Mishenina, L.N. (2000). New formed minerals in oil and gas bearing sections and oil-containing technogenic objects. Mineralogy of Technogenesis 1: 106–126. Vakhin, A.V., Kayukova, G.P., Mikhailova, A.M. et al. (2018a). Conversion of heavy oil with different chemical compositions under catalytic aquathermolysis with an amphiphilic Fe–Co–Cu catalyst and kaolin. Energy and Fuels 32 (6): 6488–6497. Vakhin, A.V., Mukhamatdinov, I.I., Aliev, F.A. et al. (2018b). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: part II–changes in composition of aromatic hydrocarbons. Petroleum Science and Technology.

235

236

4 Fundamentals of In Situ Upgrading

Vakhin, A.V., Onishchenko, Y.V., Chemodanov, A.E. et al. (2019). The composition of aromatic destruction products of Domanic shale kerogen after aquathermolysis. Petroleum Science and Technology 37: 390–395. Vakhin, A.V., Aliev, F.A., Mukhamatdinov, I.I. et al. (2021). Extra-heavy oil aquathermolysis using nickel-based catalyst: Some aspects of in-situ transformation of catalyst precursor. Catalysts 11 (2): 189. Volokitin, Y., Sakhibgareev, R., Shaymardanov, M., and Nurieva, O. (2012). Chemical and analytical work in support of West Salym field enhanced oil recovery project (ASP). In: SPE Russian Oil and Gas Exploration and Production Technical Conference and Exhibition. OnePetro. SPE-162067-MS. Weissman, J.G. and Kessler, R.V. (1996). Downhole heavy crude oil hydroprocessing. Applied Catalysis A: General 140: 1–16. Wen, S., Zhao, Y., Liu, Y., & Hu, S. (2007). A study on catalytic aquathermolysis of heavy crude oil during steam stimulation. In International symposium on oilfield chemistry. OnePetro. Xie, L.-F., Wang, F., Zhai, L.-L. et al. (2019). A sequenced batch process for integrated hydropyrolysis and hydrotreatment of a microalgae and used engine oil blend. Fuel Processing Technology 190: 47–54. Xu, H. and Pu, C. (2018). Mechanism of underground heavy oil catalytic aquathermolysis. Chemistry and Technology of Fuels and Oils 53: 913–921. Yan, T., Xu, J., Wang, L. et al. (2015). A review of upgrading heavy oils with supercritical fluids. RSC Advances 5: 75129–75140. Yapaskurt, O.V. (2014). A new typification scheme of postsedimentation transformations of terrigenous deposits of continents and their margins with allowance for geodynamic factors. Moscow University Geology Bulletin 69: 229–235. Yi, S., Babadagli, T., and Li, H.A. (2018). Use of nickel nanoparticles for promoting aquathermolysis reaction during cyclic steam stimulation. SPE Journal 23: 145–156. Yuan, P.Q., Zhu, C.C., Liu, Y. et al. (2011). Solvation of hydrocarbon radicals in sub-CW and SCW: an ab initio MD study. Journal of Supercritical Fluids 58: 93–98. Zhang, X., Ying-Cai, F., Yong-Jian, L., and Hong-Chang, C. (2010). Effects of reservoir minerals and chemical agents on aquathermolysis of heavy oil during steam injection. China Petroleum Processing & Petrochemical Technology 12: 25. Zhao, Y., Lu, W., Chen, J. et al. (2014). Research progress on hydrothermal dissolution and hydrolysis of lignocellulose and lignocellulosic waste. Frontiers of Environmental Science & Engineering 8: 151–161. Zheng, R., Pan, J., Chen, L. et al. (2018). Catalytic effects of montmorillonite on coke formation during thermal conversion of heavy oil. Energy and Fuels 32: 6737–6745. Zhou, Z., Slaný, M., Kuzielová, E. et al. (2022). Influence of reservoir minerals and ethanol on catalytic aquathermolysis of heavy oil. Fuel 307: 121871.

237

5 Catalyst for In Situ Upgrading of Heavy Oils Persi Schacht, Pablo Torres-Mancera, and Jorge Ancheyta Instituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, Mexico City, 07730, Mexico

5.1

Introduction

The catalytic upgrading of heavy and extra-heavy crude oil refers to the chemical treatment aimed at producing changes in their physical and chemical properties so that the heavy crude oil can be effectively transformed into lighter oil by a catalytic system. For application to improving in situ oil recovery, it is necessary to incorporate two fundamental components into the reservoir: one is the source of hydrogen, hydrogen gas, or a hydrogen donor, and the other one is a catalyst. The usage of different hydrogen donors has been reported, among those in a liquid state, tetraline, decalin, naphthalene, pyrene, and aromatics-rich streams stand out (Schacht et al. 2015; Omar et al. 2020). Likewise theoretical and experimental evidence of their effectiveness as hydrogen donors in the catalytic aquathermolysis process has been disclosed by Johnstone et al. (1985) and Alemán-Vázquez et al. (2016). Another hydrogen donor tested is methane, where methyl radicals are known to be sources of hydrogen, and these are the smallest radicals found and can activate free radical addition reactions (Clark et al. 1990; Strausz et al. 1999; Hao et al. 2015). In fact, the methane contained in natural gas has the highest H/C ratio among hydrocarbons, making it an important hydrogen-donor alternative under mild operating conditions to expensive hydrogen (Ovalles et al. 1995; Guo et al. 2016; He and Song 2017). Catalyst is the other fundamental component for in situ upgrading and extraction of heavy and extra-heavy crude oils. Catalyst for this purpose has been reported and classified as follows: water-soluble, oil-soluble, mineral, ionic, and dispersed liquids. The application of different hydrogen donors in combination with catalysts or catalytic complexes provides a synergistic effect to viscosity reduction of in situ heavy crude oil upgrading, which has been evidenced by different experiments carried out in batch reactors and in pilot and field tests, obtaining a higher content of light hydrocarbons, with the consequent decrease in heavy fractions, as well as in the sulfur content. Preliminary studies of water-soluble salts of transition metals have been reported with catalyst such as FeSO4 , RuCl3 , NiSO4 , and VOSO4 (Clark et al. 1990). It was proposed that water-soluble transition metal salts can accelerate the cracking of sulfur-containing components of heavy crude into alkanes, carbon dioxide, hydrogen, and hydrogen sulfide which reduces the viscosity of heavy oil. Afterward water-soluble, nonorganic metal salts have been widely used as catalysts in a steam injection process. The contact of the surface of the catalysts with the oil phase is a crucial point to provide a high efficiency of the reagents. In this sense, developments of homogeneous catalysts Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

238

5 Catalyst for In Situ Upgrading of Heavy Oils

capable of interacting directly with the interface of the water–oil-rock system to improve the quality of heavy oil have been achieved. Obviously, the usage of an appropriate catalyst is the key to the aquathermolysis reaction. Originally, researchers mainly used salts of transition metal ions, transition metal compounds, and organometallic compounds for this technology, and later superoxides began to be used. First, Strausz et al. (1999) used HF BF3 to upgrade Alberta heavy crudes. Then, a series of solid fluorinated acid was used as a catalyst for the preparation of hydrocarbon resins. Many research efforts are aimed at reducing the viscosity of heavy crude oils (Hao et al. 2015; Moein et al. 2020), specifically by the recovery method with metallic nanoparticles (Al-Marshed et al. 2015; Scott et al. 2019) because nanotechnology has a promising future for the recovery and improvement of heavy and extra-heavy crude oils, in the field, compared with other recovery methods. It is also suggested that there is the need for development of catalytic cracking catalysts of low-cost, high activity, and high selectivity along with an optimal process of catalyst synthesis (Zhao et al. 2021). This chapter focused on reviewing various types of catalysts including water-soluble catalysts, oil-soluble catalysts, mineral catalysts, and ionic liquids. Also, their synthesis methods, general characterization, and applications to the upgrading of crude oil are discussed. Nanoparticles are not included here since Chapter 6 is devoted to these types of catalysts.

5.2

General Aspects of Homogeneous Catalysts

The most important characteristic of homogeneous catalysts is that they are made up of transition metals and are highly selective. Homogeneously catalyzed reactions are mainly controlled by kinetics due to the transport of reactants to the catalyst occurs easily because there is no porous system as in the case of heterogeneous catalysts. Therefore, these catalysts do not present restrictions due to diffusional processes reagents-catalyst. Furthermore, since the active sites are well defined, most of the mechanisms of homogeneous catalysis have been elucidated, and the scrutiny of the mechanistic aspects is easy to carry out under reaction conditions by means of spectroscopic methods. On the other hand, it is well known that in catalytic action unstable species can be formed due to the contact between a hydrocarbon molecule and one of the active centers of the catalyst. The reaction products derived from the catalytic process, as well as the reaction rate, depend largely on the nature and reactivity of the reaction intermediates. Furthermore, in the case of hydrotreating catalysts for the upgrading of heavy crude oils, two fundamental components are normally required: the metal function and the acid function, resulting in the so-called “bifunctional catalysts,” where the relationship between the metal and the acid centers must be properly balanced since it plays a preponderant role in the performance of the catalyst. In bifunctional catalysts, the relationship between the metal and the acid centers decisively influences their behavior; the control step of the reaction corresponds to the transformation of the superficial carbenium ions in the acid centers (Hart 2014; Gutierrez-Acebo et al. 2018). This process is not simple since it involves a great variety of reactions involving the breaking of C—C bonds, which requires a relatively large amount of energy. When metallic and acidic functions operate in concert bifunctional catalytic mechanism improved performance in the heavy oil upgrading process is observed. For example, to break down into lighter molecules at the acid sites of the catalyst, a polyaromatic species must be partially hydrogenated. In the absence of hydrogen, said polyaromatic compound is a coke precursor, while coke corresponds to a complex compound that deactivates the catalyst. When a partially hydrogenated polyaromatic adsorbs on acid sites, it can produce carbenium ions that are susceptible to conversion to less-complex polyaromatics,

5.3 Water-Soluble Catalysts

partially saturated cyclic hydrocarbons, and acyclic species. Therefore, the presence of Brønstedand Lewis-type acid sites is important not only in the conversion of asphaltene molecules consisting of a large number of polyaromatic molecules but also in the transformation of S- and N-containing heteroatoms. This ultimately produces a reduction in viscosity, a partial elimination of sulfur and nitrogen, an increase in the amount of saturated compounds, and a reduction in the content of resins and asphaltenes (García-Dávila et al. 2018; and Barbaro et al. 2012). The processing of heavy hydrocarbons requires catalysts of the bifunctional type, that is, owing properties for cracking and hydrogenation reactions. The catalytic cracking activity is provided by the acid function, while the hydrogenation activity is provided by the metals. The acid function of the catalyst can be provided by mineral acids such as hydrochloric, sulfuric, chromic, nitric phosphoric, among others. However, these acids are very corrosive, so they are used in smaller proportions, with respect to the content of the metal that fulfills the function of hydrogenation. Therefore, the catalyst used for the heavy crude oil upgrading process is very different from that used in the normal hydrotreating during refining process. Furthermore, heavy crude oil that is extracted from different geographical areas has very different properties. For this reason, when designing a catalyst, it is important to emphasize that it has the appropriate specifications for the hydrocracking of heavy crude oil. Therefore, the key challenge is to design and synthesize catalysts that have a high activity for in situ hydroprocessing, considering the following aspects: – – – – – – – – – – – –

Low cost High activity High selectivity They are difficult to recover Catalyst dispersion in the oil phase Catalyst dosage Reservoir conditions Coke generation Degree of salinity Formation damage Type of deposit Altered wettability

5.3

Water-Soluble Catalysts

Water-soluble catalysts were the first used to crack crude oil, which is since hot water is cheap, safe, and is considered as a universal solvent. Regarding the organic components of crude oil, those containing C—S bonds have the lowest average bond energy (260 kJ/mol), allowing their breaking into smaller molecules. In fact, when the aquathermolysis reaction occurs, it releases the molecular hydrogen that helps break the C—S bond and ultimately leads to the viscosity-reducing effect of crude oil. These types of catalysts are formulated from transition metals (Al, Ca, Cu, Co, Fe, Mo, Mn, Ni, V, W, Zr, Zn, mainly) and are commonly synthesized from metal salts (Zhong et al. 2003; Li et al. 2013; Schacht et al. 2019; Duan et al. 2020). It has been observed that these water-soluble salts have no significant effect on coke generation. As example, Figure 5.1 shows the effect of metals such as Cu, Fe, Ni, or Zr in reducing viscosity. The water-soluble metal catalyst is cheap and easy to synthesize; however, its dispersion effect in the aqueous medium is unfavorable compared with other types of catalysts. There are also factors

239

5 Catalyst for In Situ Upgrading of Heavy Oils

100 80

Viscosity (cSt)

240

60 40

20 0

Feed

Fe

Cu

Ni

Zr

Water-soluble catalyst Figure 5.1

Effect of the type of metal liquid catalysts on the viscosity of the liquid product after reaction.

that can reduce the catalytic activity, such as the adsorption of water in the rock formation of the reservoir, the cognate water of the reservoir, and the minerals present; however, these catalysts are taken to be supercritical conditions that show a high catalytic activity, presenting acid/base properties to participate in the catalytic upgrading of heavy oil, extra-heavy oil, bitumen, and residues. Water-soluble catalysts are recommended for the aquathermolysis reaction. This in situ catalytic upgrading process is carried out under steam injection, where high temperature steam reacts with the oil in the presence of the catalyst (Hyne et al. 1982). The heat provided by the steam acts with the high molecular weight hydrocarbons that are present in heavy crude oil, and in combination with the catalyst, they transform into lower molecular weight hydrocarbons. The chemical reaction for aquathermolysis has been reported to be as follows: (Hyne et al. 1982) RCH2 − CH2 S − CH3 + 2H2 O → RCH3 + CO2 + H2 + H2 S + CH4 The main mechanism of the aquathermolysis reaction is the breaking of the C—S bond, generating hydrogen from the water molecules present in the medium. This process has the following advantages (Ancheyta and Speight 2007; Speight 2009): – – – – – –

Convert heavy crude oil to light crude oil Reduce viscosity to improve the extraction of heavy crude oil Higher API gravity Lower sulfur content Lower nitrogen content Higher content of distillates: naphtha ( Ni catalyst > Fe catalyst. Kukushkin et al. (2018) studied the steam cracking of heavy oil at 425 ∘ C and 2.0 MPa on iron and molybdenum-based catalysts. It was found that the yield of light fractions ( Mo > Cu > Ni > Fe > Co > Zn > Mn.

5.4

Oil-Soluble Catalysts

Metal precursors in oil-soluble dispersed catalysts precursors are created by the combination of an oxide, a sulfide, or a salt of metal from group IV through VIII including transition metal-based catalysts derived from the organic acid salt or metal–organic compounds of Mo, V, W, Cr, Fe, etc. (Suwaid et al. 2020). In this type of catalyst, the metal sulfide particles are polar (hydrophilic) in nature and are associated with the hydrophilic molecules of resins and asphaltenes in the oil due to the presence of sulfur, nitrogen, and functional groups, as well as metals Ni and V. Formulated transition metal catalysts (mainly Co, Fe, Mo Ni, V) are synthesized from organometallic complexes.

5.4.1

Synthesis Procedure

Some examples of synthesis of oil-soluble catalysts are described below: – Example 1 Preparation of ferric oleate-based catalyst. In a flask, iron (III) nitrate (Fe(NO3 )3 ) monohydrate is dissolved with distilled water to obtain a saturated solution, and an aqueous solution (1 mol/l) of sodium hydroxide (NaOH) is slowly added to this solution, and to the saturated solution with continuous stirring until the pH value of the mixture solution reached 4–6. Subsequently, oleic acid (C18 H34 O2 ) is added with a molar ratio of 1.8 : 1 (oleic acid/ferric nitrate/sodium hydroxide). The reagents are heated to 125 ∘ C and stirred for 4 hours. An upper organic layer is formed which should be separated with a funnel and washed three times with n-heptane (C7 H16 ), subsequently it is left to dry under vacuum at 80 ∘ C for 24 hours. After the solvent is evaporated, the final product is ferric oleate (C54 H99 FeO6 ) and a reddish-brown viscous oil is obtained. – Example 2 In a three-neck flask, 2-ethylhexanoic acid and 12% NaOH are placed with a molar ratio of 1 : 0.98, and gasoline is used as a solvent. This mixture is heated at 50–760 ∘ C for 30 minutes. Subsequently, a solution formed by cobalt dichloride (CoCl2 ⋅6H2 O) and 2-ethylhexanoic acid, with a ratio of 2 : 0.98, is slowly incorporated into the previous solution by dripping and with constant stirring saponification liquor. The temperature should be maintained at 80–90 ∘ C for 30 minutes. The resulting mixture is washed two or three times with water and the organic phase is collected. – Example 3 The preparation of the oil-soluble organic acid cobalt catalyst is as follows: the first step is to put 2-ethylhexanoic acid and NaOH at 12% (m/m) in a three-necked flask with the following

243

244

5 Catalyst for In Situ Upgrading of Heavy Oils

molar ratio (1 : 0.98–0.998). Gasoline is used as a solvent. The above mixture is heated at 50–70 ∘ C for 30 minutes. Subsequently, CoCl2 (the molar ratio of 2-ethylhexanoic acid and CoCl2 ⋅6H2 O is 2 : 0.98) is slowly added to the saponification liquor under stirring. The temperature must be maintained between 80 and 90 ∘ C for 30 minutes. The reaction is completed after 90 minutes. The mixture is then washed two to three times with water, and the organic phase is collected. The oil-soluble catalyst to be used in the thermal recovery of heavy oil will be obtained after distillation, dehydration, and sedimentation. The catalyst is a viscous purple liquid. – Example 4 Oil-soluble Co, Ni, Fe, and Cu carboxylates are obtained by exchange reaction between the sodium salt of tall oil and nonorganic cobalt salts (Sitnov et al. 2016). The catalyst precursor of 1.0 wt% oil is injected into the emulsion. The active species are formed in situ. After one-hour reaction, generally, all of the metals are effective in not only suppression of coke and gas formation but also showing the high levels of desulfurization, demetallization, and Conradson carbon conversion. However, the molybdenum gives the best results in terms of liquid yield (95.8–97.2 wt%) and coke (0.5–0.8 wt%). Oil-soluble dispersed catalyst based on iron has the greatest advantage in terms of cost but seems to be not very effective, with the highest coke formation (5.2 wt%) and gas yield (2.8 wt%) (Figures 5.3 and 5.4).

5.4.2

Activity of Oil-Soluble Catalysts

Suwaid et al. (2020) proposed catalysts based on transition metals (Fe, Co, Ni) – 70 g of heavy oil and 30 g of water, at 300 ∘ C for 24 hours. A significant decrease was obtained in the content of resins and asphaltenes from 26.30% and 8.26% to 16.55% and 1.49%, respectively. 100%

80%

60%

40%

Gas (wt%) Liquid (wt%) Coke (wt%)

20%

0%

e on

N

a) e( t a

) (a

l ke

ic

N

o ct

sin

e tr al

o

b

Co

)

(a

e at

n

Iro

n

Figure 5.3

en

bd

y ol

um

a)

e(

t na

si

e

th

h ap

M

a)

e(

t na

re

) b) (b e( te t a a N sin sin re re re m um um nu ni di e o a d r an yb Ch V ol M b)

e( on

e at sin

) (b

Performance of some common metal compounds.

a: Jobo crude b: Cold Lake crude

5.4 Oil-Soluble Catalysts

24

API gravity

20 16 12 8 4 0 Feed

Fe

Cu

Water-soluble catalyst Figure 5.4 reaction.

Ni

Zr

Effect of the type of water-soluble catalysts on the API gravity of the liquid product after

Vakhin et al. (2018) loaded a 300 cm3 reactor with oil and water in a 70 : 30% mass ratio, 1.0 wt% of nickel carboxylate catalyst is used at 300 ∘ C for six hours, where the experiments were carried out on a rock sample extracted at a depth of 1900 m. The results indicated that the asphaltene content was reduced by 27%, the resins by 41.5%, 22% of sulfur was removed, and no change in nitrogen content was observed. Vakhin et al. (2021) carried out the aquathermolysis of extra-heavy oil with a nickel-based catalyst in a batch reactor with steam, at 300 ∘ C and 8.825 Mpa for 48 hours with 10% nickel-based catalyst. It was observed that a decrease in resins and asphaltenes resulted in an increase in saturates. The change in the resin content was from 871.7 to 523.3 amu and in asphaltenes from 1572.7 to 1072.3 amu. Alaei et al. (2017) placed 200 ppm of MoO3 catalyst with an average particle size of 50 nm, 100 g of heavy crude in the presence of H2 /N2 and H2 S as sulfur gas at 400 ∘ C and 7.943 MPa in a batch reactor. Sulfur dropped from 4.8 to 3.15 wt%; API changes from 12∘ to 20.2∘ and the viscosity dropped from 96.05 to 4.42 cSt. Zhao et al. (2014b) placed in a 300 cm3 reactor heavy oil and water in a 4/1 ratio at 300 ∘ C and 0.98 MPa for 24 hours, with a nickel catalyst and 7% formic acid as H2 donor. Increments of 6.8% in saturates and 4.37% in aromatics were observed. Meanwhile, resins and asphaltenes were reduced by 2.0 and 0.99 wt%, respectively, the sulfur content dropped from 0.565 to 0.074 wt% and nitrogen from 0.425 to 0.406 wt%. Galukhin et al. (2015) loaded a 300 cm3 reactor with 70 g heavy oil, 30 g water, and 4 g cyclohexane at 250 ∘ C for six hours and 0.14% iron acetyl acetonate catalyst. The resins content reduces from 30.7 to 16.3 and the asphaltene content from 26.6 to 13.4. Vakhin et al. (2017) found that Fe-based catalyst with H2 donor at 200 ∘ C lowered the viscosity and fractions of resin, and asphaltenes were significantly reduced. Petrukhina et al. (2014) loaded in a 1000 cm3 reactor at 300 ∘ C for five hours, a mixture of crude oil, kaolin, tetralin, and water in a ratio of 100 : 5 : 43%. The viscosity was reduced 98% and the resins changed from 38.37 to 22.06 wt%. Li et al. (2018) mixed 50 g of the heavy crude at 200 ∘ C for 24 hours with 10% ferric oleate catalyst (The catalyst did not contain water and the heavy crude contained only 12.0 wt% of water). The next changes were detected: saturates increase from 31.1 to 32.1 wt%;

245

246

5 Catalyst for In Situ Upgrading of Heavy Oils

aromatics from 32.7 to 33.8 wt%; resins decrease from 32.7 to 29.7 wt%; asphaltenes increase from 3.5 to 4.4 wt%, sulfur diminishes from 2.59 to 2.36 wt%, and nitrogen from 1.79 to 1.52 wt%. Nguyen et al. (2016) found that the catalytic yields of the metals follow the order: Mo>Ni>Ru> Co>V>Fe. Zhao et al. (2014a) studied oil-soluble organic Co salt as a catalyst and formaldehyde as H2 donor, for the aquathermolysis reaction of heavy oil. It was shown that aromatic hydrocarbons increased by 7.13%, while the resin and asphalt contents decreased by 6.4% and 5.42%, respectively, and the sulfur content decreased by 0.5%. Rezaei et al. (2012) loaded a 250 cm3 batch reactor, 80 g of vacuum residue (>524 ∘ C, 6.0 wt% sulfur and 17.7 wt% asphaltenes) with 100 ppm MoS2 (octoate) at 445 ∘ C and 12.75 MPa and H2 for one hour. The asphaltene conversion was 75%, and the sulfur conversion was 67%. Vakhin et al. (2018) loaded a 300 cm3 batch reactor with heavy crude oil and water in a 70/30% ratio and 1% catalyst based on Co, Ni, Fe, and Co at 250 ∘ C for six hours. The results indicated that sulfur and nitrogen did not present significant changes. An increment in the content of high molecular weight aromatic hydrocarbons was observed. Zhou et al. (2010) mixed in a batch reactor, 2 g of a catalyst precursor suspension (2-ethylhexanoate) with 181 g of heavy oil at 440 ∘ C, 15.2 MPa in atmosphere of H2 . Asphaltene conversion of 81 wt% was obtained. Panariti et al. (2000) mixed in a 30 cm3 batch reactor, 10 g of vacuum residue and 1000 ppm of different catalysts and H2 at 460 ∘ C for 90 minutes. The activity of the tested metals follows the order: Mo > Ni∼Ru > Co > V > Fe, high yields of distillates were presented, sulfur reduction from 3.9 to 2.6 wt%. Zhou et al. (2018) loaded a 500 cm3 autoclave with 150 g of extra-heavy crude, 100 ppm of MoS2 catalyst dispersed in ethanol at 390 ∘ C for three hours in a hydrogen atmosphere. It was found that 30% of the residue fraction is converted, and the viscosity is reduced by 60%. Zhao et al. (2014a) placed heavy crude oil and water in an autoclave in a 4 : 1 ratio, 7% formaldehyde as hydrogen donor and 0.1% catalyst at 280 ∘ C for 24 hours. The viscosity of heavy crude was reduced by 89.5%, saturated and aromatic by 7.13% and 4.69%, respectively; resins and asphaltenes by 6.4% and 5.42%, respectively, and the sulfur content dropped by 0.5%. Zhong et al. (2003) placed heavy crude oil, 0.02% Fe catalyst, and 10 ml of tetralin in an autoclave at 240 ∘ C for 72 hours, observing that the viscosity dropped by 60%. Wen et al. (2007) tested a molybdenum oleate catalyst with a concentration of 0.5% with heavy crude oil at 240 ∘ C for 24 hours, resulting in a 90% drop in viscosity, they also added 0.1% of a molybdenum emulsion, resulting in a viscosity reduction of 90% at 200 ∘ C for 24 hours.

5.5

Mineral Catalysts

A reservoir is mainly composed of rock material, crude oil, water, and impurities. The mineral catalysts of the deposit are commonly quartz, feldspar, and clay, mainly zeolites. In particular, the clay mineral has drawn attention for its high specific surface area and high density of acid sites that have catalytic effects. It was first reported that metal-containing particles in clay minerals could change the equilibrium of the hydrothermal cracking reaction. Fan et al. (2001) suggested that Al3+ interacts with H4 SiO4 to form a strongly acidic hydroxyl group on its surface,

5.5 Mineral Catalysts

while adsorption and association of water molecules on the surface of Al3+ produce Bronsted acid sites, causing polarization of asymmetry that makes SiOOHAl groups become polarized and have high acidity, accelerating the hydrothermal cracking reaction and thus reducing the viscosity of heavy oil. On the other hand, clay mineral yield is influenced by reservoir temperature. Figure 5.3 shows that the reservoir temperature can suddenly rise to several thousand degrees Celsius near the ISC front; therefore, it is necessary to study the thermal behavior of clay minerals at high temperatures. Chin et al. (2017) have studied the relationships between the chemical compositions of clay minerals and their thermal behaviors. It was found that the types of clay minerals have an influence on the product species and quality throughout the reaction process. Fan et al. (2002) further noted that the clay minerals mainly influence the combustion reaction by catalyzing the LTO reaction, i.e. the clay mineral causes an increase in the reaction rate, which lowers the activation energy of the oxidation/cracking reaction and, consequently, the amount of fuel deposited increases. The catalytic effects of mineral catalysts are based on the existence of a variety of metal particles. Ovalles et al. (2003) studied the functions of minerals containing iron, silicon, alumina, etc., in the hydrothermal cracking reaction, and the results of their experiments showed that the viscosity-reducing effects of the above substances followed an ascending order of minerals < SiO2 < Fe2 O3 < silicon-aluminum oxide, especially the viscosity reduction that can be done more effectively using high acidity silica-alumina. However, it has been found that the hydrothermal cracking reaction almost fails under anhydrous conditions (Ogbuneke et al. 2007). It is worth mentioning that with increases in water consumption, the improvement effect also increases, which inhibit the coking of heavy oil, although this relationship is not linear. In Bagci’s (2005) study, it was confirmed that clay minerals not only lowered the activation energy of the reaction and changed the Arrhenius constant but also influenced the maximum value of the combustion reaction during the in situ combustion (ISC). Yu et al. (2017) studied the effects of reaction temperature and air/oil ratio on the catalytic process of clay minerals in the low temperature oxidation (LTO) reaction stage. It was stated that the catalytic activity of clay minerals increased with increasing temperature or air/oil ratio. Furthermore, they found that the clay had catalyzed the fuel. Deposition process and the asphaltene content increased with increasing clay content, although the viscosity of the crude oil decreased as a whole. Interestingly, it was concluded that clay minerals in situ or by injection accelerated the rate of reaction as well as the rate of heat release and thus improved the enhancement effect. Pillared clay (PILC) is made by expanding the spacing of clay layers using a pillar technique, which is widely used as catalysts or molecular sieves. Faillace et al. (2017) used pillared clay as a catalyst to upgrade heavy oil, and it was reported that it promoted the production of lighter components of paraffin oil, highlighted that the catalytic effect of pillared clay is the best.

5.5.1

Synthesis Procedure

Mineral catalysts can be easily prepared and are useful when low conversions are required in hydrocracking and hydrotreating reactions. They are usually prepared in two ways: – Only using a single catalyst of mineral origin – Using the mixture of two or more catalysts and an additional material to provide mechanical and/or textural properties; normally it is a strong base that collaborates in the hydrolysis of the water present in the heavy hydrocarbon

247

248

5 Catalyst for In Situ Upgrading of Heavy Oils

5.5.1.1 Single-Mineral Catalysts

For the preparation of catalysts that use a single mineral, there are various minerals that can be used as catalysts in the hydrocracking and hydrotreating reactions, the most common are the following: – – – – – – –

pyrite molybdenite hematite limonite siderite plentlandite pyrrhotite

The raw ore can be used as a catalyst but must be subjected to a series of physicochemical processes whose objectives are to remove undesirable mineral compounds (impurities) and the physical and chemical processes that are carried out on the mineral follow the order: (a) Crushing and pulverizing The mineral to be used as a catalyst is crushed and pulverized to the desired particle size, considering the degree of dispersion to which it is desired to work, as shown in Table 5.1. Hematite and molybdenite (containing nickel, iron, or molybdenum) can be used as catalysts only with the grinding and grinding process. (b) Drying and calcination The pulverized ore is heated for 30–60 minutes in a furnace or muffle at 120–150 ∘ C in order to remove moisture and excess acid used in the previous procedure. If the ore is other than molybdenite, it is heated for 1 to 60 minutes at a temperature of 650–800 ∘ C to remove organic material and some carbonates. It is likely that after drying and calcination of the ore, the ore will agglomerate, for which it will be necessary to pulverize again to the desired dispersion size. (c) Dispersion or microdispersion The catalyst is added after loading the crude oil in the reactor, and the temperature is raised to 150 ∘ C with intense stirring for 60 minutes. Subsequently, the reaction temperature is reached. It is recommended that the amounts of mineral catalyst to be used should not be greater than 1 wt% with respect to crude oil. Typical values of mineral catalyst range from 250 to 10 000 ppm. 5.5.1.2 Blend of Mineral Catalysts

The preparation procedure of this catalyst is similar to that used from a single mineral as explained above, differing in that two or more minerals are used in specific ratios. The materials that are needed are the following: sodium hydroxide, two or more minerals with catalytic oxides, and hydrochloric or sulfuric acid 1 M. The procedure is like that of a single mineral. Each one of the minerals is crushed, pulverized, acidified, dried, and calcined separately to later be mixed before being dispersed or mixed with bitumen, heavy crude oil, and/or extra heavy oil. The type and quantities of each of the minerals used for the preparation of natural catalysts depend on the catalytic mixture to be prepared, which is summarized in the Table 5.2. Table 5.1

Particle size and dispersion. Particle size

Desired dispersion

Diameter

Tyler mesh

Granular

>74 μm

20–400

Microdispersity

iron > Molybdenite.

5.6 Ionic Liquids

In general, the catalytic activity of most mineral catalysts is insufficient to achieve important improvement in crude oil quality, and the injection of mineral particles in the reservoir is difficult in the practice. Hence, in spite of mineral catalysts lower price, water-soluble catalysts or oil-soluble catalysts are preferred for in situ heavy crude oils upgrading purposes.

5.6

Ionic Liquids

For several decades, ionic liquids as a catalytic phase have been the object of studies. They are good solvents for a large number of organic and inorganic materials with the possibility of carrying out homogeneous reactions. They are stable in different operating conditions against many compounds, which is due mainly to the singular properties that this type of fluid possesses. Ionic liquids are mainly composed of ions, but these are distinguished from the classic molten salts (NaCl) by their low melting point, which is approximately [(Et)3 NH][AlCl4 ]-Fe2+ > (Et)3 NH][AlCl4 ]-Cu+ . The viscosity was also reduced by 64.76%, 61.43%, and 62.62%, respectively, and the molecular weight is reduced according to the same order mentioned. Fan et al. (2009) loaded an autoclave with 200 g of heavy crude oil and ionic liquid [BMIM] [AlCl4 ] at 90 ∘ C and 35 kg/cm2 . The experimental results showed that the presence of sulfur in heavy crude oils is beneficial for the reduction of viscosity. Viscosity reduction, using the ionic liquid [BMIM][AlCl4 ], was enhanced when the water content of the heavy crude oil was less than 10%. The optimum temperature for upgrading heavy oil was 65–85 ∘ C using [BMIM][AlCl4 ]. The combination of transition metal salts and [BMIM][AlCl4 ] reduced the viscosity by more than 60%, and the asphaltene content was reduced by 78%. Shaban et al. (2014) loaded an autoclave (5 mm internal diameter, 8 mm external diameter and 25 cm long) with 100 g of heavy crude oil and ionic liquids [BMIM][FeCl4 ] (0.25, 0.5, 1.0), at 90 ∘ C. The viscosity reduction was according to the amount of ionic liquid added 60%, 68%, and 72%. In terms of the metal used for ionic liquid catalysts, the order of activity displayed by these catalysts follows the trend: Co > Fe > Mo > Ni > Cu.

5.7

Catalysts Characterization

To ascertain if the different type of catalysts were synthesized according to their application for the heavy oil upgrading, they are normally characterized by one or several of the analytical techniques briefly described in Table 5.3. Hereafter, examples of characterization of water-soluble catalysts are presented. Evolution of sample weight loss, endothermic, or exothermic process during heating treatment of catalysts precursors or final formed catalysts are commonly analyzed by DTA-DSC (Figure 5.5). Often, these analyses are employed as a tool to help in the identification of the active phase and also to assess their thermal stability. In the case of water-soluble catalysts, it is common to track the precursor evolve to active phase by several characterization techniques. In Figure 5.5, the “A zone” is characterized by an increment in the mass loss rate corresponding to 30 wt% due to the removal of water in the crystalline structure. The “B zone” is characterized by small variations in the mass loss rate; in this case, a mass loss of 6 wt%, which resulted from the condensation reaction of phosphoric acid. The “C zone,” between 376 and 450 ∘ C, can be related to the nickel-molybdenum interactions. Finally, up to 650 ∘ C (D zone), the salts were transformed presumably into an oxide form (Schacht et al. 2019).

253

254

5 Catalyst for In Situ Upgrading of Heavy Oils

Table 5.3

Techniques commonly employed to water soluble catalysts characterization.

Test

Brief description of the method

Atomic absorption spectrophotometer

Atomic absorption spectroscopy instruments place a sample in a high temperature flame that yields atomic species and passes selected, element specific, illumination through the flame to detect what wavelengths of light the sample atoms absorb. Either acetylene or nitrous oxide fuels the analytical flame.

X-ray diffraction

It is a basic characterization technique for materials with a crystalline (non-amorphous) structure: metals, minerals, polymers, catalysts, plastics, pharmaceuticals, thin-film coatings, ceramics, semiconductors, and even fluids. This technique is nondestructive, which allows the recovery of the studied material without any deterioration of the material.

Thermogravimetric and differential thermal analysis

It includes the study of the evolution of the properties of a sample or compound when it is subjected to heating at high temperatures. Thermogravimetry (TG) involves the measurement of the change in mass of the sample with temperature. Differential Thermal Analysis (DTA) measures the temperature difference between a sample and a reference material, both subjected to the same heating processes.

Potentiometric titration

The potentiometric titration method consists of measuring the potential (voltage) in a solution by means of an electrode as a function of the volume of titrant.

Infrared spectroscopy

It is the method used to determine the structures of molecules with characteristics of absorption of infrared radiation according to their molecular vibration. The IR region ranges from 12 800 to 10/cm and can be divided into the near-IR (NIR) region (12 800–4000/cm), mid-IR (4000–400/cm), and far-IR or FAR-IR. (50–400/cm). The vibrational spectrum of a molecule is considered unique and therefore characteristic physical property of this molecule. Thus, among other applications, the IR spectrum can be used as a “fingerprint” to detect functional groups present in the sample by comparing it with reference spectra.

Nuclear magnetic resonance

This technique allows to solve the structure of organic and organometallic compounds, enables the determination of the stereochemistry, and provides valuable information on other magnetically active nuclei. NMR is an absorption spectroscopy; whose foundation lies in the property that some nuclei have of absorbing energy when they are subjected to the action of radio frequencies.

Mössbauer spectroscopy

Spectroscopic technique based on the resonant emission and absorption of gamma rays in solids. Examines the nuclear excitation of an atom by gamma rays in the solid state and provides information on the oxidation state and electronic environment around the atom.

UV–Vis spectroscopy

It is based on the transmission and/or absorption of electromagnetic radiation when it interacts with matter, in the wavelength environment between 190 and 900 nm, which is extended in UV–VIS–NIR equipment up to 3300 nm. It is used in the quantitative determination of the components of solutions of transition metal ions and highly conjugated organic compounds, determining the concentration of a compound in solution.

Some characterization techniques are also employed to samples recovered after heavy oil upgrading. In this case, the tests are useful to identify the active phase and also to estimate deactivation degree in some special cases. Figure 5.6 shows the XRD pattern of the solid material collected after carrying out the heavy crude oil upgrading reaction with the copper-based liquid catalyst. In this case, it was composed of crystalline particles of four different copper sulfides, namely, chalcocite (Cu2 S), copper sulfide

5.7 Catalysts Characterization

100

10 Weight loss

90

Heat changes –10

Weight (%)

70

–30

5

4

–50

60

B

C

–70

3

50 1

40

–90 2 –110

30

Microvolt endo down (μV)

80

–130

20

A

D –150

10 0 0

100

200

300

400

500

600

700

–170 800

Temperature (°C) Figure 5.5

DTA/TGA analysis of precursors of a NiMoP catalysts.

(Cu1.96 S), covellite (CuS), and copper sulfide (Cu1.81 S) matching with the JCPDS cards with numbers 00-012-0227, 00-012-0205, 00-002-0820, and 00-041-0959, respectively. As was commented previously for the catalyst prepared with nickel, the copper composing the corresponding liquid catalyst reacted in situ with the sulfur species contained in the heavy crude oil yielding a series of copper sulfides, which are reported to be part of the dominant driving force to convert asphaltene molecules via hydrocracking reactions. Scanning electron microscopy (SEM) is employed to obtain morphological information of solid catalyst particles. Microscope equipped with elemental microanalysis allows us to perform a semiquantitative estimation of the catalysts composition by mapping selected surface sections. The obtained information is useful when improvement of catalysts synthesis procedure obtained better morphological homogeneity. Figure 5.7 shows micrographs of heteropoliacid catalysts, and SEM micrographs revealed the presence of irregularly shaped porous agglomerates. A key important factor in the properties of catalysts for upgrading of crude oil is the acidity because it is related to the ability for the hydrocarbon bond breaking. Hence, measurement of acidity directly to the catalytic active phase is very important, FT-IR spectra of adsorbed pyridine is a useful tool for this purpose, since it allows the measurement of the quantity of active sites and also their differentiation as Brønsted- or Lewis-type acid sites. Figure 5.8 shows the spectra for the catalyst formulated with Ni, which displays two signals, the first one at 1539 cm−1 associated with Brønsted-type acid sites, and the second one at 1446 cm−1 related to Lewis-type acid sites. Ascribed to both types of acid sites, an additional band located at 1486 cm−1 was also observed. It is also noted that, in general, as temperature increases, the amount of pyridine that remained adsorbed decreases, which is particularly evident for the band related to the Lewis’s acid sites

255

5 Catalyst for In Situ Upgrading of Heavy Oils

Millerite

Szomolnokite, syn

Nickel sulfide

Iron phosphate

Intensity (a.u.)

Intensity (a.u.)

Sulfur

10

20

30

40 50 2θ (°)

60

70

80

10

30

50

70

2θ (°)

(a)

(c)

Intensity (a.u.)

Zirconium phosphate Zirconium oxide Hydrogen phosphate

Intensity (a.u.)

256

Chalcocite, syn Copper sulfide (Cu1.96S) Covellite Copper sulfide (Cu1.81S) 10

30

50

70

10

30

50

2θ (°)

2θ (°)

(b)

(d)

70

Figure 5.6 X-ray diffraction pattern of the residual solid recovered after the heavy crude oil upgrading reaction with different liquid catalysts: (a) Ni, (b) Cu, (c) Fe, and (d) Zr.

(a)

Figure 5.7

(b)

SEM images of the (a) HPA-Ni catalyst, and (b) HPA catalyst.

5.8 Concluding Remarks

300 °C

400 °C 300 °C 200 °C

Trasmittance (%)

Transmittance (%)

200 °C

100 °C

100 °C

50 °C

840

986

1024

1178

1360

1486

1284

50 °C 1539 (Brønsted)

1446 (Lewis) 1700

1650

1600

1550

1500

1450

1600

1400

1400

Wavenumber (cm–1)

1200

1000

800

Wavenumber (cm–1)

(a)

(c) 400 °C

400 °C 300 °C 300 °C

Transmittance (%)

Transmittance (%)

200 °C

200 °C 100 °C

50 °C

100 °C

50 °C

1448 (Lewis)

1542 (Brønsted) 1630

1491 1482 1449 Lewis

1700

1650

1600

1550

1500

1450

1400

1700

1650

1600

1550

1500

Wavenumber (cm–1)

Wavenumber (cm–1)

(b)

(d)

1450

1400

Figure 5.8 FT-IR spectra of adsorbed pyridine collected after desorption at various temperatures for the residual solid recovered after the heavy crude oil upgrading reaction with the different liquid catalysts: (a) Ni, (b) Cu, (c) Fe, and (d) Zr.

5.8

Concluding Remarks

The main advantages and disadvantages of the catalysts used for the in situ upgrading of heavy and extra-heavy crude oils are – Water-soluble catalysts They are cheaper, easier to obtained, and easy to prepare. Water-soluble catalysts have the following order of activity: Zr > Mo > Cu > Ni > Fe > Co > Zn > Mn. Their dispersion effect in the reservoir water makes it less ideal than other types of catalysts. Metallic salts can exchange cations with the rock formation of the deposit to generate active sites and promote the combustion reaction.

257

258

5 Catalyst for In Situ Upgrading of Heavy Oils

– Oil-soluble catalysts Oil-soluble catalysts can provide greater contact with the oil. They dissolve in the injected fluid and disperse well in the reservoir fluid. They need to use solvents such as tetralin, decalin, diesel, and toluene as a carrier. Oil-soluble catalysts exhibit higher preparation cost and greater environmental pollution. – Mineral catalysts Mineral catalysts are based on the existence of a variety of metallic particles. The clay mineral has drawn attention for its high surface area and lot of acid sites (clay minerals are hydrated aluminum phyllosilicates, sometimes with varying amounts of iron, magnesium, alkali metals, alkaline earths, and other cations). – Ionic liquids Ionic liquids are not volatile. They do not promote combustion, have high thermal stability, high ionic conductivity, and are effective for the desulfurization of heavy oil. The water content must be less than 10%, and they have environmental protection assets. – In terms of viscosity reduction, the different catalysts show the following order: Mineral catalysts < Water-soluble catalysts < Oil-soluble catalysts < Ionic liquids catalysts.

References Alaei, M., Bazmi, M., Rashidi, A., and Rahimi, A. (2017). Heavy crude oil upgrading using homogenous nanocatalyst. Journal of Petroleum Science and Engineering 158: 47–55. Alarbah, A., Shirif, E., Jia, N., and Bumraiwha, H. (2021). A new approach utilizing liquid catalyst for improving heavy oil recovery. Journal of Energy Resources Technology 143 (7): 1–12. Alemán-Vázquez, L.O., Torres-Mancera, P., Ancheyta, J., and Ramírez-Salgado, J. (2016). Use of hydrogen donors for partial upgrading of heavy petroleum. Energy & Fuels 30 (11): 9050–9060. Al-Marshed, A., Hart, A., Leeke, G. et al. (2015). Optimization of heavy oil upgrading using dispersed nanoparticulate iron oxide as a catalyst. Energy & Fuels 29: 6306–6316. Ancheyta, J. and Speight, J.G. (2007). Hydroprocessing of Heavy Oils and Residua. Boca Raton: CRC Press. Atta, A.M., Al-Lohedan, H.A., and Ezzat, A.O. (2021). Synthesis and application of geminal dicationic ionic liquids and poly (ionic liquids) combined imidazolium and pyridinium cations as demulsifiers for petroleum crude oil saline water emulsions. Journal of Molecular Liquids 325: 115264. Bagci, S. (2005). Effect of pressure on combustion kinetics of heavy oils. Energy Sources 27 (9): 887–898. Barbaro, P., Liguori, F., Linares, N., and Moreno Marrodan, C. (2012). Heterogeneous bifunctional metal/acid catalysts for selective chemical processes. European Journal of Inorganic Chemistry 24: 3807–3823. Chin, C., Ahmad, Z., and Sow, S. (2017). Relationship between the thermal behaviour of the clays and their mineralogical and chemical composition: example of Ipoh, Kuala Rompin and Mersing (Malaysia). Applied Clay Science 143: 327–335. Clark, P.D., Dowling, N.I., Hyne, J.B., and Lesage, K.L. (1987). The chemistry of organosulphur compound types occurring in heavy oils: 4. The high-temperature reaction of thiophene and tetrahydrothiophene with aqueous solutions of aluminium and first-row transition-metal cations. Fuel 66: 1353–1357.

References

Clark, P.D., Clarke, R.A., and Hyne, J.B. (1990). Studies on the chemical reactions of heavy oils under steam stimulation conditions. AOSTRA Journal of Research 6: 29–39. Duan, Y., Liu, Y., Chen, Z. et al. (2020). Amorphous molybdenum sulfide nanocatalysts simultaneously realizing efficient upgrading of residue and synergistic synthesis of 2D MoS2 nanosheets/carbon hierarchical structures. Green Chemistry 22: 44–53. Faillace, J.G., Fernandes de Melo, C., Lima de Souza, S.P., and Costa Marques, M.R. (2017). Production of light hydrocarbons from pyrolysis of heavy gas oil and high density polyethylene using pillared clays as catalysts. Journal of Analytical and Applied Pyrolysis 126: 70–76. Fan, H.-F., Liu, Y.-J., and Zhong, L.G. (2001). Estudios sobre los efectos sinérgicos de minerales y vapor en los cambios de composición de los petróleos pesados. Energía y combustibles 15 (6): 1475–1479. Fan, H.F., Liu, Y.J., Zhang, L.Y., and Zhao, X.F. (2002). The study on composition changes of heavy oils during steam stimulation processes. Fuel 81 (13): 1733–1738. Fan, H., Li, Z., and Liang, T. (2007). Experimental study on using ionic liquids to upgrade heavy oil. Journal of Fuel Chemistry and Technology 35 (1): 32–35. Fan, Z., Wang, T., and He, Y. (2009). Mejoramiento y reducción de viscosidad de crudos pesados por [BMIM][AlCl4 ] líquido iónico. Revista de Química y Tecnología de Combustibles 37 (6): 690–693. Galukhin, A.V., Erokhin, A., and Nurgaliev, D.K. (2015). Effect of catalytic aquathermolysis on high molecular weight components of heavy oil in the Ashalcha field. Chemistry and Technology of Fuels and Oils 50: 555–560. García-Dávila, J., Ocaranza-Sanchez, E., Sánchez, C., and Martínez-Ayala, A.L. (2018). Catalytic activity of a bifunctional catalyst for hydrotreatment of Jatropha Curcas L. seed oil. Journal of Spectroscopy 2: 1, 5326456–7. Guo, A., Wu, C., He, P. et al. (2016). Low-temperature and low-pressure non-oxidative activation of methane for upgrading heavy oil. Catalysis Science & Technology 6: 120. Gutierrez-Acebo, E., Leroux, C., Chizallet, C. et al. (2018). Metal/acid bifunctional catalysis and intimacy criterion for ethylcyclohexane hydroconversion: when proximity does not matter. ACS Catalysis 8 (7): 6035–6046. Guzmán-Lucero, D., Flores, P., Rojo, T., and Martínez-Palou, R. (2010). Ionic liquids as demulsifiers of water-in-crude oil emulsions: study of the microwave effect. Energy & Fuels 24 (6): 3610–3615. Hao, H., Su, H., Chen, G. et al. (2015). Viscosity reduction of heavy oil by aquathermolysis with coordination complex at low temperature. The Open Fuels & Energy Science Journal 8: 93–98. Hart, A. (2014). The application of bionanoparticles for in situ upgrading of heavy oil Thesis submitted to The University of Birmingham for the PhD degree, School of Chemical Engineering College of Engineering and Physical Science University of Birmingham. He, P. and Song, H. (2017). Catalytic natural gas utilization on unconventional oil upgrading. In: Advances in Natural Gas Emerging Technologies (ed. H. Al-Megren and R. Altamimi). London, UK: IntechOpen. Huseey, C.L. (1988). Room temperature haloaluminate ionic liquids. Novel solvents for transition metal solution chemistry. Pure and Applied Chemistry 60: 1763–1772. Hyne, J. B., Greidanus, J. W., Tyrer, J. D. et al. (1982). Aquathermolysis of heavy oils. The Second International Conference on Heavy Crude and Tar Sands, Caracas, Venezuela; 25–30. Johnstone, R., Wilby, A.H., and Entwistle, I.D. (1985). Heterogeneous catalytic transfer hydrogenation and its relation to other methods for reduction of organic compounds. Chemical Reviews 85: 129–170. Kukushkin, R.G., Eletskii, P.M., Zaikina, O.O. et al. (2018). Studying the steam cracking of heavy oil over iron- and molybdenum-containing dispersed catalysts in a flow-type reactor. Catalysis in Industry 10 (4): 344–352.

259

260

5 Catalyst for In Situ Upgrading of Heavy Oils

Laredo, G., Likhanova, N., Lijanova, I. et al. (2015). Synthesis of ionic liquids and their use for extracting nitrogen compounds from gas oil feeds towards diesel fuel production. Fuel Processing Technology 130: 38–45. Li, J., Chen, Y., Liu, H. et al. (2013). Influences on the aquathermolysis of heavy oil catalyzed by two different catalytic ions: Cu2+ and Fe3+ . Energy & Fuels 5: 2555–2562. Li, Y.R., Li, Q.Y., Wang, X.D. et al. (2018). Aquathermolysis of heavy crude oil with ferric oleate catalyst. Petroleum Science 15: 613–624. Liu, D., Kong, X., Li, M., and Que, G. (2009). Study on a water-soluble catalyst for slurry-phase hydrocracking of an atmospheric residue. Energy & Fuels 23 (2): 958–961. Moein, E.S., Ahmadi, K.M., Carbognani, O.L. et al. (2020). Chemical insight into nano-catalytic in-situ upgrading and recovery of heavy oil. Fuel 278: 1–9. Nguyen, M.T., Nguyen, N.T., Cho, J. et al. (2016). A review on the oil-soluble dispersed catalyst for slurry-phase hydrocracking of heavy oil. Journal of Industrial and Engineering Chemistry 43: 1–12. Ogbuneke, K.U., Snow, C.E., Andresen, J.M. et al. (2007). American Chemical Society, Division of Petroleum Chemistry 52 (2): 118. Olivier-Bourbigou, H. and Magna, L. (2002). Ionic liquids: perspectives for organic and catalytic reactions. Journal of Molecular Catalysis A: Chemical 182-183: 419–437. Olivier-Bourbigou, H., Magna, L., and Morvan, D. (2010). Ionic liquids and catalysis: recent progress from knowledge to applications. Applied Catalysis A: General 373 (1): 1–56. Omar, S., Yang, Y., and Wang, J. (2020). A review on catalytic & non-catalytic bio-oil upgrading in supercritical fluids. Frontiers of Chemical Science and Engineering 15 (1): 4–17. Ovalles, C., Hamana, A., Rojas, I., and Bolivar, R. (1995). Upgrading of extra-heavy crude oil by direct use of methane in the presence of water: deuterium-labelled experiments and mechanistic considerations. Fuel 74: 1162–1170. Ovalles, C., Vallejos, C., Vasquez, T. et al. (2003). Downhole upgrading of extra-heavy crude oil using hydrogen donors and methane under steam injection conditions. Petroleum Science and Technology 21 (1–2): 255–274. Panariti, N., Del Bianco, A., Del Piero, G., and Marchionna, M. (2000). Petroleum residue upgrading with dispersed catalysts Part 1. Catalysts activity and selectivity. Applied Catalysis A: General 204: 203–213. Petrukhina, N.N., Kayukova, G.P., Romanov, G.V. et al. (2014). Conversion processes for high-viscosity heavy crude oil in catalytic and noncatalytic aquathermolysis. Chemistry and Technology of Fuels and Oils 50 (4): 315–326. Rezaei, H., Ardakani, J.S., and Smith, K.J. (2012). Comparison of MoS2 catalysts prepared from Mo-micelle and Mo-octoate precursors for hydroconversion of Cold Lake vacuum residue: catalyst activity, coke properties and catalyst recycle. Energy & Fuels 26: 2768–2778. Schacht, E., Kondratieva, E., Gutiérrez, O.Y., and Lercher, J.A. (2015). Pathways for H2 activation on (Ni)-MoS2 catalysts. The Journal of Physical Chemistry Letters 6 (15): 2929–2932. Schacht, P., Portales-Martínez, B., Laredo, G.C. et al. (2019). Homogeneous catalyst for in-situ hydrotreating of heavy oils. Applied Catalysis A: General 577: 99–106. Scott, C.E., Carbognani, L., and Pereira, P. (2019). In situ upgrading via hot fluid and nanocatalyst injection. In: Advanced Catalytic Materials: Current Status and Future Progress (ed. J.M. Domínguez-Esquivel and M. Ramos), 129–149. Springer, Cham. Shaban, S., Dessouky, S., Badawi, A.E.F. et al. (2014). Upgrading and viscosity reduction of heavy oil by catalytic ionic liquid. Energy & Fuels 28 (10): 6545–6553.

References

Sitnov, S.A., Petrovnina, M.S., Feoktistov, D.A. et al. (2016). Intensification of thermal steam methods of production of heavy oil using a catalyst based on cobalt (Russian). OIJ 106–108. Sitnov, S.A., Vakhin, A.V., Mukhamatdinov, I.I. et al. (2019). Effects of calcite and dolomite on conversion of heavy oil under subcritical condition. Petroleum Science and Technology 37 (6): 1–7. Speight, J.G. (2009). Enhanced Recovery Methods for Heavy Oil and Tar Sands. Gulf Publishing Company. Strausz, O.E., Mojelsky, T., Payzant, J.D. et al. (1999). Upgrading of Alberta’s heavy oils by superacid-catalyzed hydrocracking. Energy & Fuels 13 (3): 558–569. Suwaid, M.A., Varfolomeev, M.A., Ameen, A.-m.A. et al. (2020). In-situ catalytic upgrading of heavy oil using oil-soluble transition metal-based catalysts. Fuel 281: 118753. Vakhin, A.V., Sitnov, S.A., Mukhamatdinov, I.I. et al. (2017). Aquathermolysis of high viscosity oil in the presence of an oil soluble iron based catalyst. Chemistry and Technology of Fuels and Oils 53: 666–674. Vakhin, A.V., Mukhamatdinov, I.I., Aliev, F.A. et al. (2018). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: Part II Changes in composition of aromatic hydrocarbons. Petroleum Science and Technology 36: 1850–1856. Vakhin, A., Aliev, F., Mukhamatdinov, I. et al. (2021). Extra-heavy oil aquathermolysis using nickel-based catalyst: some aspects of in-situ transformation of catalyst precursor. Catalysts 11 (2): 189. Wang, J., Liu, L., Zhang, L., and Li, Z. (2014). Aquathermolysis of heavy crude oil with amphiphilic nickel and iron catalysts. Energy & Fuels 12: 7440–7447. Welton, T. (1999). Room temperature ionic liquids. Solvents for synthesis and catalysis. Chemical Reviews 99: 2071–2084. Welton, T. (2004). Ionic liquids in catalysis. Coordination Chemistry Reviews 248: 2459–2477. Wen, S., Zhao, Y., Liu, Y., Hu S. (2007). A study on catalytic aquathermolysis of heavy crude oil during steam stimulation. International symposium on oilfield chemistry, Houston, Texas, U.S. (28 February – 2 March 2007). SPE-106180-MS. Wilkes, J.S., Levisky, J.A., Wilson, R.A., and Hussey, C.L. (1982). Dialkylimidazolium chloroaluminate melts: a new class of room-temperature ionic liquids for electrochemistry, spectroscopy and synthesis. Inorganic Chemistry 21: 1263–1264. Yi, Y., Li, S., Ding, F., and Yu, H. (2009). Change of asphaltene and resin properties after catalytic aquathermolysis. Petroleum Science 6 (2): 194–200. Young, K. and Seung-Kyu, P. (2018). Viscosity reduction by catalytic aquathermolysis reaction of vacuum residues. Applied Chemistry for Engineering 29 (4): 468–473. Yu, X.C., Qu, Z., Kan, C.B., and Zhao, X.J. (2017). Effective of different clay minerals on heavy oil oxidation during ignition process. Energy & Fuels 31 (11): 12839–12847. Zhao, F., Liu, Y., Fu, Z., and Zhao, X. (2014a). Using hydrogen donor with oil-soluble catalysts for upgrading heavy oils. Russian Journal of Applied Chemistry 87: 1498–1506. Zhao, R., Li, W., Lee, J.H. et al. (2014b). Precise tuning of (YBa2 Cu3 O7-δ )1-x :(BaZrO3 )x thin film nanocomposite structures. Advanced Functional Materials 24 (33): 5240–5245. Zhao, F., Liu, Y., Lu, N. et al. (2021). A review on upgrading and viscosity reduction of heavy oil and bitumen by underground catalytic cracking. Energy Reports 7: 4249–4272. Zhong, L. G., Liu Y. J., Fan, H. F., and Jiang S. J. (2003). Liaohe extra-heavy crude oil underground aquathermolytic treatments using catalyst and hydrogen donors under steam injection conditions. SPE International Improved Oil Recovery Conference in Asia Pacific, Kuala Lumpur, Malaysia (20–21 October 2003). SPE 84863.

261

262

5 Catalyst for In Situ Upgrading of Heavy Oils

Zhou, B., Zhou, Z., and Wu, Z. (2010). Hydrocarbon-soluble, bimetallic catalyst precursors and methods for making same. US Patent 7842635 B2. Zhou, Z.Y., Zhao, Y.S., Zhang, X. et al. (2018). The study of dispersed catalyst on upgrading of unconventional oil. Earth and Environmental Science 188: 012073.

263

6 Nanoparticles for Heavy Oil In Situ Upgrading Muneer A. Suwaid 1 , Sergey A. Sitnov 1 , Ameen Al-Muntaser 1 , Chengdong Yuan 1,3 , Alexey Vakhin 1 , Jorge Ancheyta 1,2 , and Mikhail A. Varfolomeev 1 1

Department of Petroleum Engineering, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Instituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, Mexico City, 07730, Mexico 3 Center for Petroleum Science and Engineering, Skolkovo Institute of Science and Technology, Moscow 121205, Russia 2

6.1

General Aspects

Heavy oil, as one of the unconventional sources, accounts for a large portion of the global oil reserves (about 60–70% of the total proved oil reserves). It is considered as an important alternative and perhaps the most readily available oil resource to complement the conventional fossil fuels and meet near- and longer-term demands (Guo et al. 2015; Muraza and Galadima 2015; Pu et al. 2015). Nevertheless, heavy oil is well known for its high viscosity and density that lead to its poor flow ability (Hashemi et al. 2014a). Therefore, it is not easy to extract this type of heavy oil. Currently, the main technologies for heavy oil recovery are thermal methods, including in situ combustion (ISC) and steam injection (steam-assisted gravity drainage [SAGD], cyclic steam injection [CSI], steam flooding, etc.) (Abuhesa and Hughes 2009; Hashemi-Kiasari et al. 2014; Suwaid et al. 2022), where “thermal effect” generated by combustion or steam is expected to reduce the viscosity, which consequently improves the mobility of heavy oil (Ali and Meldau 1979; Butler et al. 1981; Cheih 1982; Yuan et al. 2018). However, after some field application and laboratory studies, other technical and economic challenges were exposed (Bagci and Kok 2001; Lee and Noureldin 1989; Shokrlu et al. 2013). The use of steam or hot water has a very limited effect on viscosity reduction and quality improvement of produced oil. The poor quality of recovered heavy oil because of the high content of organic sulfur, nitrogen, and metal-containing compounds (mostly asphaltenes) also raises great difficulties in the transportation, storage, and processing of these heavy oils. Hence, catalysts should be used to improve the efficiency of these thermal treatment processes and achieve a permanent in situ oil upgrading. Until now, various research have been carried out on catalytic oil upgrading in different thermal treatment processes including ISC, CSS, SAGD, and “toe-to-heel” air injection (THAI), and some of these works have been previously reviewed. For instance, Hashemi et al. reviewed the potential application of nanoparticle technology for in situ upgrading and recovery enhancement, including the synthesis of nanocatalysts, facilities for application, transport behavior in porous media, modeling of reaction kinetics, recovery enhancement, and evaluation of treated products (Hashemi et al. 2014a). Muraza and Galadima (2015) reported the application of catalysts in the aquathermolysis

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

264

6 Nanoparticles for Heavy Oil In Situ Upgrading

of heavy oil. Iskandar et al. (2016) discussed the use of nanocatalyst in aquathermolysis reaction for the viscosity reduction of heavy oil. Guo et al. also reviewed some aspects about in situ catalytic upgrading for heavy and extra-heavy oil recovery (Guo et al. 2016). Li et al. reported the recent advances on the transition-metal-based catalysts for heavy oil upgrading by aquathermolysis reactions (Li et al. 2019). All these previous reviews have demonstrated that metal-based nanoparticles have emerged as one of the most promising technologies for in situ oil upgrading. However, current technologies still face enormous challenges for heavy oil upgrading and recovery, which necessitates further advances in this field. An overall critical review about these existing methods would provide a systematic knowledge for developing new technologies to effectively extract and utilize heavy oil. This work is then focused on reviewing and discussing the recent progresses in catalytic oil upgrading and recovery using different thermal treatment methods, such as (i) asphaltene adsorption and catalytic oxidation to achieve heavy oil upgrading (Franco et al. 2013b; Nassar et al. 2011a); (ii) CAtalytic upgrading Process In situ (CAPRI) incorporated with THAI (Al-Marshed et al. 2015a; Hart et al. 2015); (iii) asphaltene removal for improving the recovery efficiency utilizing the reservoir as an immense reactor and nanomaterials as catalysts/adsorbents (Hosseinpour et al. 2013, 2014); (iv) catalytic upgrading in thermal cracking for EOR (Husein and Alkhaldi 2014); (v) upgrading of heavy oil by superheated steam injection (Kondoh et al. 2016); (vi) low-temperature catalytic thermal cracking of asphaltenes (Montoya et al. 2016); (vii) catalytic oil upgrading by supercritical water cracking (SWC) (Dejhosseini et al. 2013; Golmohammadi et al. 2016; Hosseinpour et al. 2015b; Kosari et al. 2017); (viii) catalytic oxidation of heavy oil in ISC process (Fan et al. 2015); (ix) oil upgrading in aquathermolysis processes (Galukhin et al. 2015; Khalil et al. 2015; Sitnov et al. 2020; Vakhin et al. 2021; Wei et al. 2007); (x) catalytic heavy oil upgrading and recovery via electromagnetic heating (Bera and Babadagli 2017); and (xi) catalytic steam injection process for oil upgrading (Hamedi Shokrlu and Babadagli 2013; López et al. 2017). Furthermore, differently to others, this book sheds lights on the advantages on the use of nanoparticles in heavy oil industry and addresses some of the limitations and challenges facing this technology.

6.2

Synthesis

Over the last several decades, tremendous efforts have been conducted in the development of reliable synthetic protocols for nanosized materials with full control on size, shape, and other special characteristics. In general, nanosized particles can be prepared either by physical, chemical, or biological methods depending on the nature of the materials and involvement of chemical reactions (Saravanan et al. 2008). Furthermore, the synthesis of nanoparticles can also be categorized as either a “top-down” or a “bottom-up” method (An and Somorjai 2015; Sepeur 2008). In top-down synthetic method, nanosized particles are usually produced by size reduction via various physical and/or chemical treatments. Common top-down synthetic methods include mechanical milling/ball milling, chemical etching, thermal ablation, laser ablation, explosion process, and sputtering (Meyers et al. 2006). Meanwhile, in bottom-up method, the nanoparticles are formed by smaller entities such as atoms, molecules, and smaller particles as building blocks that are assembled to form the final particle. Commonly, this method is based on either chemical or biological mechanisms. Several examples of bottom-up synthetic methods are chemical/electrochemical precipitation, vapor deposition, atomic/molecular condensation, sol–gel processes, spray pyrolysis, laser pyrolysis, aerosol processes, and chemical or biological reduction. For four types of nanomaterials that are commonly used in oil and gas application, i.e. metallic nanoparticles, metal oxide nanoparticles, and carbon nanotubes (CNTs), are discussed.

6.2 Synthesis

6.2.1

Metallic Nanoparticles

They can be synthesized with either physical approaches or chemical reactions (An and Somorjai 2015). These methods have been extensively discussed in many review articles (Aiken III and Finke 1999; Bönnemann and Richards 2001). In general, there are four most common methods to synthesize metallic nanoparticles, i.e. reduction of metal salt precursors, electrochemical synthesis, reduction of organic ligands inorganometallic precursors, and metal vapor chemistry (Pachon and Rothenberg 2008). Among these methods, metal salt reduction is considered as one of the simplest and the most common methods to produce metallic nanoparticles. In this approach, the metal salt precursor is typically mixed with reducing agent in the presence of stabilizing agent to prevent particle aggregation. The overall reaction mechanism of this process can be expressed as the following equation (Pachon and Rothenberg 2008): xMn+ + + nxe− + stabilizing agent = M0n(cluster) In this process, the metallic nanoparticles are generally formed in stepwise processes involving nucleation, growth, and agglomeration (Murphy et al. 2005). Figure 6.1 presents the schematic mechanism of the metallic nanoparticles formation synthesized via metal salt reduction. Initially, the metal cations (Mn+ ) in the solution are reduced to metal atoms (M0 ) using reducing agents. Several reducing agents that can be used include hydrogen, carbon monoxide, hydrazine, hydrides, or salts such as sodium borohydride and sodium citrates, different types of solvents such as alcohols, and most recently, a reducing agent obtained from natural sources (plant extracts) (Roucoux et al. 2002). Following the reduction process, the metal atoms are rapidly clustered to form small metallic nanoparticles nuclei or commonly called “seed” particles which will then slowly grow and aggregate together to form more stable and bigger metallic nanoparticles as the final product. At this step, stabilizing agent is commonly added to control the particle growth and aggregation. Here, various sizes, shapes, and morphologies of the final nanoparticles can easily be controlled using various types of stabilizing agents such as surfactants, polymers, or organic ligands (Pachon and Rothenberg 2008).

Reduction agent

Mn+

Stabilizing agent

Reduction

Stabilization

Mn+ Mn+ n+ Mn+M n+ n+ M M Mn+ Mn+ n+ M Metal salt precursor solution

Metallic NPs nuclei

Stable metallic NPs

Figure 6.1 Schematic mechanism of metallic nanoparticles formation in metal salt reduction process. Source: Khalil et al. (2017)/with permission of Elsevier.

265

6 Nanoparticles for Heavy Oil In Situ Upgrading

Metal ions Mn+ Mn+ Mn+ n+ Mn+ Mn+ M Mn+ Mn+

Metal ions

(b)

Mn+ Mn+ Mn+ Mn+

(c)

(d)

Reaction mechanism [53]: Ad-atoms Anode: Mbulk → Mn+ + ne– Cathode: Mn+ + ne– + Stabilizer → Mcell/Stabilizer Mbulk + Stabilizer → Mcell/Stabilizer

Cathode

(a)

Anode

266

(e)

Stable metallic NPs

Figure 6.2 Schematic mechanism of electrochemical synthesis method of metallic nanoparticles: (a) oxidative dissolution of metal bulk at anode, (b) ion migration, (c) electron transfer from cathode, (d) reductive formation of zerovalent ad-atoms, and (e) particle growth and stabilization. Source: Khalil et al. (2017)/with permission of Elsevier.

The second general method to synthesize metallic nanoparticles is electrochemical synthesis. This method was initially introduced by a research group at Max Planck Institute in 1994 (Reetz and Helbig 1994). In this approach, nanoparticles are generated via electrochemical redox reaction between cathode and sacrificial anode that consists of the bulk metal. Both schematic and reaction mechanisms of this process are shown in Figure 6.2. In general, this process is initiated by the oxidation of bulk metal at the anode to form metal ions. These metal ions will then travel to the cathode where the ions are reduced to form zerovalent metal atoms (ad-atoms) as seed particles, which further grow to form stable metallic nanoparticles (Figure 6.2). In literatures, various collections of both noble and transitional metals nanoparticles such as Pd, Co, Fe, Ag, Au, Ti, and Ni, as well as bi-metallic alloy nanoparticles such as Fe–Ni, Pd–Ni, and Fe–Co have been successfully synthesized using this methods (Reetz et al. 1995; Reetz and Helbig 1994). It is reported that this method allows an easy control for size-selective particle formation by simply adjusting the current density (higher current density for small NPs, and vice versa) (Pachon and Rothenberg 2008). In the literature, several metallic nanoparticles have been successfully synthesized from their organometallic precursors (Chaudret 2005), for example Ni nanoparticles from Ni(C8 H12 )2 (Ely et al. 1999). Furthermore, metallic nanoparticles can also be fabricated using another common method called metal vapor chemistry or commonly known as chemical vapor synthesis (CVS) (Ely et al. 1999). In this method, a wide range of materials such as organometallics, carbonyls, hydrides, chlorides, and other types of volatile compounds are used as precursors to produce their corresponding metallic nanoparticles. In general, this method is initiated by rapid evaporation of metal precursors to form the atomic vapor of the metal. The metal vapor is then condensed into a cold liquid containing stabilizing agent to form the corresponding metallic nanoparticles (Pachon and Rothenberg 2008).

6.2 Synthesis

6.2.2

Metal Oxide Nanoparticles

Among different types of metal oxide nanoparticles, the most commonly used include Fe2 O3 , SiO2 , Al2 O3 , MgO, ZrO2 , CeO2 , TiO2 , and ZnO (Fernández-Garcia and Rodgriguez 2007). Generally, fabrication of metal oxide nanoparticles can be divided into two categories based upon the nature of the phase transformation, i.e. liquid–solid and gas–solid (Fernández-Garcia and Rodgriguez 2007). Between the two, the former is considered as the simplest and the most common due to its ability to easily control morphological features such as size and shape. In liquid–solid transformation approach, synthetic methods that can be used include coprecipitation (Kandpal et al. 2014; Sitnov et al. 2022; Yagi et al. 2013), sol–gel processing (Kayani et al. 2014; Niederberger 2007), microemulsion (Malik et al. 2012; Rao and Viswanathan 2012; Sanchez-Dominguez et al. 2009), solvothermal methods (Soultanidis et al. 2012), template/surface derived methods (Kloust et al. 2015; Lan et al. 2011), hydrothermal (Khalil et al. 2014; Xiao et al. 2015), nonchemical method (Najafi et al. 2015), and supercritical water (SCW) (Adschiri et al. 1992; Kosari et al. 2017). Here, both metal salts and organometallics can be used as metal precursor. Various types of surfactants and polymers are frequently used to control particle growth, aggregation, and to induce self-assembly system. It can be done in either aqueous and/or organic solvent system (Adschiri et al. 1992; Djerdj et al. 2008). 6.2.2.1 Coprecipitation Method

Kayani et al. (2014) investigated the effect of iron oxide nanoparticles on aquathermolysis, and mixed iron oxide (magnetite) nanoparticles were synthesized at room temperature and atmospheric pressure by mixing two aqueous solutions: the first one consisted of iron chloride and iron sulfates; meanwhile, the second aqueous solution consisted of ammonium hydroxide, an alkaline earth metal hydroxide, and a surfactant (polyacrylic acid or sodium lauryl sulfate). The reducing agent is assumed to be the free radicals formed during the decomposition of water molecules at the moment of collapse of cavitation bubble during ultrasonic processing. The mixing process was carried out with continuous cavitation action by means of an ultrasonic disperser for no more than 30 minutes to obtain a sol of mixed iron oxide (Fe3 O4 ). The resulting reaction mass was treated with ion-exchange resins (cation exchange resin – KU-2-8 and anion exchange resin – AV-17-8) without turning off the cavitation effect on the mass. It is worthy to note that the addition of ion-exchange resins to the reaction mass was kept until the pH value of the mass reaches a neutral value to obtain the target product. Nanosized crystallites of iron oxide were synthesized by cost-effective coprecipitation method (Sitnov et al. 2022). X-ray diffraction (XRD) pattern indicated that the magnetic nanoparticles were pure Fe3 O4 with a spinel structure. The nanoparticles were characterized by XRD, Fourier-Transform Infrared (FT-IR) spectra, transmission electron microscope (TEM), and magnetization measurement of the nanoparticles was carried out using vibrating sample magnetometry (VSM). The cell thickness of fresh nanoparticle estimated from XRD lies in the range of 5.65–8.16 nm, whereas the particle size in the TEM was 20–22 nm. These results indicated that few grains of ferrite were aggregated to form particles of about 20–50 nm in diameter. The thermogravimetric analysis of the nanoparticle was also carried out in temperature interval from room temperature to 700 ∘ C. In the study, the particles having higher magnetization value in comparison to previously prepared ferrite were synthesized, and the method used can also be a model for the synthesis of other nanoparticles in which Fe(II) is incorporated in the particle growth process. Basuki et al. (2014) created a library of magnetic nanoparticles using in situ coprecipitation of ferrous (Fe2+ ) and ferric (Fe3+ ) ions from aqueous solutions in the presence of functional block

267

268

6 Nanoparticles for Heavy Oil In Situ Upgrading

copolymers. Three different iron oxide anchoring groups, viz, phosphonic acid, carboxylic acid, or glycerol, were incorporated into well-defined diblock copolymers of poly(oligoethylene glycol acrylate) employed to stabilize the iron oxide nanoparticles. The [copolymer]:[Fe] ratio was varied to wield control over nanoparticle diameters within the range of 7–20 nm. Synthesizes have the following steps: iron salts (1 : 2 M ratio of FeCl2 /FeCl3 at 0.065 M) were premixed with different amounts of diblock copolymer in aqueous solutions. The weight ratio between the diblock copolymer and the iron salts was varied from 4 : 1 to 1 : 10. The mixtures were incubated at 40 ∘ C for 30 minutes to ensure iron complexation to the anchoring groups (phosphonic acid, carboxylic acid, or glycerol), prior to in situ coprecipitation of the IONPs induced by the addition of 28% NH4 OH. After sonication, the IONPs were purified by dialysis in water followed by multiple contiguous centrifugation and washing steps. We observed that varying the iron/deblock copolymer ratio resulted in different color solutions (ranging from orange to black) reflecting changes to the iron oxidation state (Figure 6.3). Yagi et al. (2013) carried out the synthesis of binary magnesium–transition metal oxides, MgM2 O4 (M: Cr, Mn, Fe, Co), and MgNiO2 , was performed by calcination at relatively low temperatures of 500 and 750 ∘ C for 24 hours through inverse coprecipitation of carbonate hydroxide precursors: aqueous metallic nitrate salt solutions 100 cm3 ; 0.080 M Mg(II); 0.160 M Cr(III), Mn(II), Fe(III), Co(II); or 0.080 M Ni(II) were prepared by dissolving Mg(NO3 )26 H2 O and Cr(NO3 )39 H2 O, Mn(NO3 )26 H2 O, Fe(NO3 )39 H2 O, Co(NO3 )26 H2 O, or Ni(NO3 )26 H2 O in deionized

O P

OH O OH

+ 2 Fe3+ + Fe2+

NH4OH H2O

P(OEGA)-b-P(PAEA) P(OEGA)-b-P(PAA) P(OEGA)-b-P(PGA)

Polymer/Fe ratio

Black

IONP@P(PAEA)-b- IONP@P(AA)-b- IONP@P(GA)-bP(OEGA) P(OEGA) P(OEGA)

1:4

Black

Black

Black

1:2

Brown

Brown

Black

1:1

Brown

Orange

Black

2:1

Orange

Yellow

Black

4:1

Yellow

Yellow

Brown

Brown

Orange

Yellow

Black, Unstable

Figure 6.3 In situ coprecipitation of IONPs at different polymer-to-iron ratio. Source: Basuki et al. (2014)/from Royal Society of Chemistry.

6.2 Synthesis

water, where M is mol dm3 . A sodium carbonate solution (200 cm3 , 0.070 M Na2 CO3 ) for pH control and precipitation was also prepared. These solutions were heated to 80 ∘ C with vigorous stirring (500 rpm). The metallic nitrate salt solutions were added dropwise into the sodium carbonate precipitation solution; this procedure is called “inverse coprecipitation” because a precipitation solution is added into a metallic salt solution unconventional coprecipitation. The resulting suspensions were stirred at 70–80 ∘ C for 30 minutes and then filtered. The filtered precipitates (precursors) were rinsed with deionized water (300 cm3 ) at 80 ∘ C and air-dried for 24 hours at 80 ∘ C. In our preliminary experiment, we found that the rinsing procedure reduced the contamination of the obtained carbonate hydroxide precursors by Naþ ions to less than 1 at.%, which was evaluated by energy-dispersive X-ray analysis. The precursors were pulverized by dry ball milling (Fritsch Japan PLP-7) at 500 rpm for 8 × 15 minutes at intervals of 5 minutes, followed by calcination in air at 500 or 750 ∘ C for 24 hours. The synthesis procedure for magnesium binary oxide powders is summarized in Figure 6.4. 6.2.2.2 Sol–Gel Processing

Niederberger (2007) studied the corresponding sol–gel process that can roughly be defined as the conversion of a precursor solution into an inorganic solid by chemical means. In general, the precursor is either an inorganic metal salt or a metal organic species like a metal alkoxide or acetylacetonate. In aqueous systems, metal alkoxides are the most widely used precursors, and their chemical transformation into the oxidic network involves hydrolysis and condensation reactions. In aqueous sol–gel processes, the oxygen for the formation of the oxidic compound is Metallic nitrate salt aqueous solution (100 cm3) 0.080 M Mg(NO3)2 0.160 M Cr(NO3)3, Mn(NO3)2, Fe(NO3)3, Co(NO3)2 (or 0.080 M Ni(NO3)2) stirred at 80 ˚C, 500 rpm

Sodium carbonate precipitation solution (pH adjuster aqueous solution: 200 cm3) 0.070 M Na2CO3 stirred at 80 ˚C, 500 rpm

Added dropwise

Deionized water (300 cm3) at 80 ˚C

Carbonate hydroxide suspension stirred at 70–80 ˚C, 500 rpm for 30 min Filtered, rinsed, and air-dried at 80 ˚C Carbonate hydroxide precursor

Pulverized by dry ball milling at 500 rpm for 8 × 15 min at intervals of 5 min, followed by calcination in air at 500 or 750 ˚C for 24 h

Products (Magnesium-transition metal binary oxide powders)

Figure 6.4 Synthesis procedure for magnesium–transition metal binary oxide powders. Source: Yagi et al. (2013)/IOP Publishing.

269

270

6 Nanoparticles for Heavy Oil In Situ Upgrading

≡M—X + R—O—M≡

≡M—O—M≡ + R—X

(Eq. 1)

≡M—OR + RO—M≡ O

≡M—O—M≡ + R—O—R O

(Eq. 2)

≡M—O—M≡ + RO—CR′

(Eq. 3)

≡M—O—CR′ + R—O—M≡ 2 ≡M—OR + 2 O

–2 ROH

≡M—O—M≡ + O

(Eq. 4)

Figure 6.5 Condensation steps leading to M–O–M bonds in nonaqueous sol–gel processes (alkyl halide elimination (Eq. 1), ether elimination (Eq. 2), ester elimination (Eq. 3), and aldol-like condensation (Eq. 4)).

supplied by the water molecules. In nonaqueous systems, where intrinsically no water is present, the question of the origin of the oxygen for the metal oxide arises. Analogous to the nonhydrolytic preparation of bulk metal oxide gels, the oxygen for nanoparticle formation is provided by the solvent (ethers, alcohols, ketones, or aldehydes) or by the organic constituent of the precursor (alkoxides or acetylacetonates). The most frequently found condensation steps in the formation of a metal–oxygen–metal bond are summarized in Figure 6.5. Kayani et al. (2014) studied Fe2 O3 nanoparticles that were prepared using 0.1 M iron nitrate Fe(NO3 )3 ⋅9H2 O as a precursor, gelated by 800 ml of monohydrated citric acid, and singly distilled water was used as solvent. 0.1 M iron nitrate (Fe(NO3 )3 ⋅9H2 O) was added to citric acid solution dropwise with vigorous stirring. The sol was then heated to 70 ∘ C while stirring until the gel was formed and the contained water was evaporated. The formed gel was then dried by putting it in oven at 100 ∘ C for an hour. The dried gel was then annealed at temperatures of 400 and 1000 ∘ C. First, 4.5 ml of titanium isopropoxide was taken in a 250 ml beaker. Then, 9 ml of glacial acetic acid was added followed by the addition of 100 ml water dropwise. A transparent sol was obtained within 15 minutes. Then, magnesium methoxide slurry was added to the above sol with continuous stirring. A light yellow-colored sol was obtained after five hours. The sol was kept in an oven at 80 ∘ C for gelation, and the gel obtained was then dried at 100 ∘ C to get a powder. The as-prepared sample powders were grounded and subjected to calcination at 500, 700, and 900 ∘ C to get TiO2 −MgO-mixed metal oxide nanoparticles (Djerdj et al. 2008). 6.2.2.3 Microemulsion Method

Li et al. suggested that microemulsion was a better way to synthesize the nanonickel particles. In this novel process, an oil phase (e.g. cyclohexane or methylcyclohexane) is used to prepare a nickel nanocatalyst, which can be used as a hydrogen donor to further improve the properties of the heavy oil (Wei et al. 2007). According to Song et al. (2005), the microemulsion was prepared with methylcyclohexane, water, surfactant AEO 9, and n-octanol with the mass ratio of 50 : 10 : 1 : 0.2. The preparation of catalyst in microemulsion was carried out in an atmospheric environment at 30 ∘ C, methylcyclohexane, water, surfactant AEO 9, and n-octanol, that was an auxiliary agent of the surfactant, were mixed using the magnetic stirrer until well-proportioned solution was formed. Then a specified quantity of Ni(NO3 )2 was added into the solution and mixed by means of the magnetic mixer until Ni(NO3 )2 was completely dissolved. Subsequently, by adding specified amounts of LiBH4 , the Ni ions in the water pools of the microemulsion were reduced. In order to complete reduction of the Ni ion in the microemulsion, the molar ratio of [LiBH4 ]:[Ni] was approximately 5 : 1. Finally, the microemulsion with 20 gNi/l was obtained Noorlaily et al. (2013) synthesized NiO particles with an average size of 15.4 nm from NiCl2 ⋅6H2 O via the ethylene glycol pathway. These NiO nanoparticles displayed excellent uniformity, spherical

6.2 Synthesis

morphology, and good dispersion (rated at a 65 nm diameter). Addition of NiO nanocatalyst has led to a viscosity reduction of 22% during aquathermolysis when compared with heavy oil alone. A dependence on the particle size was noticed with calcination temperature. Crystal size increased with increased temperature (4.77, 9.08, and 15.26 nm for 27, 127, and 227 ∘ C, respectively) due to the excess energy that is delivered to the crystals, thus promoting their growth. It was estimated that the activation energy for NiO growth during calcinations is about 21.33 kJ/mol. Chen et al. (2019) developed an in situ synthesis strategy for preparing well-dispersed CuO nanoparticles as aquathermolysis catalyst for viscosity reduction in Shengli heavy oil (China). A Cu(OH)2 -contained microemulsion was employed as a carrier to disperse the precursor Cu(OH)2 to the heavy oil phase. Under aquathermolysis condition (240 ∘ C, 2.5 MPa of N2 ), the Cu(OH)2 precursors would first be converted in situ to well-crystallized and size-homogeneous CuO nanoparticles naturally. MeO nanoparticles can be produced in (w/o) microemulsions using metal chloride, nitrate, or oxalate as precursors (Abdrabo and Husein 2012). However, unlike some claims made by Mihaly et al. (2011), NiO is not formed in the (w/o) microemulsion, but rather its intermediaries are precipitated in the microemulsion. NiO nanoparticles are only formed ex situ by the calcination of the (w/o) microemulsions (Mihaly et al. 2011). 6.2.2.4 Preparation of Catalysts in Supercritical Water

Synthesis of MeO nanoparticles can be also achieved in situ in SCW (Kosari et al. 2017). Five different metal oxide nanoparticles’ (NiO, CuO, ZnO, Co2 O3 , and Cr2 O) rigid cylindrical batchwise reactor (stainless steel, type 316 l) was used for nanoparticle synthesis. The capacity of the reactor was about 100 cm3 , and it was designed to sustain the working pressure and temperature up to 610 atm and 500 ∘ C, respectively. As a safety hint, the reactor was merely loaded to one-third of its capacity and was set in a preheated furnace, which was heated to the desired temperature, i.e. 500 ∘ C. Elapsing three hour reaction time, the reactor was then removed from the furnace and was quickly quenched in cold water. The formed nanoparticles inside the reactor were repeatedly rinsed with distilled water, and then all the nanoparticles were separated from the remnant aqueous solution using a high-speed centrifuge at 14 000 rpm for 15 minutes. Subsequently, the nanoparticles were washed three times with distilled water and were then dried at the room temperature.

6.2.3

Carbon Nanotubes (CNTs)

During the past several years, different types of systematic methods for CNTs synthesis have been extensively reported. Depending on the type of CNTs, their application, and their characteristics, each method has its own advantages and disadvantages. These methods have been comprehensively discussed and reviewed in many articles (Arora and Sharma 2014; Prasek et al. 2011; Szabó et al. 2010). Ever since the discovery of CNTs, several techniques have been explored for large-scale and high-quality products. The most widely used methods to produce CNTs explained in this section are electric arc discharge (Ebbesen and Ajayan 1992; Journet et al. 1997a), laser ablation (Thess et al. 1996), and chemical vapor deposition (CVD) (Terrones et al. 1997). Electric arc discharge is one of the first methods used to produce CNTs (Ebbesen and Ajayan 1992; Journet et al. 1997b). The arc discharge method employs two graphite electrodes (6–12 mm in diameter) that are separated by a short distance (1–4 mm) inside a chamber filled with an inert gas. A current of about 50–100 A is passed though the electrodes, and carbon atoms are ejected from the positive electrode (anode) and deposited on the negative electrode (cathode). As a result, the length of the anode decreases as the CNTs start forming on the cathode. Carbon is vaporized from the graphite anode in the form of crystallites, which generate small carbon clusters (mainly C3 )

271

272

6 Nanoparticles for Heavy Oil In Situ Upgrading

(Prasek et al. 2011). Next, these carbon clusters rearrange themselves into a tubular shape forming the MWCNTs, which drift toward the cathode and deposit on its surface. In 1995, Smalley and his coworkers introduced a very promising approach to produce CNTs called the laser ablation method (Guo et al. 1995). In a laser ablation or evaporation method, a powerful laser is used to ablate a carbon target in a hot helium (He) or argon (Ar) atmosphere. As the

Electrical oven Pressure gauge (0–10 bar) Tank for the mixture of gases Flow meters

Pressure gauge (–1–0 bar)

Catalyst on holder

Gas output

Gas output A B V

mV

Vacuum

Thermo couple A

Quartz tube

Variable voltage

Hydrocarbon

Hydrogen (99.9999%)

Carrier gas

Liquid hydrocarbon Thermostal

Flow meter

(a)

Gas input

Catalyst

Solid hydrocarbon (b)

Figure 6.6 Schematic diagram of a CVD setup utilizing three different types of hydrocarbons: gas, liquid, and solid. (a) shows scheme the injection of gases and liquid hydrocarbons. (b) shows an electric oven with a catalyst bed. Source: Dervishi et al. (2009)/Taylor & Francis.

6.3 Characterization

graphite target inside a furnace is heated up at about 1200 ∘ C, a pulsed laser beam incident on the target starts evaporating carbon from the graphite. The carrier gas sweeps the carbon atoms from the high-temperature zone to a cold copper collector on which they condense into nanotubes. In order to generate SWCNTs using the laser ablation technique, it is necessary to impregnate the graphite target with transition metal catalysts (Terrones et al. 1997). It is experimentally found that the SWCNT growth time in this technique is only a few milliseconds long. Generally, along with SWCNTs (long bundles) (Saito et al. 1992) and MWCNTs (closed-ended) (Guo et al. 1995), fullerenes, amorphous carbon, and other carbon by-products are produced when using the laser ablation technique. MWCNTs produced by this method have a number of layers varying from 4 to 24 and an inner diameter ranging between 1.5 and 3.5 nm (Prasek et al. 2011). The morphology and the properties of CNTs are highly influenced by many different parameters such as light intensity, furnace temperature, type of hydrocarbon and carrier gas, and the flow rate of different gases. For example, when the furnace temperature is below 800 ∘ C, no CNT growth is observed, whereas a maximum SWCNT yield is obtained at about 1200 ∘ C. Unfortunately, the laser ablation technique is very expensive because it involves high-purity graphite rods and high-power lasers (Terrones et al. 1997). CVD is a relatively slow method that produces long CNTs in large quantities (Kong et al. 1998; Ren et al. 1998). The hydrocarbon source is heated at high temperatures, typically between 700 and 1000 ∘ C, inside a quartz tube in the presence of catalytic systems. CNTs are produced from the thermal decomposition of the carbon-containing gas molecules on desirable catalytic systems. The latter present in the substrate provide nucleation sites for the nanotube growth. At high temperatures, once the hydrocarbon decomposes into hydrogen and carbon, carbon atoms dissolve and diffuse into the metal surface, and rearrange themselves into a network containing hexagons of carbon atoms and finally precipitate out in the form of CNTs. Once the metal surface is covered by amorphous carbon and its surface is “poisoned,” the carbon atoms cannot come into contact with the metal catalyst, which would result in the termination of CNT growth. The hydrocarbon source exploited in the CVD method can be in a gas state such as acetylene, methane, and ethylene, a liquid state such as benzene, alcohol, and hexane, or a solid state such as camphor, naphthalene, and many more. Figure 6.6 shows a schematic diagram of the CVD method exploiting the hydrocarbon source in any state (gas, liquid, and solid).

6.3

Characterization

The methods that are used to characterize the catalysts are quite diverse, for example XRD, Mössbauer spectroscopy, scanning electron microscopy (SEM), TEM, FT-IR, and energy-dispersive X-ray (EDX mapping). XRD, Mössbauer spectroscopy, FT-IR, and EDX mapping methods can be used to determine the composition. Morphology and perception research use SEM, TEM, and other methods. For example in Figures 6.7 and 6.8, XRD patterns for the fresh nanoparticles obtained by SCW and coprecipitation were recorded (Kosari et al. 2017; Sitnov et al. 2022). These patterns were compared with the references for the produced metal oxides according to the X-pert High Score PDF code, and thereby relevant compounds were detected. Accordingly, it is revealed that all the nanoparticles were produced in a single phase under supercritical condition. Moreover, the XRD broad peaks are indicative of the particles with nanometric size. Using XRD pattern, the average size of sample’s crystals is measured by the Debye Scherrer formula as follows: dXRD = 0.089𝜆∕𝛽 cos 𝜃

273

6 Nanoparticles for Heavy Oil In Situ Upgrading

ZnO Zn ZnS

Intensity (a.u.)

NiO

Cu2O Cu

Co2O3 Cr2O3

20

25

30

35

40

45 50 2θ (°)

55

60

65

70

Figure 6.7 X-ray diffraction patterns of spent nanocatalysts after cracking reaction in SCW (Condition: T = 450 ∘ C, t = 60 minutes, water/oil ratio (g/g): 8/1, and catalyst/oil ratio (g/g): 1/5). Source: Kosari et al. (2017)/with permission of Elsevier.

2000

Magnetite (Fe3O4)

1600 Counts

274

1200

800

400 10

20

30

40

50

60

2θ (°) Figure 6.8 X-ray diffraction diagram of the obtained Fe3 O4 catalyst, which prepared coprecipitation method. Source: Sitnov et al. (2022)/with permission of Elsevier.

in which dXRD denotes the crystalline size and stands for the X-ray wavelength (0.15418 nm). The line broadening at half the maximum intensity (FWHM) minus the instrumental line broadening is named, which sometimes is referred to as Δ(2𝜃), and 𝜃 is the Bragg angle. The average crystalline size of the obtained nanoparticles was estimated by mentioned procedure and is presented in Table 6.1.

6.3 Characterization

Table 6.1

Average crystalline size of the transition metal oxide nanoparticles. NiO

CuO

Cr2 O3

ZnO

Crystal size (nm)

12

14

24

19

17

100

120

90 100

80 70

80

60 60

50 40

40

30 20

20

10

D (nm)

0 0

Figure 6.9

25

50

75

0 100 125 150 175 200 225 250

Area particle size distribution cumulative (%)

Co3 O4

Area differencial particle size distribution (%)

Nanoparticles

Particle size distribution of the obtained catalyst.

In Figure 6.8, the obtained product was characterized by XRD analysis, which showed a double mixed iron oxide(II, III). To study the magnetite suspension’s polydispersity in the aqueous phase, a bimodal particle size distribution was applied by using an ultrasonic particle analyzer Zeta-APS (USA). The obtained data indicate that the synthesized particles are slightly more than 100 nm on average. However, the obtained data indicate as well that more than 30% of the obtained particles range in the nanometer scale (Figure 6.9). In order to acquire more information about the state of various metal oxide species, FT-IR spectra of the synthesized nanoparticles were recorded. FT-IR spectra of all samples consist of broadband and intense IR absorption in the region of 2800–3650 cm−1 centered at around 3400 cm−1 corresponding to O–H stretching vibration; a band at c. 1350 cm−1 is attributed to the absorption of atmospheric CO2 on the metallic cation; and a sharp band at c. 1630 cm−1 is ascribed to the bending mode of the OH group of the adsorbed water molecules. The absorption peak situated approximately at 490 cm−1 can be attributed to the Zn–O stretching mode in the ZnO lattice. In the spectrum of cobalt oxide, the absorption peaks around 665 and 566 cm−1 are assigned to the vibration of Co–O. In the spectrum of nickel oxide, absorption band at c. 459 cm−1 is associated with the Ni–O stretching vibration (Noorlaily et al. 2013). FT-IR spectrum of copper oxide exhibits a strong absorption broadband approximately located at 530 cm−1 due to the Cu–O vibrations. Two adsorption bands in the Cr2 O3 spectrum centered approximately at 540 and 620 cm−1 correspond to the vibration of Cr–O. The FT-IR spectra exhibit no adsorption bond pertaining to nitrate and/or hydroxynitrate intermediates. These results, in turn, confirm the formation of highly pure metal oxide nanoparticles using SCW method supporting those data obtained from the XRD patterns.

275

6 Nanoparticles for Heavy Oil In Situ Upgrading

400

Counts

300 200

Fe3O4 Cu0.86Fe2.14O4 FeCu4 CuS2

100 0 10

210

Fe1.7O4

180

Fe2O3 Ni3S2

150 Counts

276

20

30 40 2θ (coupled 2θ/θ) WL = 1.54060 (a)

50

20

30 40 2θ (coupled 2θ/θ) WL = 1.54060 (b)

50

60

120 90 60 30 0 10

Figure 6.10 X-ray analysis results of the obtained nanoparticles of iron and copper spinel (a), iron and nickel spinel (b). Source: Sitnov et al. (2018)/with permission of Elsevier.

In addition, XRD can be used to study the formation of the active form of the mono-and bi-metallic catalyst in situ from oil-soluble precursors, such as tallates of transition metal (Sitnov et al. 2018). In Figure 6.10, X-ray analysis results of the obtained nanoparticles after thermobaric effect on a mechanical mixture of iron and copper tallates (a), iron and nickel tallates (b) are shown. The investigated precursor samples transform into nanoparticles of nonstoichiometric spinel ferrites. The concentration of metal sulfides is negligible. In the spectrum of the sample, which was obtained as a result of hydrothermal influences on a mixture tallates of iron and copper, lines from magnetite (Fe3 O4 ), copper spinel ferrite Cu0.86 Fe2.14 O4 , FeCu4 compounds, and copper sulfide (CuS2 ) were observed (Figure 6.10a). A sample based on iron and nickel (see Figure 6.10b) is characterized by the presence of maghemite (γ-Fe2 O3 ), nickel spinel ferrite (Ni1.43 Fe1.7 O4 ), and nickel sulfide (Ni3 S2 ). The diversity of the catalyst nanoparticles composition reflects the complex process of their formation in the hydrothermal environment. The process of aquathermolysis at high pressure appears in this case as a hydrothermal synthesis of spinel ferrites from the formed oxides. In an ideal case, chemical compounds of the following type should be formed: NiFe2 O4 and CuFe2 O4 . However, at high temperatures, the composition of these compounds may deviate from the above stoichiometry, enriched either with iron oxide or with nickel or copper oxide, which can be seen from the XRD spectra (see Figure 6.10). Nonstoichiometry of spinels at high temperatures can be associated with thermodynamically useful formation processes exactly like a structure of mixed oxides.

6.3 Characterization

Table 6.2 Parameters of hyperfine interactions (effective local magnetic field on 57Fe nuclei (Heff ), isomer shift (𝛿 Fe ), quadrupole splitting (Δ), and partial contributions from components of the Mössbauer spectra). Parameter

H eff , ±1 k

𝜹Fe , ±0.01 mm/s

S1

460

0.66

0.01

S2

490

0.26

−0.02

S3

487

0.30

0

36.2

S4

517

0.38

−0.04

17.9

D1

—

0.33

0.32

Figure 6.11 Mössbauer transmission spectra of the sample nanoparticles of iron and nickel spinel (S1, S2 correspond to magnetite Fe3 O4 ; S3, S4 – NiFe2 O4 ; D1 – fine-dispersed particles of iron oxide at room temperature).

𝚫, ±0.01 mm/s

S, ±0.5%

31.7 9.70

4.50

v (MM/c)

Transmission (%) b

S2 S1

100 D1 S3

S4

96 –9

–6

–3

0

3

6 9 v (MM/c)

The results of X-ray spectroscopy may be confirmed by data obtained using Mössbauer spectroscopy. Figure 6.10 shows the Mössbauer spectrum of the sample catalyst after thermobaric effect on a mechanical mixture of iron tallate and nickel tallate. The parameters of hyperfine interactions calculated from specters are presented in Table 6.2. The Mössbauer spectrum of the sample based on the mechanical mixture of iron tallate and nickel tallate, shown in Figure 6.11, consists of five partial components (Table 6.2). The values found for the parameters of hyperfine interactions of the spectral components indicate a structurally and magnetically inhomogeneous multiphase state of the sample. The sample consists of three phases such as magnetite, nickel spinel ferrite and, presumably, iron oxide in a finely dispersed form. During the thermal treatment in an autoclave (in reservoir conditions), the precursors of iron and nickel tallates decompose and form three phases: magnetite FeO⋅Fe2 O3 , complex iron oxide and nickel spinel ferrite NiFe2 O4 , and a superparamagnetic fine phase of iron oxides, and probably nickel. EDX mapping of metal nanoparticles is formed because of thermal impact on a mechanical mixture of iron and copper tallate as shown in Figure 6.12. As it can be seen from Figure 6.12,

Figure 6.12 EDX mapping of metal particles formed as a result of thermal impact on a nanoparticle of iron and copper spinel.

277

278

6 Nanoparticles for Heavy Oil In Situ Upgrading

(a)

(b)

(c) Figure 6.13 The distribution of elements in the catalysts isolated after thermocatalytic treatment for (a) 1 hour; (b) 2 hours; (c) 4 hours; (d) 6 hours; (e) 12 hours.

6.3 Characterization

(c)

(d)

(e) Figure 6.13

(Continued)

overlapping of sections occurred which characterize the distribution of elements of iron, copper, and oxygen. This indicates the formation of copper spinel ferrite (CuFe2 O4 ). The results of EDX mapping also indicate the formation of mixed oxides of copper and iron, along with the formation of individual oxides, such as magnetite. The distribution of elements in the catalysts isolated after thermo-catalytic treatment for 12 hours is presented in Figure 6.13. The results indicate the ambiguous behavior in the weight compositions

279

280

6 Nanoparticles for Heavy Oil In Situ Upgrading

of Ni, C, O, and S elements. During various thermo-catalytic treatment durations, the active form of the catalyst was observed in the form of nickel sulfide. The adsorption of coke-like compounds on the catalyst nanoparticles results in the increase of the content of carbon element. The content of Na and Al metals as well as nonmetals such as O, Si, and Cl, which exists in the composition of rocks, was increased (Vakhin et al. 2021). SEM and TEM analyses are used to study the morphology and elemental composition of catalyst particles. Catalyst nanoparticles of NiO are shown in Figure 6.14 (Kosari et al. 2017). Iron oxide with 70 nm size is shown in Figure 6.15 (Sitnov et al. 2019). The adsorption of catalytic particles on

Figure 6.14 Scanning electron microscopy (SEM) images of NiO nanoparticles. Source: Kosari et al. (2017)/from ELSEVIER.

200 nm

Figure 6.15 SEM image of the iron oxide catalyst particles surface. Source: Sitnov et al. (2019)/from IOP Publishing, CC BY 3.0.

6.3 Characterization

the rock’s mineral grains is shown in Figure 6.16 (Sitnov et al. 2022). The catalyst nanoparticles of less than 100 nm are shown in Figure 6.17 (Sitnov et al. 2018). The specimen for TEM (Figure 6.18) detection was prepared by depositing a 5 μl microemulsion containing the metal nickel onto a gold grid coated with carbon. Then, the specimen was dried in a vacuum oven at 200 ∘ C in nitrogen atmosphere for 24 hours to remove water, organic solvent, and surfactant. The size of the catalyst particles in the treated specimen was measured by H700H type TEM. The nanoparticles obtained by this method were spherical and had an average diameter of 6.3 nm (Bera and Babadagli 2017). Yeletsky et al. (2019) studied the adsorption of NiMo/ SiO2 nano-catalysts on carbonate rock at 425 ∘ C. The size of catalyst particles is shown in Figure 6.19. Figure 6.16 SEM analysis of the adsorbed Fe3 O4 nanoparticles on rock surface. Source: Sitnov et al. (2022)/from ELSEVIER.

800 nm Figure 6.17 SEM image of the formed particles as a result of the thermal impact on the mechanical mixture of iron and copper spinel. Source: Sitnov et al. (2018)/from ELSEVIER.

281

282

6 Nanoparticles for Heavy Oil In Situ Upgrading

Figure 6.18

TEM pattern of nanonickel catalyst.

Figure 6.19 TEM images from fresh NiMo/SiO2 . Source: Yeletsky et al. (2019)/from ELSEVIER.

6.4

Catalytic Activity

The most striking properties of nanoparticles are their higher surface area which reflects their high efficiency in the processes of C—S—C bonds destruction in resins and asphaltenes molecules. This provides usually a decrease in heavy oils’ viscosity (Liu et al. 2015; Wen et al. 2007), as well as the ability to penetrate deep into the pore space of the reservoir rock (Maity et al. 2010; Olvera et al. 2014). Besides, some studies have shown that nanocatalysts can reduce the risk of coke formation during oil in situ upgrading at aquathermolysis stages (Zhang et al. 2007). Different nanoparticles were reported and discussed in detail for each metal (Simão et al. 2022).

6.4.1

Monometallic Catalysts in Oil Upgrading

A total of 17 different individual metals used as nanoparticles for in situ oil upgrading were found in the literature, which is summarized in Figure 6.20. They range from alkaline earth metals (Mg, Ca) to lanthanides (Ce), including also transition metals (Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Zr, Mo, and W), semimetals (Si), and posttransition metals (Al, Sn). Herein, metals are organized in the ascending order of atomic number.

6.4 Catalytic Activity

Figure 6.20 Overview of the metals contained in the reported nanoparticles or supported nanoparticles with applications in oil upgrading.

283

284

6 Nanoparticles for Heavy Oil In Situ Upgrading

6.4.1.1 Magnesium

MgO has been studied for asphaltene adsorption and catalytic oxidation to achieve heavy oil upgrading (Nassar et al. 2011a). The calculation of the Langmuir adsorption isotherms was reported for asphaltenes onto nanoparticle surface for MgO (basic nature and 545 ∘ C in lighter oil with less generation of coke and a higher removal of sulfur. The obtained upgraded oil also showed a much lower viscosity (Galarraga and Pereira-Almao 2010). The catalytic activity of monometallic MoS2 and bimetallic MoS2 and Al2 O3 nanoparticles was evaluated together with two different trimetallic nanoparticles: (i) Ni3 S2 , MoO2 , and Al2 O3 ; and (ii) Co9 S8 , MoO2 , and Al2 O3 . To obtain the trimetallic catalyst, first the aluminum precursor was added (ratio Al/ΣMe (wt.%) = 0.14; and ratio Mo/(Ni or Co) = 1), and later the precursors of the

6.4 Catalytic Activity

remaining metals were added together. The conversion of the 500 ∘ C fraction was similar to that of bimetallic catalyst (48.9% and 49.0% for NiMoAl and CoMoAl catalysts, respectively, versus 48.5% for MoAl); but the conversion of the 180–350 ∘ C fraction was from 12- to 20-fold higher in trimetallic catalysts (6.3% and 10.3% for NiMoAl and CoMoAl catalysts, respectively, versus 0.5% for MoAl). The production of coke was low for all the nanoparticles considered, as it remained always below 2% (Kadiev et al. 2016). 6.4.2.3 Supported-Nanoparticles Coated with Nanoparticles of a Different Metal

Different studies have reported the use of metallic, semimetallic, polymer, inorganic, and biogenic nanoparticles coated at the same time by smaller nanoparticles of one or several metals. In function of the nature of the support, these nanoparticles can be divided into silica, alumina, carbon, zeolite, and biogenic particles. Some of them, especially those supported by semimetals oxides like alumina or silica, could have been considered also as bimetallic catalysts due to the reported catalyst activity of some uncoated cores. 6.4.2.3.1

Silica-Supported Nanoparticles

Silica-supported nanoparticles are also of great interest in the field of oil upgrading for improving the mobility of oil in reservoir, as proved by various research works reported in the literature (Franco et al. 2013a, b; Montoya et al. 2016). For instance, fumed silica nanoparticles coated with nickel(II) oxide (NiO) or palladium(II) oxide (PdO) nanoparticles with diameters of 2.9 and 4.1 nm, respectively, have ability to perform cracking of the asphaltenes contained in heavy oil at milder temperatures than those used in thermal cracking with a much lower generation of coke during the catalytic process. The palladium oxide-coated fumed silica nanoparticles showed better activity than the nickel oxide-coated ones. With PdO–SiO2 nanoparticles, the generation of coke reduced to 0.06%, 1000-fold lower than the 61.2% determined in thermal cracking (Montoya et al. 2016). The use of silica nanoparticles coated at the same time by PdO and NiO has been reported to be more efficient catalyst in asphaltene cracking than the silica nanoparticles coated only with NiO or PdO. Even then, no coke was generated during the catalytic cracking process. NiO and PdO nanoparticles showed an average diameter of 1.3 and 2.2 nm, respectively (Montoya et al. 2016). Synthesized silica nanoparticles coated with nickel oxide nanoparticles showed a strong capacity to adsorb asphaltenes onto their surface (Franco et al. 2013a, b). The coated nanoparticles were prepared by the washing of previously synthesized SiO2 nanoparticles (90 nm of diameter) with a 5 wt.% or a 15 wt.% nickel nitrate solution, rendering NiO nanoparticles (SNi5 and SNi15 hybrids, respectively). In the two hybrids, the NiO nanoparticles had a diameter of 15 nm (Franco et al. 2013b). The SNi15 hybrids showed the highest capacity to adsorb asphaltenes, according to the calculated Langmuir adsorption curves (N ads,max of 16.376 mg/m2 ), also being higher than the value determined for crystalline SiO2 gel nanoparticles (13.158 mg/m2 ). Nevertheless, SNi5 hybrids showed a lower adsorption (8.961 mg/m2 ) that the parental SiO2 nanoparticles. Thus, in this SNi5 hybrid, the coating reduced the adsorption capacity (Franco et al. 2013b). Furthermore, the asphaltene adsorption properties of different PdO-coated and PdO–NiO-coated SiO2 were evaluated (Franco et al. 2013a). In this case, uncoated SiO2 nanoparticles showed the lowest adsorption capacity, and the Pd–Ni/Si hybrid containing 1% Pd, 1% Ni, and 98% Si exhibited the highest asphaltene adsorption capacity, followed by a PdO–SiO2 hybrid (2% and 98%, respectively). In another study, the following nanoparticles (with a ±0.2 nm error in diameter) supported over fumed silica gel were studied: (i) NiO, 2.9 nm (2 wt.%), (ii) PdO, 4.1 nm (2 wt.%), and (iii) NiO (1 wt.%) and PdO (1 wt.%), 1.3 and 2.2 nm, respectively (Franco et al. 2014). In addition to the

295

296

6 Nanoparticles for Heavy Oil In Situ Upgrading

adsorption effects reported in (Franco et al. 2013a), these silica-supported nanoparticles significantly increased the asphaltene gasification through the reductions of the asphaltene association to the catalyst surface, of the temperature at which gasification of asphaltenes is produced, of coke formation and of catalyst deactivation. The effect was more pronounced in Pd–Ni hybrids, suggesting a possible synergy of the action of the two metals. Finally, the nanoparticles can reduce the production of methane and increase the emission of CO. 6.4.2.3.2

Alumina-Supported Nanoparticles

Regarding alumina-supported Pd nanoparticles, they enabled the production of oil with not only a higher API gravity increase than the abovementioned biogenic palladium nanoparticles (Hart et al. 2016) but also are more vulnerable to be deactivated by coke. The use of bimetallic MoS2 and Al2 O3 nanoparticulated catalysts generated in situ for oil upgrading was reported (Kadiev et al. 2016). The metallic precursors (ammonium paramolybdate and aluminum nitrate) could be added simultaneously or sequentially (first Al, later Mo). To equal Al/ΣMe ratio (in wt.%) of 0.14, the sequential addition of the Al/Mo precursors was more profitable in terms of conversion of 500 ∘ C fraction (48.5% versus 41.0%) than the simultaneous addition of the precursors. Moreover, sequentially obtained bimetallic MoS2 /Al2 O3 nanoparticles also improved the conversion of MoS2 nanoparticles (40.8%). Nevertheless, in the case of the simultaneous addition, when the content of Al was a bit higher (ratio of 0.8), the conversion was increased to 62.2%. A further increase of this ratio until 4.5 decreased the conversion to 52.1% (Kadiev et al. 2016). Nanoparticles of nickel oxide supported onto nanoparticulated alumina could effectively adsorb asphaltenes (Franco et al. 2013b, c). The alumina nanoparticles used as support had a diameter of 35 ± 4 nm, and the coating of NiO nanoparticles diameter varied as a function of the amount of nickel nitrate added to prepared nanoparticles. Impregnation of alumina nanoparticles with 5 wt.% nickel nitrate solution rendered NiO nanoparticles of 16 nm of diameter (AlNi5 hybrid), and with a 15 wt.% solution, of 29 nm of diameter (AlNi15 hybrid). AlNi15 hybrid showed the highest capacity to adsorb asphaltenes and could complete their sorption rapidly at around two minutes (Franco et al. 2013b, c). N ads,max values for AlNi5 and AlNi15 were 2.51 and 10.16 mg/m2 , respectively; the latter being very promising, although a bit lower than those of silica hybrids, as reported above. Both alumina-coated hybrids showed higher adsorption capacity than the uncoated alumina nanoparticles. The same work reported the simultaneous coating of alumina nanoparticles with PdO and NiO, but in this case, the adsorption capacity is much lower (N ads,max of 0.411 mg/m2 ) (Franco et al. 2013b). 6.4.2.3.3

Carbon-Supported Nanoparticles

Guo et al. (2017) showed that the viscosity reduction of a heavy oil was achieved after a two-hour upgrading at 300 ∘ C in presence of zeolite or of different carbon nanosized catalysts (ketjenblack carbon, CNTs, and graphene nanoplatelets) coated or uncoated with NiO nanoparticles with a particle size around 12 nm. Ketjenblack produced the highest oil viscosity reduction (65%) among the four supports evaluated, whereas NiO-ketjenblack nanoparticles were more efficient than NiO-coated catalysts in terms of oil viscosity reduction (75%). According to the experimental data obtained from Fourier-transform ion cyclotron resonance mass spectrometry, the viscosity reduction and the catalytic upgrading can be explained through the conversion of carboxylic acid-containing large molecules into smaller and more saturated molecules (Guo et al. 2017). Another study reported the use of Ni–Mo/-activated carbon (Ni–Mo/AC) hydrotreating catalysts (Shi et al. 2012). Final catalyst composition was 15 wt.% MoO3 and 5 wt.% NiO. The size of the coating nanoparticles ranged from 30 to 50 nm. SEM measurements revealed the fixation

6.4 Catalytic Activity

and dispersion of the nanoparticles to the highly porous surface of the activated carbon. These novel catalysts performed a better hydrotreating performance in the upgrading of heavy vacuum gas oil (HVGO) than two commercial hydrotreating catalysts (Ni–Mo/γ-Al2 O3 ). The catalytic activities evaluated were the hydrodesulfurization (HDS) and the hydrodenitrogenation (HDN). For example, the novel catalysts could remove a 70–80% of the sulfur, whereas the commercial catalysts removed a 40–60% of the sulfur. With respect to HDN, the conversion of the novel Ni–Mo/AC catalyst was in the range of 23–30%, while the conversion of the commercial catalysts ranged from a 7% to a 17%. The generation of carbon residue was also lower: 0.01% or lower for Ni–Mo/AC, compared with 0.06–0.11% of the commercial catalysts (Shi et al. 2012). On the other hand, carbon nanofibers were coated with Ni–Co nanoparticles (Remón et al. 2016). The appropriate carbon support was impregnated with the solutions of the metal nitrates, and nanoparticles were obtained by basic precipitation with ammonium hydroxide followed by a drying and a thermal treatment at 450 ∘ C. Three different nanoparticles were obtained depending on the temperature of the preparation and pretreatment of the carbon nanofibers, rendered in a different functionalization of the support: (i) NiCo-CNFf (boiling temperature of HNO3 , and the concentration of Ni and Co is 17.4 and 2.7 wt.%, respectively); (ii) NiCo-CNFf –600 (boiling temperature of HNO3 followed by a thermal treatment at 600 ∘ C, and the concentration of Ni and Co is 18.5 and 2.3 wt.%, respectively); and (iii) NiCo-CNFf –900 (boiling temperature of HNO3 followed by a thermal treatment at 900 ∘ C, and the concentration of Ni and Co is 16.7 and 2.1 wt.%, respectively). These catalysts were used in SCW upgrading of bio-oil. The NiCo-CNFf –600 and the NiCo-CNFf –900 significantly improved the liquid content of the bio-oil, and in addition, NiCo-CNFf –900 also reduced the solid fraction compared with the thermal treatment in the absence of catalyst (Remón et al. 2016). Carbon-supported nanoparticles are used not only in oil upgrading but also in the adsorption of the asphaltenes contained in oils. They can be also employed for additional catalytic reactions, such as hydrodesulfuration. For instance, among the supported molybdenum catalysts, graphite coated by MoS2 nanoparticles (36 Å) has shown catalyst activity for hydrodesulfuration. The interaction between MoS2 and pristine highly oriented pyrolytic graphite (HOPG(0001)) seems to be too weak to allow for a high dispersion of the MoS2 within the substrate. The morphology of MoS2 nanoclusters depends strongly on the temperature, e.g. it varies from single-layer clusters at 727 ∘ C to stacked multilayer cluster at 927 ∘ C. From a catalytic point of view, the existence of hydrogen adsorbates at the cluster edges is important given that both the adsorption of the S-containing molecule and dissociation of H2 are required to facilitate the HDS reaction (Kibsgaard et al. 2006). 6.4.2.3.4

Zeolites as Support

Zeolites can be employed also as supports of metal oxide nanoparticles with oil upgrading or cracking purposes (Varzaneh et al. 2015). A commercial SAPO-34 zeolite with a specific surface area of 573 m2 /g and a pore volume of 0.289 cm3 /g was impregnated with Ce(NO3 )3 or with Zr(NO3 )3 solutions. Subsequent evaporation of the solvent, drying, and calcination rendered the final Ce/SAPO-34 and Zr/SAPO-34 catalysts with a 2–15 wt.% and 2–8 wt.% content of Ce and Zr, respectively. The catalysts were used in the cracking of naphtha. The results pointed out that the addition of a small amount of Ce or Zr nanoparticles to the zeolite structure improved its catalytic activity, compared with that of the zeolite alone for the thermal cracking. This improvement was reflected in the enrichment of the ethylene, propylene, and olefin content of the naphtha, and in the reduction of the percentage of >C5 hydrocarbons. The best effect was observed for the 2% Ce and 8% Ce-containing catalysts, followed by the 2% Zr-containing zeolite (Varzaneh et al. 2015).

297

298

6 Nanoparticles for Heavy Oil In Situ Upgrading

Zeolite Y nanoparticles exchanged with La3+ ions were used as support of SiO2 and Al2 O3 nanoparticles to obtain a novel catalyst (Karami and Mahinpey 2016). The aim of the lanthanum ions is to replace the sodium ions, which affect negatively the stability and catalytic activity of zeolite Y nanoparticles in hydrothermal treatments. Zeolites of different sizes were evaluated (150, 450, and 1500 nm). It turned out that those with a particle size of 28 nm exerted the best catalytic effect in the upgrading of 1,3,5-triisopropylbenzene (chosen as a representative asphaltene compound). The catalyst with the lowest size enabled to obtain an upgrading product with a higher content of light alkenes and a lower content of benzene, coke, and heavy compounds (Karami and Mahinpey 2016). Alternatively, the use of SO4 2− /Zr-silicalite-1 zeolite catalysts in the upgrading of heavy oil was reported (Su et al. 2017). This catalyst was prepared by the mixture of zirconium butoxide and tetramethylorthosilicate (in different wt.% ratios), then it was treated with tetrapropyl ammonium hydroxide and tetraethylorthosilicate. An autoclave treatment (210 ∘ C, two days) rendered a product that was impregnated in sulfuric acid and later calcined at 550 ∘ C to remove the organic residues. The catalyst was used in the aquathermolysis upgrading of heavy oil from Shengli oil field (China). The complex Zr-containing zeolite catalysts reduced the viscosity of the oil in a 42–62% (depending on the Zr-silicalite ratio), which is higher than the 36% reduction determined for the aquathermolysis treatment in absence of the catalyst. The oil content of asphaltenes and resins also was significantly reduced with this Zr-zeolite, which in addition increased the content of saturated hydrocarbons with respect to the original oil and that with the thermal treatment (Su et al. 2017). 6.4.2.3.5

Other Complex Inorganic Supports

A kaolin-modified mesoporous–macroporous support has been used to prepare NiO nanoparticles, thus rendering a novel catalyst (Hassan et al. 2013). To synthesize the catalyst, commercial kaolin was treated with aqueous solutions of calcium and barium acetates to increase the basicity of kaolin. Then, pores were created by sucrose addition, and particles were extruded and dried. Finally, NiO nanoparticles were added in a known proportion, and the mixture was calcined at 650 ∘ C for eight hours. The resulting catalysts showed a pore size around 200 Å. The incorporation of the nanoparticles was advantageous for the adsorption of asphaltenes and the catalytic oxidation of asphaltene (Hassan et al. 2013). Another study by López et al. (2017) reported the activity of bimetallic Mo–Ni or Mo–Co nanoparticles supported on clay or on metakaolin for a catalytic steam injection process for extra-heavy oil. Mean particle sizes, with a ±0.1 nm error, were 4.5 and 4.4 nm for Mo–Ni and Mo–Co clay-supported nanoparticles, respectively, and 8.9 and 13.5 nm for Mo–Ni and Mo–Co metakaolin-supported nanoparticles, respectively. For both clay-supported and metakaolinsupported bimetallic nanoparticles, the absorption of asphaltenes was significantly higher than that measured for the clay and metakaolin nonfunctionalized supports. In line to this, the Mo–Co nanoparticles showed a higher adsorption than the Mo–Ni nanoparticles in both supports. Additionally, the production of coke in thermal process accompanied by steam hydrogen injection was much lower in the presence of the supported bimetallic nanoparticles (2–4% for metakaolinsupported nanoparticles and 5–6% for clay-supported nanoparticles) than in the presence/the absence of the nonfunctionalized support (13–21%). Using inert atmosphere, the reduction of the coke production in thermal cracking in presence of catalyst was also noteworthy, e.g. coke production was 6.19% for metakaolin-supported Mo–Ni nanoparticles, while 25.80% for thermal cracking in the absence of the supported catalyst. These bimetallic-supported catalysts also significantly reduced the activation energies for the cracking reactions of the virgin oil (López et al. 2017).

6.4 Catalytic Activity

6.4.3

Biogenic and Complex Organic Supports

Hamedi-Shokrlu and Babadagli (2013) have used nickel nanoparticles as catalysts for in situ heavy oil upgrading during steam injection process. They devised a process consisting of previously injecting a cationic surfactant in the medium in order to change the electric charge nature to positive, then followed by nickel nanoparticles stabilized in a xantham–gum polymer, which allows nickel nanoparticles (average size of 100 nm) to spread through the oil/water interface as well as to the matrix surface. They reported that through this methodology, the nickel nanoparticles allowed for the reduction of activation energy of the main aquathermolysis reaction (the reaction corresponding to the cleavage of the C—S bonds in organosulfur compounds, together with the generation of hydrogen sulfide and other lighter components) from 69 to 38 kJ/mol. The maximum catalytic activity of nanoparticles was observed at 270 ∘ C, above which the rates of catalytic and noncatalytic aquathermolysis decrease. This methodology (surfactant + polymer + nickel nanoparticles) increased oil recovery by 22% (with 7% being due to nickel nanoparticles alone), accompanied by a 40% viscosity decrease of the produced oil after catalysis. Finally, moving to biogenic supports, biogenic palladium nanoparticles (2.8 μm size) were obtained through the bacterial reduction of sodium tetrachloropalladate(II) by Desulfovibrio desulfuricans. After killing the bacteria, the bacterial biomass was used as support of the palladium catalyst for in situ heavy oil upgrading with THAI. The effect on the reduction of viscosity and the increase of API gravity achieved by this palladium catalyst is not as good as a classical Co–Mo/Al2 O3 catalyst. However, in exchange, the Pd/biomass catalyst enabled the production of oil enriched in liquid oil fraction and a much less generation of coke content. Carbon-supported palladium nanoparticles rendered less viscous oil with similar API gravity increase and liquid/gas composition than biomass-Pd, but more contaminated with coke. In another work, Pd/Pt bio-nanoparticles supported the biogenic material and with a 5 or 20 wt.% of Pd alone or Pd/Pt, using D. desulfuricans or Bacillus benzoevorans were obtained. Particle sizes were 5.10 and 7.45 nm for 5 and 20 wt.% Pd nanoparticles, respectively. For Pd/Pt nanoparticles, sizes were 21.16 nm for 5 wt.% Pd/Pt nanoparticles, and 28.87 nm for 20 wt.% Pd/Pt nanoparticles. After treatment, the upgraded light oil content of the treated oil with any of the bionanoparticles was 89–90 wt.%, while in the thermal treatment, the treatment in the presence of biomass and the treatment with a commercial catalyst yielded an upgraded light oil content of 79.83, 84.86, and 87.25 wt.%, respectively. Commercial catalyst was an Al2 O3 -supported Ni–Mo catalyst. The upgraded light oil content of the treated oil with bio-Pd nanoparticles was slightly higher than that with bio-Pd/Pt nanoparticles. In terms of viscosity, the upgrading achieved by bio-nanoparticles is smaller than that obtained using a commercial catalyst, but both were significantly bigger than the upgrading achieved with only thermal cracking. In this last case, bimetallic bio-Pd/Pt nanoparticles were more effective than monometallic bio-Pd nanoparticles (Omajali et al. 2017). Brown et al. (2016) made use of a biogenic nanoscale magnetite (BnM; Fe3 O4 , with a particle size of 15–30 nm) for in situ upgrading of heavy oil related to the THAI-CAPRI process. A stunning 97.8% decrease in viscosity was achieved when compared with the original feed oil, and coke formation was also reduced when compared with thermal cracking (6.9 versus 10.2 wt.%, respectively). Furthermore, the addition of palladium to the biogenic magnetite resulted in a further enhancement of the oil recovery process; indeed, a viscosity reduction of 99.4% was achieved using this nanocomplex when compared with the feed oil. Meanwhile, the coke content was decreased by about 3 wt.%, and the API gravity was increased by 7.8∘ when considering feed oil as reference, against the 5.3∘ obtained by thermal cracking alone. All these resulted in an increase of the upgraded light oil content of 90 wt.% for 9.5 wt.% Pd-BnM versus 79 wt.% for thermal cracking.

299

300

6 Nanoparticles for Heavy Oil In Situ Upgrading

6.5

Conclusions

The performance of different nanoparticulated catalysts based on metal oxides for heavy oil upgrading has been discussed. The catalysts were classified systematically as the function of the metals, bi/polymetallic catalysts, and the existence or absence of a support core. In general, the nanoparticulated catalysts show good properties that make them suitable for heavy oil upgrading, and specifically, for in situ upgrading with different thermal treatment processes aiming at increasing the recovery factor of heavy oil and bitumen. Among other effects with respect to the thermal cracking, the nanoparticles of metal oxides can increase the adsorption of asphaltenes, API gravity, and the content of maltenes in the upgraded oil. Nanoparticles can also reduce coke formation and the content of heteroatoms with respect to thermal cracking. Nanoparticles of certain metal oxides such as hematite, Cobalt(II, III) oxide, and nickel(II) oxide, seem to have a better catalytic activity. Nevertheless, it is virtually impossible to determine in a quantitative form which nanoparticles exert the most efficient and profitable oil upgrading, as the total disparity of the experimental systems, heavy oil feed, reaction conditions, etc., reported in the literature makes it impossible for any type of rational comparison. Moreover, the nature of the substrate varies as studies employ oils from different reservoirs with different composition, or even they use specific oil fractions. Another critical factor is the nature of the nanoparticles used, since their sizes, the metal/semimetal oxide(s) used, the number of metals, the particle surface, and the possibility of support of the particles (and the own nature of the support, supported, unsupported) are extremely diverse. As the previous divergences were not enough, the quantification of the catalytic ability and/or the activities of the nanocatalysts is quite variable. For instance, a few parameters evaluated are adsorption, API gravity increase, the maltene content of the upgraded oil, coke production, desulfurization achieved, or the asphaltene content of the product. Consequently, there is an urgent need of establishing different standard substrates and methods that enable a more effective comparison for the catalytic applications of different metal oxide nanoparticulated catalysts in oil upgrading. In addition, it is also recognized that the use of nanoparticulated catalysts for in situ upgrading of heavy oil is a research topic of increasing interest, which apart from enhancing the heavy oil recovery changes the quality of the oil in a permanent manner.

References Abdrabo, A.E. and Husein, M.M. (2012). Method for converting demetallization products into dispersed metal oxide nanoparticles in heavy oil. Energy and Fuels 26: 810–815. Abuhesa, M.B. and Hughes, R. (2009). Comparison of conventional and catalytic in situ combustion processes for oil recovery. Energy and Fuels 23: 186–192. Adschiri, T., Kanazawa, K., and Arai, K. (1992). Rapid and continuous hydrothermal crystallization of metal oxide particles in supercritical water. Journal of the American Ceramic Society 75: 1019–1022. Agrawal, R. and Wei, J. (1984). Hydrodemetalation of nickel and vanadium porphyrins. 1. Intrinsic kinetics. Industrial and Engineering Chemistry Process Design and Development 23: 505–514. Aiken, J.D. III, and Finke, R.G. (1999). A review of modern transition-metal nanoclusters: their synthesis, characterization, and applications in catalysis. Journal of Molecular Catalysis A: Chemical 145: 1–44. Ali, S.M. and Meldau, R.F. (1979). Current steamflood technology. Journal of Petroleum Technology 31: 1332–1342.

References

Alkhaldi, S. and Husein, M.M. (2014). Hydrocracking of heavy oil by means of in situ prepared ultradispersed nickel nanocatalyst. Energy and Fuels 28: 643–649. Al-Marshed, A., Hart, A., Leeke, G. et al. (2015a). Optimization of heavy oil upgrading using dispersed nanoparticulate iron oxide as a catalyst. Energy and Fuels 29: 6306–6316. https://doi.org/10.1021/acs .energyfuels.5b01451. Al-Marshed, A., Hart, A., Leeke, G. et al. (2015b). Effectiveness of different transition metal dispersed catalysts for in situ heavy oil upgrading. Industrial and Engineering Chemistry Research 54: 10645–10655. An, K. and Somorjai, G.A. (2015). Nanocatalysis I: synthesis of metal and bimetallic nanoparticles and porous oxides and their catalytic reaction studies. Catalysis Letters 145: 233–248. Arora, N. and Sharma, N.N. (2014). Arc discharge synthesis of carbon nanotubes: comprehensive review. Diamond and Related Materials 50: 135–150. Askarian, M., Vatani, A., and Edalat, M. (2017). Heavy oil upgrading in a hydrodynamic cavitation system: CFD modelling, effect of the presence of hydrogen donor and metal nanoparticles. Canadian Journal of Chemical Engineering 95: 670–679. Bagci, S. and Kok, M.V. (2001). In-situ combustion laboratory studies of Turkish heavy oil reservoirs. Fuel Processing Technology 74: 65–79. Basuki, J.S., Jacquemin, A., Esser, L. et al. (2014). A block copolymer-stabilized co-precipitation approach to magnetic iron oxide nanoparticles for potential use as MRI contrast agents. Polymer Chemistry 5: 2611–2620. Bera, A. and Babadagli, T. (2017). Effect of native and injected nano-particles on the efficiency of heavy oil recovery by radio frequency electromagnetic heating. Journal of Petroleum Science and Engineering 153: 244–256. Biyouki, A.A., Hosseinpour, N., Bahramian, A., and Vatani, A. (2017). In-situ upgrading of reservoir oils by in-situ preparation of NiO nanoparticles in thermal enhanced oil recovery processes. Colloids and Surfaces A: Physicochemical and Engineering Aspects 520: 289–300. Bönnemann, H. and Richards, R.M. (2001). Nanoscopic metal particles – synthetic methods and potential applications. European Journal of Inorganic Chemistry 2001: 2455–2480. Brown, A.R., Hart, A., Coker, V.S. et al. (2016). Upgrading of heavy oil by dispersed biogenic magnetite catalysts. Fuel 185: 442–448. Butler, R.M., McNab, G.S., and Lo, H.Y. (1981). Theoretical studies on the gravity drainage of heavy oil during in-situ steam heating. Canadian Journal of Chemical Engineering 59: 455–460. Chaudret, B. (2005). Organometallic approach to nanoparticles synthesis and self-organization. Comptes Rendus Physique 6: 117–131. Cheih, C. (1982). State-of-the-art review of fireflood field projects (includes associated papers 10901 and 10918). Journal of Petroleum Technology 34: 19–36. Chen, D.-H. and Wu, S.-H. (2000). Synthesis of nickel nanoparticles in water-in-oil microemulsions. Chemistry of Materials 12: 1354–1360. Chen, M., Li, C., Li, G.-R. et al. (2019). In situ preparation of well-dispersed CuO nanocatalysts in heavy oil for catalytic aquathermolysis. Petroleum Science 16: 439–446. Chianelli, R.R., Rendina, D., Hauptmann, E.G. et al. (2014). Conversion of heavy tar sands with asphaltene chemical structures via catalytic coking using MoS2 catalytic material. Energy and Fuels 28: 1750–1755. Cohen, D.D., Gulson, B.L., Davis, J.M. et al. (2005). Fine-particle Mn and other metals linked to the introduction of MMT into gasoline in Sydney, Australia: results of a natural experiment. Atmospheric Environment 39: 6885–6896. Dejhosseini, M., Aida, T., Watanabe, M. et al. (2013). Catalytic cracking reaction of heavy oil in the presence of cerium oxide nanoparticles in supercritical water. Energy and Fuels 27: 4624–4631.

301

302

6 Nanoparticles for Heavy Oil In Situ Upgrading

Dervishi, E., Li, Z., Xu, Y. et al. (2009). Carbon nanotubes: synthesis, properties, and applications. Particulate Science and Technology 27: 107–125. ´ Z., and Niederberger, M. (2008). Nonaqueous synthesis of metal oxide Djerdj, I., Arˇcon, D., Jagliˇcic, nanoparticles: short review and doped titanium dioxide as case study for the preparation of transition metal-doped oxide nanoparticles. Journal of Solid State Chemistry 181: 1571–1581. Ebbesen, T.W. and Ajayan, P.M. (1992). Large-scale synthesis of carbon nanotubes. Nature 358: 220–222. Ely, T.O., Amiens, C., Chaudret, B. et al. (1999). Synthesis of nickel nanoparticles. Influence of aggregation induced by modification of poly(vinylpyrrolidone) chain length on their magnetic properties. Chemistry of Materials 11: 526–529. Fan, C., Zhang, Q., and Wei, F. (2015). Controllable oxidation for oil recovery: low temperature oxidative decomposition of heavy oil on a MnO2 catalyst. Chinese Journal of Catalysis 36: 153–159. Fernández-Garcia, M. and Rodgriguez, J.A. (2007). Metal Oxide Nanoparticles. Upton, NY: Brookhaven National Lab (BNL). Franco, C.A., Montoya, T., Nassar, N.N. et al. (2013a). Adsorption and subsequent oxidation of colombian asphaltenes onto nickel and/or palladium oxide supported on fumed silica nanoparticles. Energy and Fuels 27: 7336–7347. Franco, C.A., Nassar, N.N., Ruiz, M.A. et al. (2013b). Nanoparticles for inhibition of asphaltenes damage: adsorption study and displacement test on porous media. Energy and Fuels 27: 2899–2907. Franco, C., Patiño, E., Benjumea, P. et al. (2013c). Kinetic and thermodynamic equilibrium of asphaltenes sorption onto nanoparticles of nickel oxide supported on nanoparticulated alumina. Fuel 105: 408–414. Franco, C.A., Montoya, T., Nassar, N.N., and Cortés, F.B. (2014). NiO and PdO supported on fumed silica nanoparticles for adsorption and catalytic steam gasification of Colombian C7 asphaltenes. In: Handbook on Oil Production Research, 101–145. Hauppauge, NY: Nova Science Publishers. Galarraga, C.E. and Pereira-Almao, P. (2010). Hydrocracking of Athabasca bitumen using submicronic multimetallic catalysts at near in-reservoir conditions. Energy and Fuels 24: 2383–2389. Galukhin, A.V., Erokhin, A.A., Osin, Y.N., and Nurgaliev, D.K. (2015). Catalytic aquathermolysis of heavy oil with iron tris(acetylacetonate): changes of heavy oil composition and in situ formation of magnetic nanoparticles. Energy and Fuels 29: 4768–4773. Golmohammadi, M., Ahmadi, S.J., and Towfighi, J. (2016). Catalytic cracking of heavy petroleum residue in supercritical water: study on the effect of different metal oxide nanoparticles. Journal of Supercritical Fluids 113: 136–143. Greff, J.H. and Babadagli, T. (2011). Catalytic effects of nano-size metal ions in breaking asphaltene molecules during thermal recovery of heavy-oil. In: SPE Annual Technical Conference and Exhibition. OnePetro. SPE-146604-MS. Guo, T., Nikolaev, P., Rinzler, A.G. et al. (1995). Self-assembly of tubular fullerenes. The Journal of Physical Chemistry 99: 10694–10697. Guo, K., Li, H., and Yu, Z. (2015). Metallic nanoparticles for enhanced heavy oil recovery: promises and challenges. Energy Procedia 75: 2068–2073. Guo, K., Li, H., and Yu, Z. (2016). In-situ heavy and extra-heavy oil recovery: a review. Fuel 185: 886–902. Guo, K., Zhang, Y., Shi, Q., and Yu, Z. (2017). The effect of carbon-supported nickel nanoparticles in the reduction of carboxylic acids for in situ upgrading of heavy crude oil. Energy and Fuels 31: 6045–6055.

References

Hamedi Shokrlu, Y. and Babadagli, T. (2013). In-situ upgrading of heavy oil/bitumen during steam injection by use of metal nanoparticles: a study on in-situ catalysis and catalyst transportation. SPE Reservoir Evaluation and Engineering 16: 333–344. https://doi.org/10.2118/146661-PA. Hart, A., Greaves, M., and Wood, J. (2015). A comparative study of fixed-bed and dispersed catalytic upgrading of heavy crude oil using-CAPRI. Chemical Engineering Journal 282: 213–223. Hart, A., Omajali, J.B., Murray, A.J. et al. (2016). Comparison of the effects of dispersed noble metal (Pd) biomass supported catalysts with typical hydrogenation (Pd/C, Pd/Al2 O3 ) and hydrotreatment catalysts (CoMo/Al2 O3 ) for in-situ heavy oil upgrading with toe-to-heel air injection (THAI). Fuel 180: 367–376. Hashemi, R., Nassar, N.N., and Pereira-Almao, P. (2012). Transport behavior of multimetallic ultradispersed nanoparticles in an oil-sands-packed bed column at a high temperature and pressure. Energy and Fuels 26: 1645–1655. Hashemi, R., Nassar, N.N., and Pereira Almao, P. (2013). Enhanced heavy oil recovery by in situ prepared ultradispersed multimetallic nanoparticles: a study of hot fluid flooding for Athabasca bitumen recovery. Energy and Fuels 27: 2194–2201. Hashemi, R., Nassar, N.N., and Almao, P.P. (2014a). Nanoparticle technology for heavy oil in-situ upgrading and recovery enhancement: opportunities and challenges. Applied Energy 133: 374–387. Hashemi, R., Nassar, N.N., and Pereira Almao, P. (2014b). In situ upgrading of Athabasca bitumen using multimetallic ultradispersed nanocatalysts in an oil sands packed-bed column: Part 1. Produced liquid quality enhancement. Energy and Fuels 28: 1338–1350. Hashemi, R., Nassar, N.N., and Pereira Almao, P. (2014c). In situ upgrading of Athabasca bitumen using multimetallic ultradispersed nanocatalysts in an oil sands packed-bed column: Part 2. Solid analysis and gaseous product distribution. Energy and Fuels 28: 1351–1361. Hashemi-Kiasari, H., Hemmati-Sarapardeh, A., Mighani, S. et al. (2014). Effect of operational parameters on SAGD performance in a dip heterogeneous fractured reservoir. Fuel 122: 82–93. Hassan, A., Lopez-Linares, F., Nassar, N.N. et al. (2013). Development of a support for a NiO catalyst for selective adsorption and post-adsorption catalytic steam gasification of thermally converted asphaltenes. Catalysis Today 207: 112–118. Hosseinpour, N., Khodadadi, A.A., Bahramian, A., and Mortazavi, Y. (2013). Asphaltene adsorption onto acidic/basic metal oxide nanoparticles toward in situ upgrading of reservoir oils by nanotechnology. Langmuir 29: 14135–14146. Hosseinpour, N., Mortazavi, Y., Bahramian, A. et al. (2014). Enhanced pyrolysis and oxidation of asphaltenes adsorbed onto transition metal oxides nanoparticles towards advanced in-situ combustion EOR processes by nanotechnology. Applied Catalysis A: General 477: 159–171. Hosseinpour, M., Ahmadi, S.J., and Fatemi, S. (2015a). Successive co-operation of supercritical water and silica-supported iron oxide nanoparticles in upgrading of heavy petroleum residue: suppression of coke deposition over catalyst. Journal of Supercritical Fluids 100: 70–78. Hosseinpour, M., Fatemi, S., and Ahmadi, S.J. (2015b). Catalytic cracking of petroleum vacuum residue in supercritical water media: Impact of α-Fe2 O3 in the form of free nanoparticles and silica-supported granules. Fuel 159: 538–549. Husein, M.M. and Alkhaldi, S.J. (2014). In situ preparation of alumina nanoparticles in heavy oil and their thermal cracking performance. Energy and Fuels 28: 6563–6569. Iskandar, F., Dwinanto, E., Abdullah, M., and Muraza, O. (2016). Viscosity reduction of heavy oil using nanocatalyst in aquathermolysis reaction. Kona Powder and Particle Journal 33: 3–16. Journet, C., Maser, W.K., Bernier, P. et al. (1997a). Large-scale synthesis of carbon nanotubes. Nature 388: 756–758.

303

304

6 Nanoparticles for Heavy Oil In Situ Upgrading

Journet, C., Maser, W.K., Bernier, P. et al. (1997b). Large-scale production of single-walled carbon nanotubes by the electric-arc technique. Nature 388: 756–758. Kadiev, K.M., Khadzhiev, S.N., and Kadieva, M.K. (2013). Synthesis and use of polyfunctional catalyst nanoparticles for hydroconversion of natural bitumen. Petroleum Chemistry 53: 298–308. Kadiev, K.M., Oknina, N.V., Maksimov, A.L. et al. (2016). Hydroconversion of oil sludge in the presence of ultradisperse slurry catalyst. Research Journal of Pharmaceutical, Biological and Chemical Sciences 7: 704–714. Kandpal, N.D., Sah, N., Loshali, R. et al. (2014). Co-Precipitation Method of Synthesis and Characterization of Iron Oxide Nanoparticles, 87–90. India: NISCAIR-CSIR. Karami, D. and Mahinpey, N. (2016). Application of novel zeolite Y nanoparticles in catalytic cracking reactions. Chemical Engineering Communications 203: 251–257. Kayani, Z.N., Arshad, S., Riaz, S., and Naseem, S. (2014). Synthesis of iron oxide nanoparticles by sol–gel technique and their characterization. IEEE Transactions on Magnetics 50: 1–4. Khalil, M., Yu, J., Liu, N., and Lee, R.L. (2014). Hydrothermal synthesis, characterization, and growth mechanism of hematite nanoparticles. Journal of Nanoparticle Research 16: 1–10. Khalil, M., Lee, R.L., and Liu, N. (2015). Hematite nanoparticles in aquathermolysis: a desulfurization study of thiophene. Fuel 145: 214–220. Khalil, M., Jan, B.M., Tong, C.W., and Berawi, M.A. (2017). Advanced nanomaterials in oil and gas industry: design, application and challenges. Applied Energy 191: 287–310. Kibsgaard, J., Lauritsen, J.V., Lægsgaard, E. et al. (2006). Cluster − support interactions and morphology of MoS2 nanoclusters in a graphite-supported hydrotreating model catalyst. Journal of the American Chemical Society 128: 13950–13958. Kloust, H., Zierold, R., Merkl, J.-P. et al. (2015). Synthesis of iron oxide nanorods using a template mediated approach. Chemistry of Materials 27: 4914–4917. Kondoh, H., Tanaka, K., Nakasaka, Y. et al. (2016). Catalytic cracking of heavy oil over TiO2 –ZrO2 catalysts under superheated steam conditions. Fuel 167: 288–294. Kong, J., Soh, H.T., Cassell, A.M. et al. (1998). Synthesis of individual single-walled carbon nanotubes on patterned silicon wafers. Nature 395: 878–881. Kosari, M., Golmohammadi, M., Ahmadi, S.J. et al. (2017). On the catalysis capability of transition metal oxide nanoparticles in upgrading of heavy petroleum residue by supercritical water. Journal of Supercritical Fluids 126: 14–24. Lan, L., Ge, S., Liu, K. et al. (2011). Synthesis of Ni2 P promoted trimetallic NiMoW/γ-Al2 O3 catalysts for diesel oil hydrotreatment. Journal of Natural Gas Chemistry 20: 117–122. Larbey, R.J. (1994). Issues surrounding the use of lead in gasoline – energy, economic and environmental. Science of the Total Environment 146: 19–26. Lee, D.G. and Noureldin, N.A. (1989). Effect of water on the low-temperature oxidation of heavy oil. Energy and Fuels 3: 713–715. Leliveld, R.G. and Eijsbouts, S.E. (2008). How a 70-year-old catalytic refinery process is still ever dependent on innovation. Catalysis Today 130: 183–189. Li, C., Huang, W., Zhou, C., and Chen, Y. (2019). Advances on the transition-metal based catalysts for aquathermolysis upgrading of heavy crude oil. Fuel 257: 115779. Liu, X., Yang, Z., Li, X. et al. (2015). Preparation of silica-supported nanoFe/Ni alloy and its application in viscosity reduction of heavy oil. Micro & Nano Letters 10: 167–171. López, D., Giraldo, L.J., Salazar, J.P. et al. (2017). Metal oxide nanoparticles supported on macromesoporous aluminosilicates for catalytic steam gasification of heavy oil fractions for on-site upgrading. Catalysts 7: 319. Maity, S.K., Ancheyta, J., and Marroquín, G. (2010). Catalytic aquathermolysis used for viscosity reduction of heavy crude oils: a review. Energy and Fuels 24: 2809–2816.

References

Malik, M.A., Wani, M.Y., and Hashim, M.A. (2012). Microemulsion method: a novel route to synthesize organic and inorganic nanomaterials: 1st nano update. Arabian Journal of Chemistry 5: 397–417. Meyers, M.A., Mishra, A., and Benson, D.J. (2006). Mechanical properties of nanocrystalline materials. Progress in Materials Science 51: 427–556. Mihaly, M., Comanescu, A.F., Rogozea, A.E. et al. (2011). NiO–silica based nanostructured materials obtained by microemulsion assisted sol–gel procedure. Materials Research Bulletin 46 (10): 1746–1753. Montoya, T., Argel, B.L., Nassar, N.N. et al. (2016). Kinetics and mechanisms of the catalytic thermal cracking of asphaltenes adsorbed on supported nanoparticles. Petroleum Science 13: 561–571. Muraza, O. and Galadima, A. (2015). Aquathermolysis of heavy oil: a review and perspective on catalyst development. Fuel 157: 219–231. Murphy, C.J., Sau, T.K., Gole, A.M. et al. (2005). Anisotropic metal nanoparticles: synthesis, assembly, and optical applications. The Journal of Physical Chemistry B 109: 13857–13870. Najafi, M., Abbasi, A., Masteri-Farahani, M., and Janczak, J. (2015). Sonochemical preparation of bimetallic (Cu/Mo) oxide nanoparticles as catalysts for dye degradation under mild conditions. Polyhedron 93: 76–83. Nassar, N.N. and Pereira-Almao, P. (2010). Capturing H2 S (g) by in situ-prepared ultradispersed metal oxide particles in an oilsand-packed bed column. Energy and Fuels 24: 5903–5906. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2011a). Metal oxide nanoparticles for asphaltene adsorption and oxidation. Energy and Fuels 25: 1017–1023. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2011b). Effect of surface acidity and basicity of aluminas on asphaltene adsorption and oxidation. Journal of Colloid and Interface Science 360: 233–238. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2011c). Effect of the particle size on asphaltene adsorption and catalytic oxidation onto alumina particles. Energy and Fuels 25: 3961–3965. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2011d). Comparative oxidation of adsorbed asphaltenes onto transition metal oxide nanoparticles. Colloids and Surfaces A: Physicochemical and Engineering Aspects 384: 145–149. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2012). Thermogravimetric studies on catalytic effect of metal oxide nanoparticles on asphaltene pyrolysis under inert conditions. Journal of Thermal Analysis and Calorimetry 110: 1327–1332. Nassar, N.N., Hassan, A., Luna, G., and Pereira-Almao, P. (2013a). Kinetics of the catalytic thermooxidation of asphaltenes at isothermal conditions on different metal oxide nanoparticle surfaces. Catalysis Today 207: 127–132. Nassar, N.N., Hassan, A., and Pereira-Almao, P. (2013b). Clarifying the catalytic role of NiO nanoparticles in the oxidation of asphaltenes. Applied Catalysis A: General 462: 116–120. Nassar, N.N., Hassan, A., and Vitale, G. (2014). Comparing kinetics and mechanism of adsorption and thermo-oxidative decomposition of Athabasca asphaltenes onto TiO2 , ZrO2 , and CeO2 nanoparticles. Applied Catalysis A: General 484: 161–171. Niederberger, M. (2007). Nonaqueous sol–gel routes to metal oxide nanoparticles. Accounts of Chemical Research 40: 793–800. Niquille-Röthlisberger, A. and Prins, R. (2006). Hydrodesulfurization of 4,6-dimethyldibenzothiophene and dibenzothiophene over alumina-supported Pt, Pd, and Pt-Pd catalysts. Journal of Catalysis 242: 207–216. Noorlaily, P., Nugraha, M.I., Khairurrijal, K. et al. (2013). Ethylene glycol route synthesis of nickel oxide nanoparticles as a catalyst in aquathermolysis. Materials Science Forum 737: 93–97.

305

306

6 Nanoparticles for Heavy Oil In Situ Upgrading

Olvera, J.N.R., Gutiérrez, G.J., Serrano, J.A.R. et al. (2014). Use of unsupported, mechanically alloyed NiWMoC nanocatalyst to reduce the viscosity of aquathermolysis reaction of heavy oil. Catalysis Communications 43: 131–135. Omajali, J.B., Hart, A., Walker, M. et al. (2017). In-situ catalytic upgrading of heavy oil using dispersed bionanoparticles supported on gram-positive and gram-negative bacteria. Applied Catalysis B: Environmental 203: 807–819. Ovalles, C., Filgueiras, E., Morales, A. et al. (2003). Use of a dispersed iron catalyst for upgrading extra-heavy crude oil using methane as source of hydrogen☆. Fuel 82: 887–892. Pachon, L.D. and Rothenberg, G. (2008). Transition-metal nanoparticles: synthesis, stability and the leaching issue. Applied Organometallic Chemistry 22: 288–299. Prasek, J., Drbohlavova, J., Chomoucka, J. et al. (2011). Methods for carbon nanotubes synthesis. Journal of Materials Chemistry 21: 15872–15884. Pu, W.-F., Yuan, C.-D., Jin, F.-Y. et al. (2015). Nanoparticle technology for heavy oil in-situ upgrading and recovery enhancement: opportunities and challenges. Energy and Fuels 29: 1151–1159. Rao, C.V. and Viswanathan, B. (2012). Microemulsion synthesis and electrocatalytic properties of carbon-supported Pd–Co–Au alloy nanoparticles. Journal of Colloid and Interface Science 367: 337–341. Reetz, M.T. and Helbig, W. (1994). Size-selective synthesis of nanostructured transition metal clusters. Journal of the American Chemical Society 116: 7401–7402. Reetz, M.T., Helbig, W., and Quaiser, S.A. (1995). Electrochemical preparation of nanostructural bimetallic clusters. Chemistry of Materials 7: 2227–2228. Remón, J., Arauzo, J., García, L. et al. (2016). Bio-oil upgrading in supercritical water using Ni-Co catalysts supported on carbon nanofibres. Fuel Processing Technology 154: 178–187. Ren, Z.F., Huang, Z.P., Xu, J.W. et al. (1998). Synthesis of large arrays of well-aligned carbon nanotubes on glass. Science 80 282: 1105–1107. Roucoux, A., Schulz, J., and Patin, H. (2002). Reduced transition metal colloids: a novel family of reusable catalysts? Chemical Reviews 102: 3757–3778. Saito, R., Fujita, M., Dresselhaus, G., and Dresselhaus, M.S. (1992). Electronic structure of graphene tubules based on C 60. Physical Review B 46 (3): 1804. Sanchez-Dominguez, M., Boutonnet, M., and Solans, C. (2009). A novel approach to metal and metal oxide nanoparticle synthesis: the oil-in-water microemulsion reaction method. Journal of Nanoparticle Research 11: 1823–1829. Saravanan, P., Gopalan, R., and Chandrasekaran, V. (2008). Synthesis and characterisation of nanomaterials. Defence Science Journal 58 (4): 504–516. Sepeur, S. (2008). Nanotechnology: Technical Basics and Applications. Vincentz Network GmbH & Co KG. Shi, Y., Chen, J., Chen, J. et al. (2012). Preparation and evaluation of hydrotreating catalysts based on activated carbon derived from oil sand petroleum coke. Applied Catalysis A: General 441: 99–107. Shokrlu, Y.H., Maham, Y., Tan, X. et al. (2013). Enhancement of the efficiency of in situ combustion technique for heavy-oil recovery by application of nickel ions. Fuel 105: 397–407. Simão, A., Domínguez-Álvarez, E., Yuan, C. et al. (2022). On the use of metallic nanoparticulated catalysts for in-situ oil upgrading. Fuel 313: 122677. Sitnov, S.A., Mukhamatdinov, I.I., Vakhin, A.V. et al. (2018). Composition of aquathermolysis catalysts forming in situ from oil-soluble catalyst precursor mixtures. Journal of Petroleum Science and Engineering 169: 44–50. https://doi.org/10.1016/j.petrol.2018.05.050. Sitnov, S., Mukhamatdinov, I., Tubman, Y. et al. (2019). Investigating the effect of iron oxide(II, III) nanoparticles on aquathermolysis of heavy oil. In: IOP Conference Series: Earth and Environmental Science. IOP Publishing, 12006.

References

Sitnov, S., Mukhamatdinov, I., Aliev, F. et al. (2020). Heavy oil aquathermolysis in the presence of rock-forming minerals and iron oxide(II, III) nanoparticles. Petroleum Science and Technology 38: 574–579. Sitnov, S.A., Khelkhal, M.A., Mukhamatdinov, I.I. et al. (2022). Iron oxide nanoparticles impact on improving reservoir rock minerals catalytic effect on heavy oil aquathermolysis. Fuel 327: 124956. Song, L.X., Lu, Z.Y., and Liao, Q.L. (2005). Study on preparation of magnetic Fe3 O4 nano-particle with double micro emulsion. Journal of Functional Materials 36: 1762–1764. Soultanidis, N., Zhou, W., Kiely, C.J., and Wong, M.S. (2012). Solvothermal synthesis of ultrasmall tungsten oxide nanoparticles. Langmuir 28: 17771–17777. Su, L., Guan, Z., Li, Q. et al. (2017). Synthesis of SO4 2− /Zr-silicalite-1 zeolite catalysts for upgrading and visbreaking of heavy oil. Journal of Nanoparticle Research 19: 1–10. Suwaid, M.A., Varfolomeev, M.A., Al-Muntaser, A.A. et al. (2022). Using the oil-soluble copper-based catalysts with different organic ligands for in-situ catalytic upgrading of heavy oil. Fuel 312: 122914. Szabó, A., Perri, C., Csató, A. et al. (2010). Synthesis methods of carbon nanotubes and related materials. Materials (Basel) 3: 3092–3140. Takahashi, T., Higashi, H., and Kai, T. (2005). Development of a new hydrodemetallization catalyst for deep desulfurization of atmospheric residue and the effect of reaction temperature on catalyst deactivation. Catalysis Today 104: 76–85. Tarboush, B.J.A. and Husein, M.M. (2012). Adsorption of asphaltenes from heavy oil onto in situ prepared NiO nanoparticles. Journal of Colloid and Interface Science 378: 64–69. Terrones, M., Grobert, N., Olivares, J. et al. (1997). Controlled production of aligned-nanotube bundles. Nature 388: 52–55. Thess, A., Lee, R., Nikolaev, P. et al. (1996). Crystalline ropes of metallic carbon nanotubes. Science 80 273: 483–487. Trovarelli, A. (1996). Catalytic properties of ceria and CeO2 -containing materials. Catalysis Reviews 38: 439–520. Vakhin, A.V., Aliev, F.A., Mukhamatdinov, I.I. et al. (2021). Extra-heavy oil aquathermolysis using nickel-based catalyst: some aspects of in-situ transformation of catalyst precursor. Catalysts 11: 189. Valkovic, V. (1978). Trace Elements in Petroleum. PPC Books. Varzaneh, A.Z., Towfighi, J., Kootenaei, A.H.S., and Mohamadalizadeh, A. (2015). Effect of cerium and zirconium nanoparticles on the structure and catalytic performance of SAPO-34 in steam cracking of naphtha to light olefins. Reaction Kinetics, Mechanisms and Catalysis 115: 719–740. Wei, L.I., Zhu, J., and Qi, J. (2007). Application of nano-nickel catalyst in the viscosity reduction of Liaohe extra-heavy oil by aqua-thermolysis. Journal of Fuel Chemistry and Technology 35: 176–180. Wen, S., Zhao, Y., Liu, Y., and Hu, S. (2007). A study on catalytic aquathermolysis of heavy crude oil during steam stimulation. In: International Symposium on Oilfield Chemistry. OnePetro. SPE-106180-MS. Xiao, A., Zhou, S., Zuo, C. et al. (2015). Hydrothermal synthesis of mesoporous metal oxide arrays with enhanced properties for electrochemical energy storage. Materials Research Bulletin 61: 54–57. Yagi, S., Ichikawa, Y., Yamada, I. et al. (2013). Synthesis of binary magnesium–transition metal oxides via inverse coprecipitation. Japanese Journal of Applied Physics 52: 25501. Yao, H.C. and Yao, Y.F.Y. (1984). Ceria in automotive exhaust catalysts: I Oxygen storage. Journal of Catalysis 86: 254–265. Yeletsky, P.M., Reina, T.R., Bulavchenko, O.A. et al. (2019). Phenanthrene catalytic cracking in supercritical water: effect of the reaction medium on NiMo/SiO2 catalysts. Catalysis Today 329: 197–205.

307

308

6 Nanoparticles for Heavy Oil In Situ Upgrading

Yuan, C., Varfolomeev, M.A., Emelianov, D.A. et al. (2018). Oxidation behavior of light crude oil and its SARA fractions characterized by TG and DSC techniques: differences and connections. Energy and Fuels 32: 801–808. Zamani, A., Maini, B., and Almao, P.P. (2013). Propagation of Nanocatalyst Particles Through Athabasca Sands. J Can Pet Technol 52 (04): 279–288. Paper Number: SPE-148855-PA. https://doi.org/10.2118/148855-PA. Zamani, A., Maini, B., and Pereira-Almao, P. (2012). Flow of nanodispersed catalyst particles through porous media: effect of permeability and temperature. Canadian Journal of Chemical Engineering 90: 304–314. Zhang, X. and Klabunde, K.J. (1992). Superoxide (O2 − ) on the surface of heat-treated ceria. Intermediates in the reversible oxygen to oxide transformation. Inorganic Chemistry 31: 1706–1709. Zhang, S., Liu, D., Deng, W., and Que, G. (2007). A review of slurry-phase hydrocracking heavy oil technology. Energy and Fuels 21: 3057–3062.

309

7 Catalytic Mechanism and Kinetics Guillermo Félix 1 , Alexis Tirado 1 , Ameen Al-Muntaser 1 , Mikhail A. Varfolomeev 1 , Chengdong Yuan 1,3 , and Jorge Ancheyta 1,2 1

Department of Petroleum Engineering, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Mexicano del Petróleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, Mexico City, 07730, Mexico 3 Center for Petroleum Science and Engineering, Skolkovo Institute of Science and Technology, Moscow, Russia 2 Instituto

7.1

Introduction

The decrease in light hydrocarbons reserves, which are mainly composed by products of high commercial value and quality (gasoline, diesel, and jet fuel), and the increase in the energy demand worldwide have driven increased research on heavy oil, extra-heavy oil, and bitumen recovery technologies. In addition, heavy crude oils production, transportation, and refining present great challenges due to their inherent properties, such as high viscosity, low API gravity and hydrogen/carbon (H/C) ratio, high content of heteroatoms, metals, and asphaltenes (Al-Marshed et al. 2015; Ancheyta and Speight 2007; Castañeda et al. 2014; Elahi et al. 2019; Hinkle and Batzle 2006; Hongfu et al. 2002; Orozco Castillo 2016; Puron et al. 2014; Scott et al. 2019). Therefore, it is necessary to improve the mobility of these heavy fractions. One popular method is by heating or partially upgrading the heavy oil in the reservoirs to be extracted for further transportation and conversion in refineries into high-value products. There are two fundamental routes for the upgrading of heavy and extra-heavy crude oils: carbon rejection and hydrogen addition. Carbon rejection process is focused on increasing the ratio of H/C through the removal of carbon atoms. However, a disadvantage of these processes is that the amount of generated products compared with the feed is lower. On the other hand, hydrogen addition technologies increase the H/C ratio by the hydrogenation reactions of high molecular weight fractions present in heavy and extra-heavy crude oils using catalysts to increase conversion toward light fractions in a hydrogen-rich atmosphere. The most novel and promising hydrogen addition technology is slurry phase hydrocracking, where the heavy fractions are upgraded by employing a highly dispersed catalyst since it can handle feeds with high amounts of asphaltenes, metals, and heteroatoms like heavy and extra-heavy crude oils (Bellussi et al. 2013; Liu et al. 2009; Machefer et al. 2013; Matsumura et al. 2005; Nguyen et al. 2016; Quitian et al. 2015). The thermal enhanced oil recovery (EOR) methods are widely employed for producing hydrocarbons from the heavy oil and bitumen reservoirs through techniques such as steam-assisted gravity drainage (SAGD), cyclic steam stimulation (CSS), and steam flooding. These techniques are attractive because of the significant reduction in viscosity due to heat transferred by steam.

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

310

7 Catalytic Mechanism and Kinetics

Nevertheless, this reduction is just temporary since by losing the amount of heat transferred, the heavy oil decreases its mobility and may even have suffered from polymerization reactions, causing an increase in the original viscosity. Experimental studies have reported the generation of H2 S, CO2 , CO, H2 , and light hydrocarbon gases during the steam injection processes due to the chemical interaction of steam with high molecular-weight compounds such as resins and asphaltenes under high-temperature and high-pressure conditions (Al-Muntaser et al. 2022; Barroux et al. 2013; Hyne 1986; Hyne et al. 1982b; Kapadia et al. 2015; Lin et al. 2020; Muraza and Galadima 2015; Qu et al. 2021; Speight 2013). These chemical reactions, called aquathermolysis, are favored through the interaction of some minerals that present catalytic effects, which would avoid the increase in viscosity and even improve the yield of light hydrocarbons and gases. In this same sense, some technologies capable of upgrading the viscosity and properties of heavy and extra-heavy crude oils in reservoirs using dispersed or nanocatalysts are also promising options for conventional thermal recovery methods, offering a more cost-efficient process with high recovery factors, decreasing the production of undesired products like greenhouse gases emissions, and minimizing the need for diluent used, such as light hydrocarbons (Da Silva De Andrade 2014; Elahi et al. 2019; Félix et al. 2017; Nguyen et al. 2017; Pham et al. 2020; Suwaid et al. 2020; Zhang et al. 2007). The in situ upgrading of heavy and extra-heavy oils is a process used inside the well, whereby lighter products are generated by introducing an ultrafine dispersed or liquid catalyst, a hot fluid to increase the temperature inside the well, e.g. heavy oil, vacuum residue, or bitumen, and hydrogen or a hydrogen donor to favor the hydrogenation reactions and produce more valuable products. These components are mixed and injected into the reservoir to accomplish the partial upgrading (Bueno Zapata et al. 2019; Duran Armas 2021; Elahi et al. 2020). There are different technologies for in situ upgrading (such as steam injection, vapor extraction, in situ combustion, or slurry-phase hydrocracking) using different types of catalysts, or hydrogen donors, as well as diverse operating conditions, so that the reaction mechanisms in each of them are different. The use of ultradispersed nanocatalysts in hydrocracking with heavy and extra-heavy crude oils favors the long-chain hydrocarbons cracking reaction in the presence of hydrogen atmosphere to form lighter fractions with smaller molecular weight, which causes viscosity reduction, decrease of heteroatoms content, and thus improving mobility of heavy and extra-heavy crude oils (Bueno Zapata et al. 2019; Elahi et al. 2020, 2019; Rodriguez-DeVecchis et al. 2017). Various kinetic studies have been published for the in situ and ex situ upgrading of heavy crude oil using ultradispersed nanocatalysts in a hydrogen (hydrocracking) atmosphere. Almost all these kinetic studies use the lumping approach to model the reactions since lumping models are the most applied for heavy feedstocks due to their complex composition. The lumps are based on distillation boiling points or fractions solubility (SARA), and the number of kinetic parameters depends on the reaction schemes, which consider up to 13 reaction rate coefficients. In the case of in situ upgrading experiments, the data to develop the kinetic studies are obtained at laboratory bench scale reactors, attempting to simulate the reservoir conditions. The literature on kinetic modeling for aquathermolysis reaction with or without a catalyst is scarce. The reported kinetic models provide insights on the performance of the reaction that can be further used to understand reaction mechanisms as well as to predict the behavior of yields and selectivities of different reaction products and reactant species. Different techniques have been used for the development of kinetic models focusing on the elemental composition, generation of certain gases, and/or the change in the fractional composition of hydrocarbon reserves (Kapadia et al. 2015; Tirado et al. 2022a). Therefore, the description of these models differs from other more developed processes. For the case of heavy and extra-heavy crude oils ex situ upgrading, the use of kinetic models for reactor modeling is well understood. Nevertheless, for in situ upgrading of heavy and extra-heavy

7.2 Reaction Mechanism During Heavy Crude Oil Upgrading

crude oils as hydrocracking or aquathermolysis processes, where the mechanisms represent the reaction pathways that take place in a medium of reservoir conditions, it is necessary to achieve a better understanding of the kinetic models for adequate control, and design of field-scale processes. Therefore, this chapter focuses on the analysis of the different kinetic models reported in the literature for the in situ and ex situ upgrading of heavy and extra-heavy crude oils using dispersed phase catalysts in a hydrogen (hydrocracking) and steam (aquathermolysis) atmosphere.

7.2

Reaction Mechanism During Heavy Crude Oil Upgrading

7.2.1

Reaction Mechanism During Hydrocracking of Heavy Crude Oils

Hydrocracking reactions can be carried out by different mechanisms depending on the catalyst type: In the absence of a catalyst, hydropyrolysis reactions (thermal hydrocracking) can be achieved at high pressures and temperatures. When there is a noble metal as catalyst, the cleavage of carbon–carbon bonds followed by hydrogenation of the fragments (hydrogenolysis) also occurs in these reactions. Catalytic cracking occurs when you have acid catalysts via carbenium and carbonium ions. Finally, when bifunctional catalysts (noble metal and a Brønsted acid component) are present, there is a synergy of hydrogenation–dehydrogenation reactions (Scherzer and Gruia, 1996; Weitkamp 2012). The results of Du et al. (2015) indicate that the hydrocracking mechanism with suspension phase catalysts is by free radicals and not by carbon ions, since the composition of the thermal hydrocracking and catalytic hydrocracking product of heavy oil and model reactant (5 wt% n-butylbenzene dissolved in n-pentane) is the same. The presence of Mo catalyst can overcome the undesirable products (gas, residue, and coke), since the dispersed catalyst split hydrogen increasing the concentration of free radicals, which promote the cracking of aromatic and naphthenic hydrocarbons, nonetheless, there is no generation of isomerization products. During in situ and ex situ upgrading, the catalytic hydrocracking process can be divided into two main stages: hydrocracking and hydrogenation. The main route to exposed free carbon radicals generated during cracking reactions is by the cleavage of C—C bonds and C-heteroatoms (S, N, O), or metals, provoked by the coordination effect of transition metal atoms within the catalyst. These radicals are then absorbed by free hydrogen atoms generated by catalyst active sites. Accordingly, the hydrodemetallization, hydrodesulfurization, hydrodeoxygenation, and hydrodenitrogenation reactions play an important role in enhancing the properties and quality of these heavy hydrocarbons by reducing the content of impurities (Angeles et al. 2014; Balasubramanian and Pushpavanam 2008; Fukuyama and Terai 2007; Gray et al. 1992; Li et al. 2014; Miki et al. 1983; Nguyen et al. 2016; Zhang et al. 2007). Commercial supported catalysts are commonly applied for ex situ applications since they have shown high conversion during heavy crude oil upgrading owing to the bifunctional activity to crack high molecular weight fractions and produce protons during the reaction. Nevertheless, these catalysts have a decay in the activity triggered by deposition of impurities or high carbon content fractions, such as coke and metal, and blocking active catalysts sites. Therefore, the recommendation for heavy and extra-heavy crude oils during the in situ upgrading is to use for ultradispersed solid or liquid catalysts since these catalysts can be in situ sulfided, have higher surface area due to the particle size, neglect activity losses, enhance diffusion through the oil, and can be synthesized in situ lowering the cost. Nonetheless, these catalysts have the main objective of improving hydrogenation reactions, leading to different reaction pathways (Al-Attas et al. 2019; Al-Marshed et al. 2015; Ancheyta et al. 2018; Browning et al. 2016; del Bianco et al. 1994; Hashemi et al. 2014).

311

7 Catalytic Mechanism and Kinetics

Additional mechanisms to the chemical reaction, such as mass transfer, can be commonly found in multiphase reactors, and these physical processes (mass transfer of molecules between phases) and chemical reactions occur simultaneously inside the reactor. The mechanism that controls the entire process is the one that takes the longest to complete, so the importance of the mechanism is defined by the rate at which it occurs. In order to develop kinetic models representing the reaction mechanism taking place in heavy and extra-heavy crude oil upgrading process, specially designed tests are necessary due to transport and diffusion phenomena that can cover the intrinsic kinetic and the behavior of the reagent and products that cannot be predicted since under these conditions the reaction rate is not controlled by the chemical reactions but by these phenomena. In the particular case of the use of dispersed catalysts in heavy oil hydrocracking reactions, the mass transfer mechanism between the gas–liquid–solid phases must also be taken into account, in addition to the chemical reaction. Hence, it is necessary to find the appropriate conditions to obtain experimental data and develop intrinsic kinetic models where these phenomena are neglected, and the kinetic parameters obtained (reaction rate coefficients, activation energies, and collision factors) represent precisely the reactions mechanisms performed. Based on this, the kinetic models can be accurately used in reactor modeling (ex situ upgrading) or reservoir simulation (in situ upgrading) (Ancheyta 2020; Angeles et al. 2014; Perego and Peratello 1999). The change in the molecular composition of heavy and extra-heavy oils during in situ upgrading using dispersed catalyst in a hydrogen atmosphere is carried out by three main reaction mechanisms. Figure 7.1 shows a representation of the global reaction pathways that take place on heavy fractions. Thermal cracking reactions are favored in the temperature range at which the process

H3C HO

CH3

g

in

2

+7H2

ki ac cr al

ck

ra

+H

m

S

c ro

er

yd

Hydrogenation

S

H

ng

CH3

Th

312

S

H3C H3C OH

H3C

H2C

CH3 H3C CH2

H3C

OH

H 3C

H3C CH3

S

HO

H3C

S S

+H

2

+ 2H S + 2H O 2 2

Figure 7.1 Global reaction pathways of thermal cracking, hydrocracking, and hydrolysis of complex molecule present in heavy oils, vacuum residues, and bitumens.

7.2 Reaction Mechanism During Heavy Crude Oil Upgrading

is carried out and the reaction rate increases exponentially with the temperature. In this way, complex molecules such as asphaltenes and resins (present mainly in residue and vacuum gas oil fractions) are cracked, breaking C—S, C—C bonds through a free radical mechanism producing lower molecular weight fractions. However, this thermal process leads to olefin and coke precursor formation, which are nondesired products during the upgrading of heavy oil. When catalysts and hydrogen are added, hydrogenation reactions prevent the formation of these precursors by saturating the radicals with active hydrogen. Moreover, some of these reactions focused on the removal of heteroatoms such as sulfur, nitrogen, oxygen, and metals. Catalysts also promote C—C bond insertion through the carbonium ions mechanism and subsequent hydrogenation of those bonds resulting in saturated compounds with lower boiling temperatures (Al-Attas et al. 2019; Ancheyta 2013; del Bianco et al. 1994; Nguyen et al. 2016).

7.2.2

Reaction Mechanism During Aquathermolysis Process of Heavy Crude Oils

The development of kinetic models requires appropriate knowledge of the reaction mechanisms involved. Several authors have studied the diverse reaction pathways carried out under aquathermolysis conditions (Aliev et al. 2021; Hyne 1986; Katritzky et al. 1996; Tumanyan et al. 2015). This set of chemical reactions is typically predominant at a higher temperature of 200 ∘ C since, at lower temperatures, chemical reactions are characterized by slow thermal maturation of the contents in the reservoir. While at higher temperatures of 300 ∘ C, thermal cracking reactions are favored (either in the presence or absence of water). These reactions involve the cleavage of C-heteroatom bonds (S, N, O), where experimental studies demonstrate that C—S bonds exhibit the highest sensitivity due to the larger electronegativity of the sulfur, being more likely to break down. Studies with organosulfur model compounds such as thiophene and tetrahydrothiophene have indicated the yield of different chemical reactions under aquathermolysis conditions (Clark et al. 1983; Lin et al. 2020). The results showed that the gases generation was similar to those obtained during aquathermolysis reactions of heavy oils (H2 S, CO, CO2 , CH4 ), demonstrating that these compound types are the main reactive species during steam injection processes (Clark et al. 1983). Lin et al. (2020) carried out a series of experiments to investigate the aquathermolysis reaction mechanisms of tetrahydrothiophene. Figure 7.2 shows the scheme of the different reaction pathways contemplated based on analysis of reaction products. The generation of compounds such S

+H+

H S

(1)

OH SH

(2)

(2)

H2S + HC

OH + H2S

(1)

OH HO

H2S

(3) H

SH O

O

+H+ (4)

+

H

H O

O

Polymerization

SH (2)

CO

HC

H (5)

(6)

+H+ H2S + HC

CO + H2O ↔ CO2 + H2

Figure 7.2 Reaction pathways for tetrahydrothiophene aquathermolysis: (1) Hydrolysis, (2) Pyrolysis, (3) Isomerization, (4) Rearrangement, (5) Hydrodesulfurization, (6) Water–Gas shift reaction, HC: hydrocarbons. Source: Lin et al. (2020)/American Chemical Society.

313

314

7 Catalytic Mechanism and Kinetics

R

R C

C

Asphaltene

OH

S

C

+

C

HS

Ö H C

S

C

+

C

H SH

NixSy

R

C

H

R

CH2 + CO

O CO + H2O C

CO2 + H2

H + H2S NiSH H2O

Figure 7.3 Reaction pathways for aquathermolysis in the presence of nickel sulfide catalyst. Source: Aliev et al. (2021)/MDPI/CCBY4.0.

as mercapto-alcohols and mercapto-aldehydes indicates ring breakage by protonation of alpha carbon. It is considered the trigger to carry out a series of chemical reactions such as isomerization, decarbonylation, condensation, and hydrodesulfurization even water-gas shift reactions (WGSRs). The reaction mechanisms of heavy oils in contact with steam are not fully understood due to the diverse chemical compounds present in petroleum, the high structural complexity of some molecules, and the combinations of series and parallel reactions involved. Through the characterization techniques, it has been observed that during the aquathermolysis experiments, the average molecular weight and aromaticity of resins and asphaltenes are modified mainly due to the breaking of sulfur bridges, alkyl aromatic, and naphthenic-aromatic compounds. Meanwhile, the water provides hydrogen to saturate free radicals produced, increasing the content of aromatics and saturated compounds along with a small amount of gas (Al-Muntaser et al. 2021; Vakhin et al. 2020). Hyne (1986) reported that during aquathermolysis experiments carried out in a H2 S corrosion-resistant autoclave reactor, the production of gas increased rapidly compared with experiments performed using a quartz tube inside the reactor. This behavior was attributed to the catalytic effect of metals in the reactor walls; nonetheless, the concentration of the gas generated when the equilibrium was reached in both experiments was alike. Based on this observation, different authors have investigated the catalytic effect of minerals during aquathermolysis reaction (Fan 2003; Li et al. 2019; Maity et al. 2010). Aliev et al. (2021) reviewed the effect of different catalysts on the aquathermolysis reaction mechanism. The reaction scheme presented in Figure 7.3 is proposed, which indicates the cracking of complex molecules such as asphaltenes and resins by breaking the C—S bonds. The effect of different catalyst types was analyzed, showing that some chemical reactions are favored. Then, the global aquathermolysis process can be described as follows: RCH2 CH2 SCH3 + 2H2 O → RCH3 + CO2 + H2 + H2 S + CH4

(7.1)

7.3 Description of Reported Kinetic Models for In situ and Ex situ Hydrocracking of Heavy Crude Oil Applications The information reported in the literature on kinetic models for hydrocracking of heavy fractions for in situ and ex situ applications includes the operating conditions, the type of feedstock and catalyst used to obtain the experimental data, the reactions schemes, and the lumps employed in each kinetic model. Furthermore, the methodology followed for each work to solve the mass balance

7.3 Description of Reported Kinetic Models for In situ and Ex situ Hydrocracking of Heavy Crude Oil Applications

equations, the performed optimization algorithm to obtain the kinetic parameters values, along with the accuracy of the model, are described. Currently, there are three main types of approaches to kinetic modeling: discrete lumps, continuous mixture, and single events. However, the most-used approach for the kinetic modeling of heavy and extra-heavy crude oils is discrete lumping since the number of compounds present in these types of oils is huge, and single-event approach is based on individual reaction between molecules, hindering the calculation of the elevated number of reaction rates and parameters. While the continuous mixture approach can handle huge number of compounds, nonetheless, the kinetic parameters are highly dependent on the feed employed (Ancheyta et al. 2005; Balasubramanian and Pushpavanam 2008; Becker et al. 2016; Marafi et al. 2010; Sadighi and Zahedi 2013). All the kinetic models (Álvarez et al. 2019; Asaee et al. 2014; Browning et al. 2019; Cai et al. 2022; Coronel-García et al. 2021; Da Silva De Andrade 2014; Duran Armas 2021; Félix and Ancheyta 2019a, b; Galarraga et al. 2012; Hassanzadeh and Abedi 2010; T. Huang et al. 2017; Kim et al. 2017; Loria et al. 2011; Ma et al. 2022; Nguyen et al. 2013; Orozco Castillo 2016; Ortega García et al. 2017; Pham et al. 2021; Rodriguez-DeVecchis et al. 2017) found in the literature, were developed with experimental data obtained at moderate reaction conditions or reservoir conditions. All the models are based on the different hydrocracking reaction pathways of heavy feedstock using dispersed or liquid catalysts in a hydrogen atmosphere. The kinetic information aims to be applied for slurry phase hydrocracking in refineries (ex situ) as well as in hydrocracking at oil extraction wells (in situ) with the injection of hydrogen or some precursor to produce the hydrogen atmosphere. There are two main ways to lump the heavy and extra-heavy fraction: by distillation boiling points and by solubility. The lumping scheme proposed by Sánchez et al. (2005) is mostly taken as a basis to develop other kinetic models (Figure 7.4a), where the chemical species of heavy oils are grouped in pseudocomponents based on boiling point ranges: unconverted Residue (R, 538+ ∘ C), Vacuum Gas Oil (VGO, 343–538 ∘ C), Distillates (D, 204–343 ∘ C), Naphtha (N, IBP-204 ∘ C), and Gases (G). Although almost all kinetic models found in the literature are based on lumping approach use distillations cuts of heavy oil, few of these works use the lumping approach separating the pseudocomponents into fractions (SARA) based on the solubility (Ashoori et al. 2017; Félix and Ancheyta 2019b). Although the continuous mixture method of modeling reaction kinetics is carried out representing the oil as a continuous mixture rather than a mixture of lumps, contemplating all the compounds inside the petroleum. A continuous mixture is obtained, describing the precise distribution of component concentrations, inclusive of the constituents that are not identified but the difference between their properties is relatively minor. The important objective is to find a distribution of the species that represents the conversion of heavier components to lighters in a continuous way, such as boiling points or distillation curves (Becker et al. 2016; Elizalde et al. 2016, 2009; Elizalde and Ancheyta 2014; Laxminarasimhan et al. 1996). Since the application of these kinetic models is focused on in situ and ex situ upgrading in a hydrogen atmosphere, the reaction mechanisms included in all these reaction pathways only consider the heavy and extra-heavy crude oil components and hydrogen. Meanwhile, for in situ applications (reservoirs), where water is present and can remove sulfur from the oil by aquathermolysis reaction reducing the viscosity in the same way as hydrocracking does (Barroux et al. 2013; Fan et al. 2004; Kapadia et al. 2013; Lin et al. 2020; Tirado et al. 2022b). Consequently, the kinetic models in hydrogen atmosphere consider all these reaction pathways in a global hydrocracking reaction. Reproducing these reaction mechanisms carried out under reserve conditions during in situ upgrading in laboratory-scale reactors is complicated, thus, the manner to develop kinetic models

315

316

7 Catalytic Mechanism and Kinetics R

R

(a)

k1 VGO k2

k5 D

VGO

k1

k6

k4

k9

k8

D

k3

k3

N

(b)

k2

k7

k10

N

k5

k4

k6

G

UCO k1 k6

G

k3

G As

(c)

k1

k4

k11 k5

VGO

k7

C

k11

k7

As k12

k2

k3

k1

k7

Re k13

k9 k10 Ar k8

k6

Sa

G

k4

(d)

k5

k12

k3

D

C

k4 Re

k2

k2

G

(e) k9 k10

k5 Ar

k6

k8 Sa

Figure 7.4 Five-lump kinetic model proposed by (a) Sánchez et al. (2005)/with permission from American Chemical Society, (b) Adapted from Nguyen et al. (2013)/Elsevier, (c) Kim et al. (2017)/Elsevier, (d) Félix and Ancheyta (2019b)/Elsevier (12 reaction rate constants), (e) Félix and Ancheyta (2019b)/Elsevier (9 reaction rate constants).

needs to be strictly analyzed. The experimental settings for the reported kinetic models for ex situ and in situ upgrading of heavy and extra-heavy crude oils are collected in Table 7.1. Batch reactors were used in the earliest published kinetic experiments, although this setup type does not perfectly simulate the characteristics at reservoir pore spaces, it is a valuable tool for determining how the catalyst and operating conditions affect heavy oil upgrading. It also aids in determining the magnitude of the various reactions that are carried out. Following that, tubular and packed-bed reactors are utilized to more accurately replicate reservoir conditions. The kinetic models reported for in situ and ex situ upgrading of heavy and extra-heavy crudes according to the number of lumps used are described below.

7.3.1

Four-lump Kinetic Models

A four-lump kinetic model, shown in Figure 7.5a, (Félix and Ancheyta 2019a), was developed for the hydrocracking of heavy crude oil using Mo-based catalyst, the pseudocomponents were based on the boiling points ranges similar to those reported by Sánchez et al. (2005): R, VGO, D, and N. The reaction pathways were based on the conversion from heavier fractions to lighter fractions. Only the liquid fractions equations were reported, since the G and solid (Coke, C) fractions were proposed to be calculated through another complementary kinetic model (Félix and Ancheyta 2019b) which

Table 7.1

Operating conditions used for the development of kinetic models.

Feedstock

Catalyst

Reactor

Temperature (∘ C)

Residence time (h)

Pressure (MPa)

References

Athabasca bitumen (10.99 ∘ API) Athabasca bitumen (9.5 ∘ API)

NiWMo

Batch

320–380

3–72

3.45

Hassanzadeh and Abedi (2010)

NiWMo

Batch

320–380

3–70

3.45

Galarraga et al. (2012)

Atmospheric residue (16.2 ∘ API)

Molybdenum naphthenate

Batch

420–430

0.5–1

15.2

Nguyen et al. (2013)

Vacuum residue (10.4 ∘ API)

Ammonium phosphomolybdate

Batch

390–435

0.33–4

7

Asaee et al. (2014)

Vacuum residue

MoS2

Batch

380–400

1–16

9.5

Kim et al. (2017)

Iranian heavy oil (3.77 ∘ API)

Oil-soluble dispersed

Batch

405–435

1–10

9

T. Huang et al. (2017)

Heavy crude oil (12.8 ∘ API)

PEG300

Batch

310–370

4–72

7

Coronel-García et al. (2021)

Maya crude oil (19.98 ∘ API) Heavy crude oil (11.97 ∘ API)

Ammonium heptamolybdate

Batch

400

1–24

5.5

Martinez-Grimaldo et al. (2011)

Molibdenite

Batch

360–400

2–5

3.9

Félix and Ancheyta (2019a,b)

Karamay atmospheric residue (15.2 ∘ API)

Mo-containing ionic liquid

Batch

390–430

1–8

12

Cai et al. (2022)

Safaniya vacuum residue (2.2 ∘ API) and Arabian light atmospheric residue (14.6 ∘ API)

Mo-octoate

Semi-batch

420–440

0.5–4

15

Álvarez et al. (2019)

Safaniya vacuum residue (2.2 ∘ API) Athabasca bitumen (11 ∘ API) Athabasca bitumen (9.2 ∘ API) Vacuum Residue (2.4 ∘ API)

Mo-octoate

Semi-batch

—

0.5–6

15

Browning et al. (2019)

NiWMo

Up-flow reactor

320–380

9–51

2.76

Loria et al. (2011)

—

Sandpack with up-flow

340–370

12–48

3.45

Rodriguez-DeVecchis et al. (2017)

NiWMo

Downflow reactor

320–395

20–50

3.45

Da Silva De Andrade (2014) Orozco Castillo (2016)

Mexican heavy oil (9.9 ∘ API)

NiWMo

Plug flow reactor

320–360

24–96

10.34

Vacuum residue (−0.4 ∘ API) Vacuum residue (1.1 ∘ API)

NiWMo

Plug flow reactor

320–360

48–120

10.34

Orozco Castillo (2016)

NiMo

Carbonate rock

320–360

12–72

10

Elahi et al. (2019)

Vacuum residue (1.06 ∘ API)

NiMo

Dolomite corepack

335–365

12–72

10

Duran Armas (2021)

Heavy crude oil (12 ∘ API) Vacuum residue (4.64 ∘ API)

Liquid Ni–Mo based

CSTR

350–370

12–180

9.8

Ortega García et al. (2017)

Mo-octoate

CSTR

410–450

1–4

16

Pham et al. (2021)

318

7 Catalytic Mechanism and Kinetics

R

k1

k2

As k5

k7

VGO

k5

Re

k4 k2

k1

k8

D

k4 Ar

k6 k3

k3

Sa

N

(a)

k6

(b)

Figure 7.5 Four-lump kinetic model reported by (a) Félix and Ancheyta (2019a)/with permission from American Chemical Society, (b) Félix and Ancheyta (2019b)/Elsevier.

is presented later. The reaction rate (r i ) equations follow a first-order reaction: rR = −(k1 + k2 + k3 )𝜔R

(7.2)

rVGO = k1 𝜔R − (k4 + k5 )𝜔VGO

(7.3)

rD = k2 𝜔R + k4 𝜔VGO − k6 𝜔D

(7.4)

rN = k3 𝜔R + k5 𝜔VGO + k6 𝜔D

(7.5)

Complementary reactions rate equations: rG = k4 𝜔As + k7 𝜔Re + k9 𝜔Ar + k10 𝜔Sa

(7.6)

rC = k13 𝜔As

(7.7)

where kj are the reaction rate coefficients for each reaction pathway j, and 𝜔i the mass fractions of component i. Another four-lump kinetic model (Félix and Ancheyta 2019b) for the heavy crude oil hydrocracking using dispersed Mo-based catalyst was lumped based on solubility properties in As, Re, Ar, and Sa. The reaction mechanism (Figure 7.5b) includes the parallel and in series reactions of the SARA fractions, where the high molecular weight compounds produce lower molecular weight fractions. Furthermore, the Ar and Re fractions can agglomerate to produce Re and As fractions, respectively. It is assumed that all reactions are of first-order giving the following mass balance equations:

7.3.2

rAs = −(k1 + k2 + k3 )𝜔As + k7 𝜔Re

(7.8)

rRe = k1 𝜔As + k8 𝜔Ar − (k4 + k5 + k7 )𝜔Re

(7.9)

rAr = k2 𝜔As + k4 𝜔Re − (k6 + k8 )𝜔Ar

(7.10)

rSa = k3 𝜔As + k5 𝜔Re + k6 𝜔Ar

(7.11)

Five-lump Kinetic Models

Hassanzadeh and Abedi (2010) studied kinetics for the in situ upgrading of Athabasca bitumen reactions using a water-soluble Ni–W–Mo catalyst. The reaction scheme, which is observed in Figure 7.4a, only considers the five-lump kinetic model reported by Sánchez et al. (2005), since it is assumed that coke formation is negligible in each experiment. First-order of reaction was

7.3 Description of Reported Kinetic Models for In situ and Ex situ Hydrocracking of Heavy Crude Oil Applications

considered for all reaction rate equations and the mass balance equations are given by the following expressions: rR = −(k1 + k2 + k3 + k4 )𝜔R

(7.12)

rVGO = k1 𝜔R − (k5 + k6 + k7 )𝜔VGO

(7.13)

rD = k2 𝜔R + k5 𝜔VGO − (k8 + k9 )𝜔D

(7.14)

rN = k3 𝜔R + k6 𝜔VGO + k8 𝜔D − k10 𝜔N

(7.15)

rG = k4 𝜔R + k7 𝜔VGO + k9 𝜔D + k10 𝜔N

(7.16)

A kinetic study was carried out by Galarraga et al. (2012) to analyze the improvement and output prediction of Athabasca bitumen obtained via steam-assisted gravity drainage at reservoir conditions using an ultradispersed trimetallic catalyst. The same kinetic model as the before mentioned was used (Figure 7.4a), considering different boiling points ranges in the reaction of residue (545+ ∘ C) toward VGO (343–545 ∘ C), D (216–343 ∘ C), N (IBP-216 ∘ C), and G. Therefore, the mass balance equations are similar to those reported by Hassanzadeh and Abedi (2010) with the difference that the weight fraction is replaced by a dimensionless concentration (the concentration at any time divided by the initial concentration). Furthermore, the general reaction for the residue conversion into light products assuming first-order reaction with only one reaction rate coefficient (k0 ) were performed. Once the evaluation of the ultradispersed catalyst was carried out for the hydrocracking of Athabasca bitumen at bench-scale, kinetic studies on a larger-scale reactor were developed. Loria et al. (2011) use a tubular pilot plant reactor to conduct experiments at moderate reaction severity (9–51 hours of residence times 𝜏, which for continuous reactors was assumed to be the ratio of the reactor volume to the volumetric flow rate, of, and hydrogen-to-oil ratio of 625 standard cm3 /cm3 ) employing ultradispersed catalysts (Ni–W–Mo). The reaction scheme and the lumps were similar to those employed by Galarraga et al. (2012) with different boiling points ranges for VGO (343–550 ∘ C) and R (550+ ∘ C) fractions. It was considered that due to the use of ultradispersed catalyst, the axial dispersion, external, and internal mass diffusion are minimal for continuous reactor gas-oil hydrocracking, resulting in the following set of equations: 𝜔(𝜏f )

𝜏f

d𝜔R = −k1

∫𝜔(𝜏0 ) 𝜔(𝜏f )

∫𝜔(𝜏0 )

∫𝜏0 𝜏f

d𝜔VGO = k1

𝜔(𝜏f )

𝜏f

d𝜔D = k2

∫𝜔(𝜏0 ) 𝜔(𝜏f )

∫𝜔(𝜏0 )

∫𝜏0 𝜏f

d𝜔N = k3

𝜔(𝜏f )

∫𝜔(𝜏0 )

∫𝜏0

∫𝜏0 𝜏f

d𝜔G = k4

∫𝜏0

𝜔R d𝜏 − k2

𝜔R d𝜏 − k5 𝜔R d𝜏 + k5 𝜔R d𝜏 + k6 𝜔R d𝜏 + k7

𝜏f

∫𝜏0 𝜏f

∫𝜏0 𝜏f

∫𝜏0 𝜏f

∫𝜏0 𝜏f

∫𝜏0

𝜔R d𝜏 − k3

𝜏f

∫𝜏0

𝜔VGO d𝜏 − k6 𝜔VGO d𝜏 − k8 𝜔VGO d𝜏 + k8 𝜔VGO d𝜏 + k9

𝜏f

∫𝜏0 𝜏f

∫𝜏0 𝜏f

∫𝜏0 𝜏f

∫𝜏0

𝜔R d𝜏 − k4

𝜏f

𝜔R d𝜏

∫𝜏0

𝜔VGO d𝜏 − k7 𝜔D d𝜏 − k9

𝜏f

∫𝜏0 𝜏f

∫𝜏0

𝜔D d𝜏 − k10 𝜔D d𝜏 + k10

𝜔D d𝜏

𝜏f

∫𝜏0 𝜏f

∫𝜏0

𝜔VGO d𝜏

(7.17) (7.18) (7.19)

𝜔N d𝜏

(7.20)

𝜔N d𝜏

(7.21)

where 𝜏 0 is the initial residence time, 𝜏 f the final residence time, 𝜔(𝜏 0 ) the weight percent at the initial residence time, and 𝜔(𝜏 f ) the weight percent at the final residence time. Additionally, the activity of the catalysts was compared (𝜗i ) in the batch and pilot plant (tubular) reactors at same reaction and residence time, respectively, as follows: 𝜗i =

rPFR,i rBATCH,i

(7.22)

319

320

7 Catalytic Mechanism and Kinetics

where r BATCH is the reaction rate of the batch reactor, and r PFR the reaction rate of the tubular reactor. The activity function for all fractions is near to 1 which implies that the conversion of heavier molecules into lighter molecules for the Athabasca bitumen with ultradispersed catalyst is the same in these two reactors. Da Silva De Andrade (2014) proposed a kinetic model for in situ upgrading of Athabasca bitumen vacuum residue and pitch (asphaltene rich phase) with ultradispersed catalyst. All experiments were conducted in a downflow reactor at hydrogen-to-oil ratio of 90 standard cm3 /cm3 . The reaction mechanism and the mass balance equations reported by Hassanzadeh and Abedi (2010) were used since they used the same five-lump model with equal boiling points. Orozco Castillo (2016) developed the kinetic studies for in situ upgrading process of vacuum residue and heavy oil with operating conditions similar to Mexican carbonate reservoirs in a pilot plant reactor employing a similar catalyst to previous studies based on Ni–W–Mo species. The boiling point ranges for petroleum fractions and the reaction pathways (Figure 7.4a) are the same to those reported by Loria et al. (2011) and Sánchez et al. (2005), respectively. Some studies have reported experiments of sand material throughout thermal reactions for heavy feedstock obtaining more realistic insights for the thermal reactions performance in porous media at conditions near to those used in reservoir operation. Rodriguez-DeVecchis et al. (2017) studied the thermal cracking of Athabasca bitumen in a porous medium with an up-flow tubular reactor at hydrogen-to-oil ratio of 120 standard cm3 /cm3 . The reaction scheme and the mass balance equations are similar to those described in the work of Orozco Castillo (2016), with the only difference that the R conversion was considered of second-order, which produces the following equations: rR = −(k1 + k2 + k3 + k4 )𝜔2R rVGO = rD = rN = rG =

k1 𝜔2R k2 𝜔2R k3 𝜔2R k4 𝜔2R

(7.23)

− (k5 + k6 + k7 )𝜔VGO

(7.24)

+ k5 𝜔VGO − (k8 + k9 )𝜔D

(7.25)

+ k6 𝜔VGO + k8 𝜔D − k10 𝜔N

(7.26)

+ k7 𝜔VGO + k9 𝜔D + k10 𝜔N

(7.27)

Another kinetic study carried out by Elahi et al. (2019) describes the analysis for in situ upgrading inside a carbonate rock (porous media) by injecting vacuum residue from Mexican heavy oil, hydrogen, and a NiMo ultradispersed nanocatalyst throughout a continuous experimental setup. As in the previous works, the operating conditions are alike to the reservoir conditions (hydrogen-to-oil ratio of 150 standard cm3 /cm3 ). The five-lump kinetic model scheme and petroleum fractions are the same as those employed by Hassanzadeh and Abedi (2010). Duran Armas (2021) developed a model for the in situ upgrading of residue fraction from Aguacate heavy oil using Ni–Mo ultradispersed nanocatalyst in a carbonate corepack (four Dolomite cores). All the experimental data were performed at near-reservoir conditions similar to Rodriguez-DeVecchis et al. (2017). The reaction mechanism, the five lumps boiling points range, and the mass balance set of equations are similar to that used by Elahi et al. (2019). The kinetic model for the hydrocracking of the Arabian light atmospheric residue in a batch reactor using molybdenum naphthenate as Mo dispersed catalyst was developed by Nguyen et al. (2013). The experiments were carried out at severe conditions as can be seen in Table 7.1 employing the reaction scheme shown in Figure 7.4b taking into account the following boiling points ranges: R (510+ ∘ C), VGO (350–510 ∘ C), D (180–350 ∘ C), and N (IBP-180 ∘ C). The first order was assumed for all the reactions (for hydrogen and lumps), the reaction from R to N was neglected, and the

7.3 Description of Reported Kinetic Models for In situ and Ex situ Hydrocracking of Heavy Crude Oil Applications j

stoichiometric coefficients (vi of component i for reaction pathway j) of the reactions and molar concentration in the liquid (CiL ) were considered by the following set of equations: L L rR = v1R k1 CRL CH + v2R k2 CRL CH 2

(7.28)

2

L L L L L + v3VGO k3 CVGO CH + v4VGO k4 CVGO CH rVGO = v1VGO k1 CRL CH 2

2

L L L L rD = v2D k2 CRL CH + v3D k3 CVGO CH + v5D k5 CDL CH 2

2

(7.30)

2

L L L L rN = v4N k4 CVGO CH + v5N k5 CDL CH + v6N k6 CNL CH 2

(7.29)

2

2

(7.31)

2

L L L L L L L L rG = v1G k1 CRL CH + v2G k2 CRL CH + v3G k3 CVGO CH + v4G k4 CVGO CH + v5G k5 CDL CH + v6G k6 CNL CH 2

2

2

2

2

2

(7.32) L L L L L L L L rH2 = v1H k1 CRL CH + v2H k2 CRL CH + v3H k3 CVGO CH + v4H k4 CVGO CH + v5H k5 CDL CH + v6H k6 CNL CH 2

2

2

2

2

2

2

2

2

2

2

2

(7.33) Kim et al. (2017) reported a kinetic model for the hydrocracking of a residue fraction using MoS2 as an oil soluble catalyst in a batch reactor. The reaction scheme observed in Figure 7.4c includes five pseudocomponents: Unconverted oil (UCO, 560+ ∘ C), VGO (320–560 ∘ C), D (IBP-320 ∘ C), G, and C. All lumps crack from heavier fractions into a lighter fraction except for C, which is only produced from UCO. Different reaction order values were considered for each lump: 2 for UCO, 1 for VGO, and 0 for D and G, as follows: rUCO = −(k1 + k3 + k6 + k7 )𝜔2UCO rVGO =

k1 𝜔2UCO

− (k2 + k4 )𝜔VGO

rD = k2 𝜔VGO + rG = rC =

k3 𝜔2UCO k6 𝜔2UCO

k7 𝜔2UCO

(7.34) (7.35)

− k5

(7.36)

+ k4 𝜔VGO + k5

(7.37) (7.38)

Another kinetic model was developed for the hydrocracking of Iranian heavy crude oil with an oil-soluble dispersed catalyst in a batch reactor (T. Huang et al. 2017). The reaction mechanism (Figure 7.4a) and the system of differential equations obtained from mass balance are equal to that reported by Hassanzadeh and Abedi (2010). The boiling point ranges of each lump are similar to those reported by Nguyen et al. (2013), only with different R (524+ ∘ C) and VGO (350–524 ∘ C) boiling points. The upgrading of heavy oil was conducted by Ortega-Garcia et al. (2017) in a continuous reactor (CSTR, hydrogen-to-oil ratio of 62.9 standard cm3 /cm3 ) using a liquid–acid catalyst to develop a kinetic model at moderate conditions. The five-lump reaction mechanism (Figure 7.4a) is based on the following boiling points ranges: R (540+ ∘ C), Heavy gasoil (HGO, 343–540 ∘ C), Light gasoil (LGO, 221–343 ∘ C), N (IBP-221 ∘ C), and G. The ordinary differential equations system for the kinetic model is similar to that reported by Hassanzadeh and Abedi (2010) with the difference in the name of fractions of VGO and D, which were named HGO and LGO, respectively. Two more five-lump kinetic models were reported by Félix and Ancheyta (2019b) developed with the same experimental data detailed previously. The first reaction scheme (Figure 7.4d) and the ordinary differential equations system (Eqs. (7.8)–(7.11)) are very similar to the four-lump kinetic model based on SARA fraction; nonetheless, the production of C from As is added in this model, and for the mass balance equations, the term −k9 𝜔As is added to the As rate equation (Eq. (7.8)) and the C rate equation is the following: rC = k9 𝜔As

(7.39)

321

322

7 Catalytic Mechanism and Kinetics

The second kinetic model (Figure 7.4e), moreover considers the reactions between SARA fractions (as in the four-lump model), includes the reaction of all SARA fractions to produce G fraction, in order to obtain the following mass balance equations: rAs = −(k1 + k2 + k3 + k4 )𝜔As + k11 𝜔Re

(7.40)

rRe = k1 𝜔As + k12 𝜔Ar − (k5 + k6 + k7 + k11 )𝜔Re

(7.41)

rAr = k2 𝜔As + k5 𝜔Re − (k8 + k9 + k12 )𝜔Ar

(7.42)

rSa = k3 𝜔As + k6 𝜔Re + k8 𝜔Ar − k10 𝜔Sa

(7.43)

rG = k4 𝜔As + k7 𝜔Re + k9 𝜔Ar + k10 𝜔Sa

(7.44)

An oil-soluble Mo dispersed catalyst was employed for the slurry-phase hydrocracking of vacuum residue in a CSTR (1500 standard cm3 /cm3 ) to develop a kinetic model (Pham et al. 2021). The reaction scheme and the reaction rate equations are similar to those reported by Rodriguez-DeVecchis et al. (2017) with similar assumption of reaction second-order for the cracking of the residue, the only difference is the boiling points ranges of the lumps: R (524+ ∘ C), VGO (343–524 ∘ C), D (177–343 ∘ C), N (IBP-177 ∘ C). Coronel-Garcia et al. (2021) elaborated a kinetic study for the upgrading of heavy oil using nickel nanoparticles in polyethylene glycol (PEG300) as fine-dispersed catalyst in a batch reactor at moderate conditions. The reaction scheme and lumping approach are the same as Sánchez et al. (2005), while the reaction rate equations are similar to those reported by Rodriguez-DeVecchis et al. (2017), considering a different reaction order for the R fraction, because the authors found 2.5 to be the best order to fit the experimental data. In such a case, the system of differential equations is: rR = −(k1 + k2 + k3 + k4 )𝜔2.5 R rVGO = rD = rN = rG =

k1 𝜔2.5 R k2 𝜔2.5 R k3 𝜔2.5 R k4 𝜔2.5 R

(7.45)

− (k5 + k6 + k7 )𝜔VGO

(7.46)

+ k5 𝜔VGO − (k8 + k9 )𝜔D

(7.47)

+ k6 𝜔VGO + k8 𝜔D − k10 𝜔N

(7.48)

+ k7 𝜔VGO + k9 𝜔D + k10 𝜔N

(7.49)

The kinetic model for the hydrocracking of heavy residue using Mo-based ionic liquid as dispersed catalyst was developed by Cai et al. (2022). The reaction scheme is based on the model reported by Kim et al. (2017) (Figure 7.4c) employing different boiling points for D ( kj at T1 for T2 > T1

(7.130)

kj ≥ 0

(7.131)

(a4 k4 + a5 k5 + a6 k6 ) −

As

Stoichiometric coefficients were determined based on the ratio of average molecular weights of products and reactants concerning Eq. (7.113). Zhang et al. (2021) carried out an experimental study to determine the optimal operating conditions for aquathermolysis of heavy oil from Liaohe oil field using different catalysts (NiO, α⋅Fe2 O3 , and Co3 O4 ). Experimental data obtained at temperatures of 200, 220, 240, and 260 ∘ C and reaction times of 12, 24, 36, and 72 hours were used to develop a five-lump kinetic model based on a proposed reaction scheme (Figure 7.8d), which consider 11 reaction pathways to describe the distribution of SARA fractions and gases production for catalytic aquathermolysis reaction: dyAs (7.132) = −(k1 + k11 + k10 )yAs + k2 yRe dt dyRe = −(k2 + k3 + k5 + k7 )yRe + k1 yAs + k4 yAr (7.133) dt dyAr (7.134) = −(k4 + k8 + k9 )yAr + k3 yRe dt dySa = −k6 ySa + k10 yAs + k5 yRe + k8 yAr (7.135) dt dyGas = k11 yAs + k7 yRe + k9 yAr + k6 ySa (7.136) dt where the reactions rate constants shall meet Arrhenius equation and not be negative. Qu et al. (2021) carried out a kinetic study to analyze the behaviors of Mackay River oil sand during thermal cracking (absence of water) and aquathermolysis (with 30 wt% of water/oil ratio)

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions

processes at 200, 250, 300, and 350 ∘ C and reaction times at 6, 12, 18, and 24 hours. A kinetic model was proposed with 15 reaction pathways (Figure 7.8e): dyAs = −(k54 + k53 + k52 + k51 )yAs + k45 yRe + k35 yAr (7.137) dt dyRe = −(k45 + k43 + k42 + k41 )yRe + k54 yAs + k34 yAr + k24 ySa (7.138) dt dyAr = −(k35 + k34 + k32 + k31 )yAr + k53 yAs + k43 yRe + k23 ySa (7.139) dt dySa = −(k24 + k23 + k21 )ySa + k52 yAs + k42 yRe + k32 yAr (7.140) dt dyGas = k51 yAs + k41 yRe + k31 yAr + k21 ySa (7.141) dt Félix et al. (2022a) reported a kinetic model for noncatalytic aquathermolysis experiments of Ashal’cha heavy crude oil. The kinetic parameters for the 10 reaction pathways considered in the scheme presented in Figure 7.8f were estimated using experimental data obtained in a temperature range of 250–300 ∘ C and up to 72 hours of reaction time: dyAs = −(k1 + k2 + k3 + k4 )yAs (7.142) dt dy (7.143) rRe = Re = k1 yAs − (k5 + k6 + k7 +)yRe dt dy rAr = Ar = k2 yAs + k5 yRe − (k8 + k9 )yAr (7.144) dt dy (7.145) rSa = Sa = k3 yAs + k6 yRe + k8 yAr − k10 ySa dt dy rG = G = k4 yAs + k7 yRe + k9 yAr + k10 ySa (7.146) dt The kinetic study contemplates the irreversible chemical reaction for the global conversion of asphaltenes (Eq. (7.115)) in order to apply the following constraint for the calculation of kinetic coefficients: rAs =

k1 + k2 + k3 + k4 − kGAsp = 0

(7.147)

In addition, the reaction rate coefficients shall meet the temperature-dependence Arrhenius equation and be positive.

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions 7.5.1

Hydrocracking Kinetic Models

The estimation of kinetic parameter is an important step during kinetic modeling since there are several different ways to solve the differential equations and to optimize the values of these parameters. Until now, there are two main methods, both of which have shown good results, to solve the mass balance differential equations: analytical and numerical. Meanwhile, for the optimization process there are two approaches employed: linear and nonlinear regression. In this case, nonlinear regression has exhibited more reliable and better results compared with linear regression; nonetheless, care has to be taken since multiple solution can be obtained during this process (Ancheyta and Sotelo-Boyás, 2000; Félix et al. 2019; Sámano et al. 2020).

337

338

7 Catalytic Mechanism and Kinetics

The assumptions in kinetic modeling are made in order to simplify the reaction scheme or equations and thus the calculations. However, these assumptions cannot be done without the knowledge of the process whereby they must be based on the previous experimental information of the phenomena. Whereby, assumptions are another important aspect in the kinetic modeling since with nontheoretical assumptions, it can reach good results but no logical explanations of the behavior would be carried out. Therefore, the methodology for the estimation of kinetic parameters and the assumptions made are important aspects to be analyzed during kinetic modeling. In addition, the average absolute error (AAE, Eq. (7.148)) is calculated in most of the kinetic models reported in the literature to compare the fit with the experimental data despite the different configuration, operating conditions, feed properties, catalyst used, etc.: ∑nc | yexp,i −ycalc,i | i=1

yexp,i

x100 (7.148) N where nc is the number of components, ycalc,i and yexp,i are the calculated and experimental yields of component i, and N is the total number of observations. The differential equations stated in the Hassanzadeh and Abedi (2010) model were resolved by an analytical solution following the integrating factor technique; meanwhile, the Monte Carlo technique was employed to estimate the fittest reaction coefficient parameters in the nonlinear optimization of the objective function based on the sum of square errors (SSE, Eq. (7.149)), obtaining a considerable low difference of the data of experiments and calculation: AAE =

SSE =

nc ∑

( yexp,i − ycalc,i )2

(7.149)

i=1

The mass balance equations for the Galarraga et al. (2012) model were solved numerically by the trapezoidal rule, and the kinetic constants were optimized by the minimization of the objective function considering the following constraints: (1) All kinetic rate coefficients were fixed as nonnegative. (2) The sum of the kinetic constants involving the residue as reactant (k1 –k4 ) must be the same as the global kinetic rate coefficient of the residue conversion (k0 ). (3) All the kinetic rate coefficients must follow the temperature linearity of the Arrhenius equation, since the kinetic parameters at a given temperature must hold higher values than at lower temperatures. The model fits very well the experimental data, obtaining a low AAE less than 5%. The methodology followed by Loria et al. (2011) to estimate the kinetic rate coefficients was by using the quasi-linearization method, where the mass balance equations are converted into a matrix and multiplied by the matrix of kinetic reaction rate coefficients to solve the matrix system numerically. Furthermore, the least-square nonlinear method was employed to minimize the SSE equation and find the optimal kinetic reaction rate coefficients. The developed kinetic model fitted quite well to the experimental data of Athabasca bitumen upgrading since the AAE was less than 7%. Da Silva De Andrade (2014) developed a program in Visual Basic of EXCEL to calculate the kinetic parameters of the model for the catalytic in situ upgrading of Athabasca bitumen, deasphalting pitch, and vacuum residue. The differential equations were solved numerically, and the reaction rate coefficients were optimized using an objective function (OF, Eq. (7.150)) that was

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions

minimized by calculating the mass fraction with the collision factors and activation energies. These kinetic parameters were calculated by linear regression with the Arrhenius equation (Eq. (7.151)) using the reaction rate coefficients: ( ) nr ∑ OF = WPI PI + WR nr − (7.150) rj2 ( kj = A0,j exp −

EA,j

)

j=1

(7.151)

RT

where nr is the number of reactions, PI the performance index (normal objective function, SSE), rj2 the determination coefficient of reaction j, W PI the performance index weight factor, and W R the Arrhenius fitting index weight factor. The kinetic model presents good accuracy with experimental data, but it varied with each feedstock: less than 4% of AAE was reported for Athabasca bitumen and vacuum residue, while less than 8% of AAE was obtained for pitch. In order to validate the accuracy of the regression method reported by Da Silva De Andrade (2014), experimental data for Athabasca bitumen with an ultradispersed hydroprocessing catalyst published by Loria et al. (2011) were used. Figure 7.9 shows a comparison of the calculated profiles by both models with experimental data at 360 ∘ C, presenting an adequate adjustment with experimental values. However, the model of Da Silva De Andrade (2014) exhibits an AAE of 3.11%, while the model of Loria et al. (2011) obtained a value of 6.25%. Da Silva De Andrade (2014) presented the best precision of both models due to the higher level of reactivity revealed in the lighter oil cuts (VGO and distillates fractions). The work of Orozco Castillo (2016) lacks information about the methods used for the optimization of the kinetic rate coefficients or for solving the mass balance equation, while the kinetic modeling for the experimental data of Rodriguez-DeVecchis et al. (2017) was done following the same methodology as Da Silva De Andrade (2014) obtained an AAE less than 7%. The same methodology was employed for solving the set of ordinary differential equations and the optimization of the reaction rate coefficients in the kinetic model published by Elahi et al. (2019). The obtained kinetic model fits well to the experimental data with a maximum AAE of 6%. Furthermore, Duran 0.6

Mass fraction (%)

0.5 0.4 0.3 0.2 0.1 0 0

Figure 7.9

5

10

15 20 25 Reaction time (h)

30

35

40

Comparison of experimental and calculated data with kinetic parameters of ( ) Adapted from ∘ × ) Da Silva De Andrade (2014) at 360 C: ( ) Residue, ( ) Gases, ( ) Naphtha, ( ) Loria et al. (2011) and ( Distillates, ( ) VGO.

339

340

7 Catalytic Mechanism and Kinetics

Armas (2021) also employed the method of Da Silva De Andrade (2014) for solving the ordinary differential equations, using a usual objective function based on SSE. However, the obtained kinetic model has a considerable deviation from the experimental data since the AAE was around 9%. The Petzold–Gear backward differentiation formulae (BDF) method was used to solve the set of ordinary differential equations extracted from bass balance for the kinetic model proposed by Nguyen et al. (2013), whereas the Levenberg–Marquardt algorithm was performed for the optimization of reaction rate coefficients. The stoichiometric coefficients of G lump (v1G , v2G , v3G , v4G , v5G , v6G ), which are temperature-independent, were determined by stoichiometric equations, while other stoichiometric coefficients were accomplished considering constraints subject to the mass balance closure. Since the hydrogen consumption displayed high errors for some reactions, the stoichiometric coefficients for this compound were not well estimated. Despite this, the kinetic model represents very well the experimental results at the conditions studied. To solve the system of the mass balance differential equations in the kinetic model proposed by Asaee et al. (2014), a numerical solution similar to Loria et al. (2011) was used subject to the minimization of an objective function (Eq. (7.152)) similar to those used by other works. In this function, the weighting factors (wi ) are employed to allow the equal error distribution between the different fractions in the objective function. The calculation of yields was done using the activation energies and collision factors estimated with reaction rate constant by linear regression following the Monte Carlo approach and then optimizing with the interior point algorithm. A constraint is considered for the estimation of the activation energies, where the activation energies value of heavier lumps conversion has to be lower than the cracking of lighter lumps, and the values for the generation of high molecular weight compounds must-have higher activation energies than production of the lighter compounds. A sensitivity analysis was done to establish that the kinetic parameters were correctly estimated and the accuracy provided by the model is good enough to predict the experimental data with the AAE less than 9%: OF =

nc ∑

wi ( yexp,i − ycalc,i )2

(7.152)

i=1

There is no detailed information for the method to solve the system of differential equations proposed by Kim et al. (2017). The algorithm of Levenberg–Marquardt was employed to optimize the kinetic parameters and give a good enough accuracy to fit the experimental data. The kinetic model reported by S. Huang et al. (2017) follows the same method to solve and optimize the kinetic parameter than Hassanzadeh and Abedi (2010) (Monte Carlo approach), and the accuracy of the kinetic model provided by the optimized reaction rate constants seems to be good enough to fit the experimental data since the AAE of the model was not provided. The proposed model by Ortega-Garcia et al. (2017) was capable of replicating the particular experimental data satisfactorily following a methodology for kinetic parameters estimation, where the use of a combination of the Runge–Kutta fourth-order integration and the Marquardt–Levenberg optimization methods in order to minimize an objective function similar to aforementioned with the only difference that the SSE is multiplied by a covariance matrix. The methodology to achieve the best values of reaction rate coefficients reported by Félix and Ancheyta (2019a,b) consists of the following: (i) the initial values are obtained using the Monte Carlo algorithm, (ii) a fourth-order Runge–Kutta procedure was used to find the solution of the differential equations, (iii) the nonlinear regression optimization is performed by Levenberg–Marquardt algorithm to determine the appropriate values of kinetic coefficients for the experimental data, and (iv) a sensitivity analysis was carried out as soon as the parameter values have been determined to guarantee that they present the best values. Using this methodology, the

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions

four-lump kinetic model based on boiling points ranges predicts accurately the experimental data due to an AAE less than 3.3% that was displayed at the temperatures studied. The kinetic models based on SARA fractions that were developed predict well enough the experimental data: four-lump and five-lump (with 12 reaction rate constants) models present an AAE of 5.4%, the five-lump model with 9 reaction rate constants presents an AAE of 14.7%, while the six-lump model presents an AAE of 16.6%. This difference in the AAE between models is because the C fraction, present in the models of 9 (five-lump model) and 13 (six-lump model) reaction rate coefficients, is difficult to predict due to the high sensibility to the temperature. A similar method to estimate the kinetic parameters than was done by Félix and Ancheyta (2019a, 2019b) was carried out by Pham et al. (2021), and even a sensitivity analysis was accomplished to assure the optimal values of the kinetic reaction rate coefficients, obtaining an AAE less than 10% for the model. An analytical method was applied in the model of Coronel-Garcia et al. (2021) to solve the differential equations, and the Levenberg–Marquardt algorithm was employed in the nonlinear optimization of the kinetic parameters. Furthermore, in order to obtain the optimal values of the reaction rate constant, a sensitivity analysis was implemented, and the model presents a low value of AAE (less than 5%) showing a good fit of the experimental data. The methodology followed by Cai et al. (2022) employs a numerical method (Runge–Kutta–Fehlberg) to solve the set of differential equations extracted from the reaction pathways, and the Universal Global Optimization method was used to estimate the reaction rate coefficients by nonlinear regression. In order to ensure the global minimum of the objective function, a sensitivity analysis was carried out with the estimated kinetic parameters, and the AAE obtained for the kinetic model was 18.1% due to the difficulty to predict the C fraction. Alvarez et al. (2019) used numerical methods to solve mass balance equations and nonlinear regression to estimate the kinetic parameter minimizing the variance among of the experimental and calculated information with Eq. (7.149). The estimation of the stoichiometric coefficients in the reaction rate equations was carried out in two different ways: one stoichiometric coefficient for each reaction was done numerically, and the second coefficient (G lump), to complete the mass balance closure, was calculated stoichiometrically. A substantial error occurs since some of the experimental data points seem to be underestimated or overestimated by the model. In the methodology followed by Browning et al. (2019), the kinetic parameters for the model reported were estimated by nonlinear least squares optimization employing the trust-region-reflective algorithm minimizing the objective function given in Eq. (7.149). The kinetic model fits the experimental data for almost all lumps except for the gas lumps (G and H2 S). In the case of the continuous mixtures model developed by Martinez-Grimaldo et al. (2011), there is no information about the methodology followed to solve the equations presented or the optimization algorithm used for parameter estimation. However, the obtained parameters fit the experimental points at short reaction times (below 7.5 hours), and when the reaction time exceeds this value, the error increases considerably. All these methodologies used in the kinetic parameter estimation during the kinetic modeling for in situ and ex situ hydrocracking of heavy oils with dispersed catalyst and hydrogen atmosphere are summarized in Table 7.5.

7.5.2

Aquathermolysis Kinetic Models

During the development of kinetic models focus was on aquathermolysis reaction, and some assumptions have been considered for the simplification of the models as well as the different methods that were used to estimate the kinetic parameters; nonetheless, some of these studies do not describe in detail the methodology that was used.

341

342

7 Catalytic Mechanism and Kinetics

Table 7.5

Methods used to estimate kinetic parameters.

Reference

Number of parameters

Reaction order

AAE (%)

Hassanzadeh and Abedi (2010)

10

1

—

Galarraga et al. (2012)

10

1

5

6

1

—

10

1

9

Nguyen et al. (2013) Asaee et al. (2014) Kim et al. (2017)

Optimization algorithm

Monte Carlo EXCEL SOLVER LM Interior point

7

2

—

LM

T. Huang et al. (2017)

10

1

—

Monte Carlo

Coronel-García et al. (2021)

10

2.5

5

LM

Félix and Ancheyta (2019a)

6

1

3.3

LM

Félix and Ancheyta (2019b)

6

1

5.4

LM

Félix and Ancheyta (2019b)

9

1

14.7

LM

Félix and Ancheyta (2019b)

12

1

5.4

LM

Félix and Ancheyta (2019b)

13

1

16.6

LM

Álvarez et al. (2019)

5

1

—

—

Browning et al. (2019)

5

1

—

Trust-region-reflective

Loria et al. (2011)

10

1

11.9

—

Rodriguez-DeVecchis et al. (2017)

10

2

7

—

Da Silva De Andrade (2014)

10

1

4

—

Orozco Castillo (2016)

10

1

7

—

Elahi et al. (2019)

10

1

6

—

Duran Armas (2021)

10

1

9

—

Ortega García et al. (2017)

10

1

—

LM

Pham et al. (2021)

10

2

10

LM

7

2

18.1

Universal Global Optimization

Cai et al. (2022) LM = Levenberg-Marquardt.

Most of the kinetic studies for aquathermolysis assume that the steam is in excess in the system, so the water concentration could be considered constant and included in the reaction rate coefficients. Therefore, the reaction rate equations become first-order reactions. In the reported models by Lin et al. (2020), Hamedi-Shokrlu and Babadagli (2014), and Zhang et al. (2020), the estimation of the kinetic parameters was carried out by integrating the batch reactor mass balance equation (Eq. (7.72)) in order to predict the gases formation from different feeds during aquathermolysis reaction: 1 𝛼 ∕n = kt + c

(7.153)

where c is an integration constant. Different values of reaction orders were assumed in Eq. (7.153) to evaluate their fit with the experimental data (mainly H2 S generation) through the least-squares method. The reaction order value that presents the best fit through the determination coefficients (r 2 ) is substituted into Eq. (7.153). In the same way, the values of kinetic coefficients for each temperature were estimated based on gas production data relating to reaction time, while activation energies and preexponential factors were determined using linear regression of the Arrhenius equation.

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions

The kinetic model reported by Lamoureux-Var and Lorant (2007) was reduced using six closure equations and fixing the preexponential factors to an arbitrary but realistic value: Aa1 = Aa2 = Ab1 = Ab2 = 1014 s−1

(7.154)

The estimation was carried out through the experimental data obtained for aquathermolysis of rock samples and an extended Levenberg–Marquardt algorithm, where quadratic error functions were defined to minimize the difference between the measured and calculated mass distribution values. The results of these kinetic parameters were used in the model reported by Barroux et al. (2013) as the initial parameter values to fit the kinetic model to experimental data were obtained during the aquathermolysis experiments on an Athabasca oil sand. On the other hand, Lamoureux-Var and Lorat (2005) do not report a solution of the described kinetic model. Belgrave et al. (1997) utilized a hybrid of Gauss–Newton and steepest descent techniques to estimate the stoichiometric and kinetic coefficients by the least squares minimization of the SSE. On the other hand, Kapadia et al. (2010) solved the system of 18 chemical reactions that determined initial values for each reaction parameter by converting pseudocomponents at each reaction time and temperature. The objective function was defined as the SSE at each time for all experiments; so the minimization of the global residual by a subspace trust-region method was performed based on the interior-reflective Newton method for parameters estimation. According to the Kapadia et al. (2013) model, the kinetic parameters for the reversible water–gas shift reaction (k7 –k8 ) were taken from the literature (Guntermann 1982; Hajdo et al. 1985), while the Levenberg–Marquardt method was used to determine the catalytic decay parameters by minimizing the SSE. However, since only the H2 S generation profile was reported at certain reaction times and temperatures, the decay parameters for other gases were estimated by assuming that their evolution during aquathermolysis followed the same trend as that for the percentage of bitumen gasified. Therefore, two decay parameters for H2 S generation and two for other gases were obtained. The differential equation system of the Zhang et al. (2021) model (Eqs. (7.132)–(7.136)) was solved by the numerical Runge–Kutta method, which produces better results in fewer steps. The objective function was defined by the minimum of SSE. The kinetic models developed by Tirado et al. (2022b) and Félix et al. (2022a) utilized a sequential methodology to reduce the number of kinetic parameters to be estimated simultaneously, avoiding convergence problems during the calculation and optimization of kinetic coefficients. The global asphaltenes conversion by linear and nonlinear regression analysis in order to meet the constraints of the model was estimated. The estimation of the kinetic parameters for nonlinear regression was performed using an optimization algorithm (Figure 7.10) capable of guaranteeing that the calculated kinetic parameters present the minimum difference between experimental and

Monte carlo Reported values Initialization of parameters

Experimental data, objective function and reaction equations

Perturbations Graphical analysis Sensitivity analysis

Nonlinear optimization Interior point algorithm Constraints

Statistical analysis Residual Parity plot

Figure 7.10

Optimal kinetic parameters Yes

Global minimum of the objective function?

No

General algorithm to ensure the optimal value of the estimated kinetic parameters.

343

344

7 Catalytic Mechanism and Kinetics

calculated data. The algorithm is based on nonlinear optimization methods, statistical analysis, and sensitivity analysis. Both models applied the average absolute error (AAE, Eq. (7.148)) as the objective function to provide uniform importance to each component during optimization. Since the use of objective functions, such as SSE, may cause the optimization algorithm to focus on those parameters of components with higher magnitude order, limiting the optimization of other parameters. This point was considered due to the difference in yields between gases generated and SARA fractions during aquathermolysis reaction. The calculation and optimization of kinetic coefficients were performed by a Runge–Kutta method and trust-region-reflective algorithm, respectively. This optimization algorithm was selected for its capability to handle bound constraints.

7.5.3

Methodology to Ensure the Optimal Set of Kinetic Parameters

For kinetic modeling, nonlinear optimization is the most used method to estimate the kinetic parameters. This is done by minimizing an objective function involving the experimental and calculated data. Moreover, most of the proposed kinetic models have several kinetic parameters to be estimated. Thus, multiple solutions can be obtained, and the initial value or guess during the solution of the nonlinear process is the crucial step to estimate the optimal solution (global minimum of the objective function). Therefore, a recommended general methodology (Figure 7.10) to ensure the optimal set of the estimated kinetic parameters is described in detail in this section. Previously to start the optimization process, the data employed for a kinetic model must be precise and adequate (the mass balance closure must be appropriate and expressed in yield preferably). A reaction scheme must be proposed based on the knowledge of the reactions taking place. Once the reaction rate equations are defined from the reaction scheme and the experimental data, the kinetic parameters can be estimated. The first step in the optimization process is the initial guess of the kinetic parameters, which is commonly done by using reported values (if the kinetic model is reported in the literature) or by the Monte Carlo algorithm. In this algorithm, an initial value (usually 1 or 0.1) is considered for all parameters, and then a random number is established as an initial value for one parameter, keeping the original values of the other parameters. The mass balance equations are solved with the new parameter value, and the objective function is evaluated, which is done at least 1000 times for each parameter (Alcázar and Ancheyta 2007; Félix et al. 2019; Félix et al. 2022a, b; Sámano et al. 2020; Tirado et al. 2022a). As mentioned previously, analytical and numerical methods are the most used to solve the mass balance equations. The use of the analytical methods is done by integrating the differential equations when are exact; however, when these are inexact, the integrating factor can be employed to convert them into exact differential equations, whereas the numerical methods used to solve the differential equations are several, as can be seen in Table 7.5. Nonetheless, the most employed method among all these reported is the Levenberg–Marquart (Marquardt 1963) due to the good results obtained and low iterations needed (Hassanzadeh and Abedi 2010; Sámano et al. 2020). There are different types of objective functions used for kinetic modeling, as those described previously; nevertheless, all of them are based on the comparison of experimental and calculated data. The most employed objective function is SSE; however, this objective function presents a disadvantage with dissimilar yield values, i.e. different orders of magnitude, which may sometimes result in wrong parameter values due to SSE that gives more importance to some components than others when optimizing. In other words, the optimization algorithm tends to optimize those parameters in which components with higher yields are present (Asaee et al. 2014; Coronel-García et al. 2021; Sadighi and Zahedi 2013). Félix et al. (2022b) and Tirado et al. (2022a) used the AAE as an objective function, which provides more balanced weighting to all components, and the difference

7.5 Methods for Calculating Kinetic Parameters and Model Assumptions

in the order of magnitude is substantially reduced. This is also done by assigning weight factors to each lump in the SSE. There are several types of weighting methods, but the most employed are 1∕y2exp , and the ratio of the arithmetic mean yields of residue to the arithmetic mean yields of that lump at each temperature rounded to an integer (Asaee et al. 2014; Gauthier et al. 2006). In order to prevent a local minimum of the objective function, where a nonlinear algorithm wrongly obtains an optimal value, a sensitivity analysis should be performed on each kinetic parameter value. This is done by applying a perturbance (typically in a range of ±50%) in the value of one parameter, maintaining the original values of the other parameters, and evaluating the value of the objective function used with these new values. A plot is performed with the value of the objective function in the “y” axis versus the percentage of perturbation in the “x” axis. The optimal value of the kinetic parameters (or the global minimum of the objective function) is obtained when the lowest value of the evaluated function is reached at 0% of perturbation (original value). Otherwise, a new optimization process needs to be done to the parameters since the minimum value was obtained in a different perturbation value for any parameter (Alcázar and Ancheyta 2007; Félix et al. 2019; Sámano et al. 2020). Additionally, for a sensitivity analysis, a statistical analysis is performed, and the residuals at different reaction times are calculated (Eq. (7.155)) and plotted against the number of experiments to observe any pattern in the residual values. For the optimization to be considered accurate, the residual values must not show any pattern, the numbers of positive and negative residuals must be balanced to avoid under- and overestimations, and the residual range must be low for better accuracy of predictions. A parity plot is also considered in the statistical analysis whereby the slope and the y-intercept are calculated, which must be close to unity and zero, respectively: Residual =

nc ∑

yexp,i − ycalc,i

(7.155)

i=1

In order to highlight the importance of an appropriate estimation of the kinetic parameters, two cases (aquathermolysis and hydrocracking of heavy oils) reported in the literature illustrate the use of the proposed methodology for estimating the best set of reaction rate coefficients based on sensitivity analysis. 7.5.3.1

Study Case in Hydrocracking of Heavy Oils

Félix et al. (2019) analyzed the kinetic data of Hassanzadeh and Abedi (2010) using the optimization algorithm described (Figure 7.10) to calculate the reaction rate coefficients and lower the difference between the experimental and calculated data. Through the set of numerical methods, such as the Monte Carlo method, Runge–Kutta method, Levenberg–Marquardt Algorithms, and sensitivity analysis, the optimized kinetic parameters displayed a lower difference regarding the experimental data, as is observed in Figure 7.11. In addition, the kinetic parameters optimized by the algorithm presented a dependence on temperature according to the Arrhenius equation. 7.5.3.2

Study Case in Aquathermolysis of Heavy Oils

Tirado et al. (2022c) estimated the kinetic parameters to predict the distribution of SARA fractions of oil and gas generation for the catalytic aquathermolysis of Liaohe extra-heavy crude oil using experimental data reported by Zhang et al. (2021) and the optimization algorithm (Figure 7.10). The differential equation system (Eqs. (7.132)–(7.136)) was solve using the Runge–Kutta method, Levenberg–Marquardt algorithm, and the average absolute error as an objective function. The results of the parameters exhibited a good fit with experimental data, obtaining values of AAE lower than 4%, as is shown in Figure 7.12. In addition, the kinetic coefficient values presented greater agreement with the reaction mechanisms reported for the aquathermolysis reaction,

345

7 Catalytic Mechanism and Kinetics

Calculated mass fraction

0.5 0.4 0.3 0.2 0.1 0 0

0.1

0.2 0.3 Experimental mass fraction

0.4

0.5

25

45

20

40

Resins yield (%)

Asphaltenes yield (%)

Figure 7.11 Parity plot of the kinetic model reported by ( ) Adapted from Hassanzadeh and Abedi (2010) and optimized by ( ) Félix et al. (2019) with the same experimental data.

15 10 5 0

0

10

20

30

40

50

60

70

30 25 0

10

20

0

10

20

30

40

50

60

70

80

30 40 50 60 Reaction time (h)

70

80

Saturates yield (%)

40

35

30 25

20 15

35

20

80

40 Aromatics yield (%)

0

10

20

30 40 50 60 Reaction time (h)

70

80

35 30 25 20 15

0.2 Gas yield (%)

346

0.16 0.12 0.08 0.04 0

0

10

20

30 40 50 60 Reaction time (h)

70

80

Figure 7.12 Comparison of (symbols) experimental and (lines) prediction yields with kinetic parameters obtained by Tirado et al. (2022c) at 200 ∘ C ( ), 220 ∘ C ( ), 240 ∘ C ( ), and 260 ∘ C ( ).

7.6 Results and Discussion

indicating that the light fractions are mainly produced from asphaltenes and resins. However, a comparison with the trends reported by Zhang et al. (2021) was not carried out since the reported kinetic parameter values do not present adequate trends to predict the experimental data.

7.6

Results and Discussion

7.6.1

Hydrocracking

7.6.1.1

Global Reaction Order for Residue Conversion

The reviewed kinetic models for hydrocracking of different heavy feedstock were developed for ex situ and in situ applications using nanocatalysts in hydrogen atmosphere. The residue or asphaltenes conversion increases as reaction conditions increase in severity, being lower at moderate conditions. Meanwhile, as can be seen in Figure 7.13, the conversion increases exponentially with the increase of temperature and residence time at severe operating conditions. The different values in maximum conversion (Figure 7.14) reported in all studies depend on the operating conditions and settings, such as temperature and time of reaction, feedstock properties, reactor type, catalyst, and hydrogen pressure. The heavy fraction (residue or asphaltenes) experimental conversion of heavy crude oils is used to obtain the reaction order obtaining different values depending on the catalyst used and the reaction conditions. For instance, during the hydrodesulfurization reaction of individual model compounds, the reaction order obtained is commonly one; however, when blends of model compounds or petroleum distillates are used, the reaction order tends to be two. The individual reaction rate of each reactive apparently causes some synergy increasing the reaction order. The order of reaction for hydrocracking of petroleum fractions varies depending on the reaction

50.00 40.00

310 °C 320 °C 330 °C 335 °C 340 °C 350 °C

30.00 20.00 10.00 0.00

0

30

60 90 120 Residence time (h)

150

Residue conversion (%)

Residue conversion (%)

60.00

180

100 90 80 70 60 50 40 30 20 10 0

360 °C 365 °C 370 °C 380 °C 390 °C

0

20

40

120

140

160

(b)

100 90 80 70 60 50 40 30 20 10 0

100.00

395 °C 400 °C 405 °C 410 °C 420 °C

Residue conversion (%)

Residue conversion (%)

(a)

80 100 60 Residence time (h)

90.00 80.00 425 °C 430 °C 435 °C 440 °C 450 °C

70.00 60.00 50.00 40.00 30.00

0

10

20 30 40 Residence time (h)

(c)

50

60

0

10

20 30 40 Residence time (h)

50

60

(d)

Figure 7.13 Residue conversion at different residence times and reaction temperatures: (a) 310–350 ∘ C, (b) 360–390 ∘ C, (c) 395–420 ∘ C, and (d) 425–450 ∘ C.

347

an za de h a G nd Residue or Asphaltene conversion (%) D a aS la Ab rra ed ilv i g (2 a D aD L et a 01 aS e ilv An oria l. ( 0) O ro a D drad et a 201 zc 2 o e A e (2 l. (2 ) Ca nd 0 0 sti rad 14) 11) Ro Oro llo ( e (2 (pit dr zc 20 01 ch o ) 1 ig ue Ca 6) 4) ( zsti (H VR D eV llo eav ) ec (20 y o i ch 1 is 6) l) ( El et a VR l a ) D hi e . (20 ur an t al 17) N Ar . (2 gu m 01 9 ye as n ( ) A et a 202 sa e l. 1) T O . H e et (201 rte ua 3) n al g Fé a-G g e . (20 ta lix a 1 an rcia l. ( 4) 2 d A et a 017 nc l. he (2 ) Co ro Ph yta 017 ne am ( 2 ) l A -Ga et 019 lv ar rcia al. ( a) e 2 A z e et a 02 lv 1 t ar al. l. ( ) 2 ez ( et 201 021 ) 9 al Fé . ( ) (A 20 lix R an Cai 19) ) d A et a (SR nc l he . (2 ) yt 02 a ( 2) 20 19 b)

7 Catalytic Mechanism and Kinetics

100

91.9

80

74.4

56.6

52.9

50.3

44

40

80.6

90.5

84.5

82.90

83.3

65.3

62.5

60

90.9

88.1

31.5

46.4

47

35.3

38.2

32.3

20 0

H as s

348

Figure 7.14 Maximum ( ) residue or ( ) asphaltenes conversion in experiments for developing of the kinetic models.

conditions, catalysts, and the complexity of the feedstocks used, i.e. heavy oil, extra-heavy oil, residue, particularly in terms of the content and properties of asphaltenes (Fogler 2020). The first-order of reaction assumption is done by almost all the kinetic models reported and only the kinetic models described by Rodriguez-DeVecchis et al. (2017), Kim et al. (2017), Pham et al. (2021), and Cai et al. (2022) assumed a second order of reaction for the residue conversion, while the experimental data described by Coronel-Garcia et al. (2021) follow a 2.5 order of reaction for residue conversion. Suggesting the reaction rate for residue fraction in all these experiments varies different from the concentrations obtained. This is because when the reaction order obtained is higher than 2, the experiments for kinetics with are done with ultradispersed nanocatalysts (with hydrogenation activity) which hydrogenate the free radical (which varies with the concentration of residue) generated by the high temperature (above 400 ∘ C) when high molecular weight compounds cracks (Félix and Ancheyta 2019a; Martínez and Ancheyta 2014). In addition, experiments for kinetics with reaction order of 2 (Rodriguez-DeVecchis et al. 2017) were performed in a porous medium (silica sand) with catalytic behavior, thus increasing the apparent reaction order similar to supported catalyst facilitating the cracking of reagents (Marques et al. 2011). However, the reaction order assumption is only verified for few works (Duran Armas 2021; Elahi et al. 2019; Orozco Castillo 2016), as can be seen in Figure 7.15. In order to corroborate the reaction order supposition, the conversion (X) or yield fraction ( y) and the reaction or residence time are employed, by plotting ln(1/(1 − X)), − ln (1 − X), or ln( yR, 0 /yR )(versus ) 𝜏 or t for first order ( n−1 ) n−1 of reaction (being k0 the slope of the straight line), and 1∕yR − 1∕yR,0 versus 𝜏 or t for nth-order of reaction (being (n−1)k0 the slope). This methodology was carried out to some of the experimental data reported to verify the assumption for the order of reaction, and the conclusions are displayed in Figures 7.16−7.18. Among all the kinetic models, only one work (Ortega García et al. 2017) reported the first order of reaction (Figure 7.16c) correctly since the experimental data have a better fit than the second order of reaction (Figure 7.16g); while for other kinetic models (Asaee et al. 2014; S. Huang et al. 2017; Nguyen et al. 2013), which supposed first order of reaction (Figures 7.16a–c), the experimental data were more accurately fitted with a second order of reaction (Figure 7.16e,f,h). One difference between the reported experimental data is that

7.6 Results and Discussion

0.4 0.3

320 °C

0.2

340 °C 360 °C

0.1

1n (ωR0/ωR)

0

0

20

40 60 80 Residence time (h) (a)

1n (ωR0/ωR)

0.8

0.5

0.6 340 °C

0.2

0.5

0.4

0.4

0.3

0.3

320 °C

0.2

340 °C

0.1

360 °C

0

0

20 40 60 Residence time (h) (c)

80

320 °C

0.4

0

100

1n (ωR0/ωR)

1n (ωR0/ωR)

0.6

360 °C

0

20

40 60 80 100 120 Residence time (h) (b)

335 °C

0.2

350 °C

0.1 0

365 °C

0

20 40 60 Residence time (h) (d)

80

Figure 7.15 Verification of first-order assumption for the global residue conversion reaction. (a) Orozco Castillo (2016)/University of Calgary for heavy oil, (b) Orozco Castillo (2016)/University of Calgary for vacuum residue, (c) Elahi et al. (2019)/Elsevier, (d) Duran Armas (2021)/University of Calgary.

the experiments for first order of reaction were obtained at moderate conditions (temperature lower than 400 ∘ C) in which hydrogenation reactions predominate, while for reaction order of 2 or higher, the experimental information were acquired at severe conditions (temperatures higher than 400 ∘ C) where more hydrocracking-dominated reactions are carried out. In Figures 7.17 and 7.18, three different reaction orders (1, 2, and 2.5) were tested for the kinetic data of Pham et al. (2021), Rodriguez-DeVecchis et al. (2017), and Cai et al. (2022), where a reaction order of 2 was assumed (Figures 7.17c,d, and Figure 7.18d, respectively). Nonetheless, the reaction order of 2.5 shows the best fit for the experimental residue conversion in the three cases (Figures 7.17e,f and 7.18f). It was stated that the high value of the reaction rate is because the sand and liquid ionic catalysts act as a bifunctional catalyst at moderate conditions: cracking the molecules and hydrogenating the free radicals. Coronel-Garcia et al. (2021) assumed the reaction order to be 2.5 (Figure 7.18e), whereas the fit of the experimental residue conversion is better for first or second order (Figure 7.18a,c, respectively) since both have a similar adjustment to the experimental data. Although a study was carried out to identify the order of reaction from the experimental data, it was inappropriately conducted because only the experimental residue is plotted against t at different temperatures and orders of reaction (0.5–3), and the data behavior can be different at different temperatures hindering the observation of the fit; meanwhile, with linear regression analysis, all the data should follow a straight line at different temperatures and the fit is better observed. 7.6.1.2

Calculation of Reaction Rate Coefficients

The reaction rate coefficient values in all kinetic models were estimated and optimized by different methods, and any of these methods is effective to achieve the best set of kinetic parameters as long as they are suitably applied, and all the imposed constraints are taken into account. Hence, it is a crucial step to ensure that the set of reaction rate coefficients is optimal for experimental

349

7 Catalytic Mechanism and Kinetics

0.020

R2 = 0.909

0.4 R2 = 0.9656

420 °C 430 °C

0.2

1/ωR–1/ωR0

1n (ωR0/ωR)

0.6

0.0 0.2 0.4 0.6 0.8 Residence time (h)

(a)

3.0 2.5 2.0 1.5 1.0 0.5 0.0

R2

405 °C 420 °C

R2 = 0.8354

0

2 4 6 Residence time (h)

435 °C

R2

2.0

0.010

420 °C R2 = 0.9764

0

0.2 0.4 0.6 0.8 Residence time (h) R2 = 0.9653

R2 = 0.9836

1.0

350 °C 370 °C

R2 = 0.9853

0.5 0.0

405 °C 420 °C 435 °C

R2 = 0.9593

2 4 6 Residence time (h)

8

(f)

0.25

1.5

1

(e)

0.12 0.10 0.08 0.06 0.04 0.02 0.00 0

= 0.9885

430 °C

0.005

8

(b)

2.5 1n (ωR0/ωR)

= 0.9263

1/ωR–1/ωR0

R2 = 0.9488

R2 = 0.9317

0.015

0.000

1

1/ωR–1/ωR0

1n (ωR0/ωR)

0

0.20

R2 = 0.9483

0.15 350 °C 370 °C

0.10 R2 = 0.9601

0.05 0.00

0

30

60 90 120 150 180 Residence time (h)

0

30

60 90 120 150 180 Residence time (h)

(g)

(c) 0.035

1.6 R2 = 0.9434

1.2

R2 = 0.9662

0.8

R2

390 °C 405 °C 420 °C 435 °C

= 0.9812

R2 = 0.8808

0.4

1/ωR–1/ωR0

1n (ωR0/ωR)

350

0.028

R2 = 0.9956

0.021

390 °C 405 °C

R2 = 0.9926

0.014

420 °C 435 °C

R2 = 0.9981 R2 = 0.9033

0.007 0.000

0.0 0

1 2 3 Residence time (h)

(d)

4

0

3 1 2 Residence time (h)

4

(h)

Figure 7.16 Linear regression analysis for first-order of reaction with the residue conversion data of (a) Adapted from Nguyen et al. (2013)/Elsevier, (b) Adapted from T. Huang et al. (2017)/Elsevier, (c) Adapted from Ortega-Garcia et al. (2017)/American Chemical Society, (d) Adapted from Asaee et al. (2014)/Elsevier; second-order of reaction in residue conversion of (e) Adapted from Nguyen et al. (2013)/Elsevier, (f) Adapted from T. Huang et al. (2017)/Elsevier, (g) Adapted from Ortega-Garcia et al. (2017)//American Chemical Society, (h) Adapted from Asaee et al. (2014)/Elsevier.

data, and for this, it is necessary to carry out a series of steps described in Section 7.5.3 (Félix et al. 2019; Sámano et al. 2020). The results of the kinetic parameters for the hydrocracking kinetic models based on the lumping approach are summarized in Table 7.6. Most of the parameters for the hydrocracking kinetic models were in general properly obtained, since the values of the reaction rate coefficients increase with the temperature. Nonetheless, some works (Coronel-García et al. 2021; Galarraga et al. 2012; Hassanzadeh and Abedi 2010; Kim et al. 2017; Loria et al. 2011; Orozco

7.6 Results and Discussion

R2 = 0.7477

1.5

R2 = 0.9289

1.0

410 °C 425 °C 440 °C 450 °C

R2 = 0.9202

0.5

R2 = 0.9694

1/ωR–1/ωR0

1n (ωR0/ωR)

2.0

0.0 0

1 2 3 Residence time (h)

4

0.018 0.015 0.012 0.009 0.006 0.003 0.000

R2 = 0.9103

0

(a) R2 = 0.8481

340 °C 350 °C 360 °C 370 °C

0.4 R2

= 0.7812

0.0

R2 = 0.869

0.015 0.010

R2 = 0.9803 R2 = 0.9921

0.005

10 20 30 40 Residence time (h)

50

0

R2 = 0.8427

410 °C 425 °C 440 °C 450 °C

R2 = 0.9955 R2 = 0.9675 R2 = 0.9866

0

1 2 3 Residence time (h)

(c)

1 2 3 Residence time (h)

4

(e)

4

1/ωR1.5–1/ωR01.5

(b) 1/ωR–1/ωR0

410 °C 425 °C 440 °C 450 °C

R2 = 0.9993

0.000 0

0.06 0.05 0.04 0.03 0.02 0.01 0.00

50

(d)

0.6

0.2

10 20 30 40 Residence time (h)

0.020 1/ωR1.5–1/ωR01.5

1n (ωR0/ωR)

0.8

340 °C 350 °C 360 °C 370 °C

R2 = 0.8201

0.005

R2 = 0.9358

0.004

340 °C 350 °C 360 °C 370 °C

0.003 0.002

R2 = 0.8389

0.001 0.000 0

10 20 30 40 Residence time (h)

50

(f)

Figure 7.17 Linear regression analysis for first-order of reaction with the residue conversion data of (a) Adapted from Pham et al. (2021)/Elsevier, (b) Adapted from Rodriguez-DeVecchis et al. (2017)/Elsevier; second-order of reaction in residue conversion of (c) Adapted from Pham et al. (2021)/Elsevier, (d) Adapted from Rodriguez-DeVecchis et al. (2017)/Elsevier; 2.5 order of reaction in residue conversion of (e) Adapted from Pham et al. (2021)/Elsevier, (f) Adapted from Rodriguez-DeVecchis et al. (2017)/Elsevier.

Castillo 2016) do not follow this dependency with temperature and present incongruences concerning the Arrhenius equation law, whereas for the continuous lumping approach, the parameters are presented in Table 7.7. These parameter values are highly dependent on the composition of the feedstock, the catalyst used, and the reaction conditions. Since certain kinetic models present similar reactions pathways, some correlations between the individual reaction rate coefficients (k1 , k2 , k3 , k4 , k5 , k6 , k7 , k8 , and k9 ) for these kinetic models and the global reaction rate coefficients for residue hydrocracking (kR = k1 + k2 + k3 + k4 ), for VGO hydrocracking (kVGO = k5 + k6 + k7 ), and for distillates hydrocracking (kD = k8 + k9 ) were calculated in order to find a tendency of the reported reaction rate coefficients with the temperature using the ki /kglobal ratios. As can be seen in Figure 7.19, k1 /kR values are in the range of 0.4–1 for almost all kinetic models, while for k2 /kR , the values are in the range of 0–0.4, for k3 /kR are lower than 0.2, and for k4 /kR are lower than 0.1. Suggesting that the residue conversion is more selective to generate VGO than D, N, and G. Moreover, it can be seen that some ratios increase with temperature and others decrease depending on the individual reaction rate coefficients, which is mainly due to some of these parameters are more sensitive to temperature than others.

351

7 Catalytic Mechanism and Kinetics

0.08 R2 = 0.9422

2.0

R2 = 0.9603

1.5

R2

1.0

310 °C 330 °C

= 0.8673

350 °C

0.5

R2

370 °C

= 0.9119

1/ωR–1/ωR0

1n (ωR0/ωR)

2.5

0.06

390 °C R2 = 0.8752

0.04

0

20 40 60 Residence time (h)

410 °C

0.02

420 °C 430 °C

0

80

(a) 390 °C 400 °C

R2 = 0.6971

1.0

410 °C

0.5

420 °C 430 °C

0.0 2 4 6 Residence time (h)

1/ωR1.5–1/ωR01.5

1n (ωR0/ωR)

1.5

0

0.18 0.15 0.12 0.09 0.06 0.03 0.0

8

310 °C 330 °C

R2 = 0.9642 R2 = 0.874 R2 = 0.9365

(b)

20 40 60 Residence time (h)

350 °C 370 °C

80

R2 = 0.9195 310 °C

R2 = 0.9629

330 °C

R2 = 0.8718 R2 = 0.9356

20 40 60 Residence time (h)

(c)

350 °C 370 °C

80

1/ωR1.5–1/ωR01.5

(e)

0.2

0

8

R2 = 0.8463

0

0.3

0.1

2 4 6 Residence time (h)

(d)

2.0

0.0

400 °C

0.0

0.0

1/ωR–1/ωR0

352

0.025 0.020

390 °C

0.015

R2

0.010

400 °C

= 0.9307

410 °C 420 °C

0.005

430 °C

0.000 0

2 4 6 Residence time (h)

8

(f)

Figure 7.18 Linear regression analysis for first-order of reaction with the residue conversion data of (a) Adapted from Coronel-Garcia et al. (2021)/Elsevier, (b) Adapted from Cai et al. (2022)/Elsevier; second-order of reaction in residue conversion of (c) Source: Adapted from Coronel-Garcia et al. (2021)/Elsevier, (d) Adapted from Coronel-Garcia et al. (2021)/Elsevier; 2.5 order of reaction in residue conversion of (e, f) Adapted from Coronel-Garcia et al. (2021)/Elsevier.

A similar trend was found for the individual and global parameters ratios for VGO hydrocracking since for k1 /kR ratio, k5 /K VGO , the ratios were in the range of 0.4–1 (Figure 7.20), while k6 /kVGO were lower 0.4, and k7 /kVGO were in the range of 0–0.25. Demonstrating that the main selectivity of hydrocracking of VGO is toward D. Finally, k8 /kD and k9 /kD for distillates hydrocracking are presented in Figure 7.21, where k8 /kD values presented a higher value (range of 0.5–1) than k9 /kD (range of 0–0.5). This trend of the ratios for hydrocracking of D reveals that the main reaction is to produce N. 7.6.1.3

Activation Energies

The values of reaction rate coefficients at different temperatures were used to estimate the activation energies reported in Table 7.8. Activation energies ranging from 7 to 320 kJ/mol were obtained by Hassanzadeh and Abedi (2010) for the 10 reactions included in the hydrocracking kinetic model, where the lowest values are for the conversion of VGO → N and N → G, and the highest values are for the conversion of R → G and VGO → D. The activation energies (157–342 kJ/mol) reported by Galarraga et al. (2012) indicated that the minimum energy-requiring reactions are the conversion in series from R → VGO and VGO → D. The activation energies present a tendency because the higher values are for the generation of light products from heavier fractions, whereas the activation energies for the reaction rate coefficients k7 , k9 , and k10 (Table 7.8) were not possible

Table 7.6

Reaction rate coefficients and pre-exponential factor reported for the in situ and ex situ hydrocracking kinetic models.

Temperature (∘ C)

k1

k2

k3

k4

k5

k6

k7

k8

k9

k 10

Hassanzadeh and Abedi (2010) (Athabasca bitumen) 320

1.92 × 10−3

9.87 × 10−5

3.63 × 10−4

1.51 × 10−5

4.51 × 10−5

5.15 × 10−4

1.44 × 10−4

1.00 × 10−3

1.99 × 10−4

2.06 × 10−3

350

6.60 × 10−3

1.73 × 10−3

4.18 × 10−3

6.76 × 10−5

2.01 × 10−3

5.55 × 10−4

4.38 × 10−4

3.02 × 10−3

4.42 × 10−4

2.71 × 10−3

380

2.71 × 10−2

2.75 × 10−2

2.94 × 10−2

6.28 × 10−3

1.26 × 10−2

5.93 × 10−4

1.20 × 10−3

8.22 × 10−3

9.13 × 10−4

3.47 × 10−3

A0 , h−1

5.84 × 109

3.80 × 1022

2.39 × 1017

1.62 × 1023

3.50 × 1022

0.002 38

1.45 × 106

8.89 × 106

3.13 × 103

0.596

Galarraga et al. (2012) (Athabasca bitumen) 320

1.18 × 10−3

1.51 × 10−4

1.79 × 10−4

3.62 × 10−5

9.78 × 10−4

3.81 × 10−6

3.62 × 10−8

3.19 × 10−6

1.54 × 10−4

1.24 × 10−4

350

6.81 × 10−3

4.49 × 10−3

2.49 × 10−3

5.09 × 10−4

2.10 × 10−3

3.13 × 10−4

5.19 × 10−4

5.86 × 10−4

0

2.02 × 10−3

380

4.48 × 10−2

2.68 × 10−2

2.87 × 10−2

1.06 × 10−2

1.78 × 10−2

2.26 × 10−3

0

2.86 × 10−3

0

0

Loria et al. (2011) (Athabasca bitumen) 320

2.14 × 10−3

1.31 × 10−3

3.00 × 10−4

6.00 × 10−5

6.70 × 10−3

4.91 × 10−3

1.05 × 10−3

350

8.45 × 10−3

6.10 × 10−3

1.13 × 10−3

7.30 × 10−4

2.82 × 10−3

1.41 × 10−3

4.60 × 10−4

360

−2

9.51 × 10

−3

2.51 × 10

−3

−4

1.81 × 10

−3

7.30 × 10

−4

3.20 × 10−4

−2

1.13 × 10

−2

9.02 × 10

4.50 × 10

−4

1.40 × 10

−4

7.00 × 10−5

4.77 × 1018

2.38 × 109

1.32 × 10

−2

380

2.65 × 10

3.09 × 10

A0 , h−1

2.11 × 109

6.76 × 1011

2.00 × 1013

8.50 × 10

−3

7.16 × 1012

5.64 × 108

Da Silva De Andrade (2014) (Athabasca bitumen) 320

2.20 × 10−3

1.14 × 10−3

4.10 × 10−4

6.00 × 10−5

6.70 × 10−4

0

350

8.99 × 10−3

4.96 × 10−3

1.77 × 10−3

7.10 × 10−4

3.15 × 10−3

2.00 × 10−5

4.00 × 10−5

360

−2

1.34 × 10

7.61 × 10

−3

−3

−3

−3

−5

2.30 × 10−4

380

2.95 × 10−2

2.80 × 10−2

A0 , h−1

4.27 × 109

6.91 × 1011

3.26 × 10

6.00 × 10

0

1.80 × 10

4.93 × 10

1.10 × 10−2

9.37 × 10−3

1.61 × 10−2

3.80 × 10−4

1.02 × 10−2

8.33 × 1011

5.88 × 1019

4.21 × 1011

3.15 × 1022

4.441 × 1048

k 11

k 12

k 13

Table 7.6

(Continued)

Temperature (∘ C)

k1

k2

k3

k4

k5

k6

k7

k8

Da Silva De Andrade (2014) (Athabasca pitch) 320

6.10 × 10−4

1.60 × 10−4

6.00 × 10−5

0

3.00 × 10−5

0

350

1.70 × 10−3

5.90 × 10−4

2.90 × 10−4

1.00 × 10−5

1.40 × 10−4

1.80 × 10−4

380

9.15 × 10

−3

−3

−3

−4

−4

1.37 × 10−2

A0 , h−1

2.84 × 109

1.81 × 1010

1.43 × 1037

4.00 × 10

2.56 × 1011

1.58 × 10

3.66 × 1011

1.70 × 10

1.78 × 1018

6.30 × 10

Da Silva De Andrade (2014) (Athabasca vacuum residue) 320

1.26 × 10−3

6.00 × 10−4

1.20 × 10−4

1.00 × 10−5

4.20 × 10−4

0

380

4.41 × 10−3

1.24 × 10−3

3.70 × 10−4

9.00 × 10−5

1.45 × 10−3

9.20 × 10−4

395

7.92 × 10

−3

−3

−4

−4

−3

3.92 × 10−3

A0 , h−1

6.83 × 103

1.66 × 10 3.745

4.60 × 10 16.145

2.10 × 10

5.85 × 107

2.17 × 10

7.05 × 102

9.15 × 1025

Orozco Castillo (2016) (Mexican heavy oil) 320

3.38 × 10−3

1.22 × 10−6

4.47 × 10−6

1.67 × 10−4

4.21 × 10−4

1.07 × 10−4

340

3.17 × 10−3

1.90 × 10−5

6.13 × 10−5

2.76 × 10−4

1.28 × 10−3

1.56 × 10−4

360

6.34 × 10−3

1.63 × 10−4

7.98 × 10−4

4.60 × 10−4

3.99 × 10−3

2.26 × 10−4

A0 , h−1

5.46 × 1001

5.64 × 1027

1.86 × 1030

1.48 × 103

1.18 × 1012

1.49 × 1001

Orozco Castillo (2016) (Vacuum residue) 320

9.90 × 10−4

2.00 × 10−4

2.00 × 10−5

6.00 × 10−5

4.40 × 10−4

6.10 × 10−4

340

2.37 × 10−3

5.50 × 10−4

1.30 × 10−4

1.20 × 10−4

1.05 × 10−3

1.07 × 10−3

360

5.09 × 10−3

1.44 × 10−3

7.40 × 10−4

4.10 × 10−4

2.41 × 10−3

1.83 × 10−3

A0 , h−1

1.97 × 108

7.22 × 109

4.72 × 1019

4.85 × 108

1.79 × 108

1.99 × 104

k9

k 10

k 11

k 12

k 13

Rodriguez-DeVecchis et al. (2017) (Athabasca bitumen) 340

5.12 × 10−3

4.24 × 10−3

4.00 × 10−5

1.23 × 10−3

2.28 × 10−3

5.50 × 10−4

3.51 × 10−3

350

−3

9.38 × 10

7.44 × 10

−3

−5

1.69 × 10−3

4.04 × 10−3

1.54 × 10−3

5.43 × 10−3

360

1.69 × 10−2

1.28 × 10−2

8.00 × 10−5

2.29 × 10−3

7.02 × 10−3

4.16 × 10−3

8.27 × 10−3

370

2.97 × 10−2

2.18 × 10−2

1.10 × 10−4

3.08 × 10−3

1.20 × 10−2

1.09 × 10−2

1.25 × 10−2

A0 , h−1

3.09 × 1014

1.51 × 1014

6.93 × 1012

3.37 × 106

1.03 × 1015

3.81 × 1018

3.07 × 1013

6.00 × 10

Elahi et al. (2019) (Vacuum residue) 320

1.00 × 10−3

2.60 × 10−4

7.50 × 10−5

2.90 × 10−4

1.80 × 10−4

1.00 × 10−6

4.00 × 10−4

3.80 × 10−4

340

2.90 × 10−3

8.00 × 10−4

3.90 × 10−4

5.00 × 10−4

5.40 × 10−4

2.30 × 10−6

8.00 × 10−4

1.00 × 10−3

360

5.90 × 10−3

2.10 × 10−3

9.70 × 10−4

8.10 × 10−4

1.50 × 10−3

4.90 × 10−6

1.40 × 10−3

2.90 × 10−3

A0 , h−1

1.7 × 109

8.8 × 1010

3.6 × 1013

2.6 × 103

5.8 × 1010

9.5 × 104

1.3 × 105

4.3 × 1010

Duran Armas (2021) (Vacuum residue) 335

2.81 × 10−3

7.30 × 10−4

3.40 × 10−4

2.40 × 10−4

5.80 × 10−4

2.90 × 10−4

8.20 × 10−4

3.00 × 10−4

1.10 × 10−3

350

3.60 × 10−3

1.12 × 10−3

5.80 × 10−4

3.70 × 10−4

8.60 × 10−4

4.70 × 10−4

1.21 × 10−3

1.00 × 10−3

1.53 × 10−3

365

6.25 × 10−3

3.18 × 10−3

9.70 × 10−4

1.10 × 10−3

1.29 × 10−3

7.40 × 10−4

2.23 × 10−3

1.59 × 10−3

2.79 × 10−3

A0 , h−1

6.55 × 104

4.78 × 1010

2.24 × 106

2.37 × 1010

1.58 × 104

8.58 × 104

1.30 × 106

4.08 × 109

3.92 × 105

Nguyen et al. (2013) 420

7.00 × 10−2

1.50 × 10−1

5.60 × 10−2

0

1.70 × 10−1

430

1.30 × 10−1

2.30 × 10−1

7.00 × 10−2

1.30 × 10−2

2.10 × 10−1

405

1.28 × 10−3

8.66 × 10−4

5.82 × 10−4

1.05 × 10−4

2.30 × 10−6

6.07 × 10−4

3.01 × 10−4

1.93 × 10−4

1.24 × 10−4

1.82 × 10−3

420

3.87 × 10−3

2.22 × 10−3

1.49 × 10−3

2.34 × 10−4

7.70 × 10−6

1.93 × 10−3

8.92 × 10−4

4.95 × 10−4

3.92 × 10−4

4.63 × 10−3

435

1.11 × 10−2

5.48 × 10−3

3.67 × 10−3

5.05 × 10−4

2.49 × 10−5

5.82 × 10−3

2.52 × 10−3

1.22 × 10−3

1.18 × 10−3

1.13 × 10−2

A0 , min−1

1.73 × 1019

7.11 × 1015

4.31 × 1015

1.45 × 1012

9.19 × 1018

9.14 × 1019

1.89 × 1018

1.59 × 1015

1.59 × 1019

1.10 × 1016

Asaee et al. (2014)

(Continued)

Table 7.6

(Continued)

Temperature (∘ C)

k1

k2

k3

k4

k5

k6

k7

k8

k9

k 10

Kim et al. (2017) 380

2.40 × 10−3

3.33 × 10−2

3.00 × 10−4

7.60 × 10−3

1.00 × 10−4

1.00 × 10−4

1.10 × 10−3

400

9.60 × 10−3

1.56 × 10−1

1.00 × 10−4

5.67 × 10−2

1.00 × 10−4

8.00 × 10−4

2.20 × 10−3

T. Huang et al. (2017) 405

5.47 × 10−2

2.71 × 10−2

6.30 × 10−3

2.15 × 10−2

4.58 × 10−2

8.70 × 10−3

3.08 × 10−2

0

420

9.59 × 10−2

3.72 × 10−2

1.48 × 10−2

2.70 × 10−2

9.75 × 10−2

1.20 × 10−2

4.84 × 10−2

1.20 × 10−2

435

−1

−2

−2

−2

−1

−2

−2

3.00 × 10−2

1.32 × 10

4.70 × 10

2.83 × 10

3.57 × 10

1.85 × 10

1.63 × 10

8.07 × 10

Ortega-Garcia et al. (2017) 350

2.45 × 10−4

6.10 × 10−3

2.55 × 10−4

1.04 × 10−5

1.68 × 10−3

3.84 × 10−3

2.94 × 10−4

2.14 × 10−3

1.52 × 10−2

3.32 × 10−3

370

2.89 × 10−4

6.80 × 10−3

3.45 × 10−4

1.47 × 10−5

1.76 × 10−3

4.29 × 10−3

3.51 × 10−4

2.40 × 10−3

1.79 × 10−2

3.69 × 10−3

Félix and Ancheyta (2019a) 360

3.20 × 10−2

0

0

2.19 × 10−3

3.39 × 10−3

0

380

3.44 × 10−2

6.50 × 10−2

4.35 × 10−2

2.30 × 10−3

3.55 × 10−3

0

−2

−2

−2

−3

−2

1.09 × 10−2

390

3.61 × 10

9.51 × 10

400

7.77 × 10−2

1.29 × 10−1

1.18 × 10−1

6.50 × 10−2

1.17 × 10−2

2.71 × 10−2

A0 , h−1

6.01 × 103

6.92 × 108

1.70 × 1013

6.42 × 1018

4.08 × 107

5.29 × 1024

6.51 × 10

5.43 × 10

1.12 × 10

Félix and Ancheyta (2019b) (4 lumps, 8 constants) 360

0

9.50 × 10−2

2.09 × 10−2

1.67 × 10−2

1.94 × 10−2

7.78 × 10−2

0

380

1.19 × 10−1

9.50 × 10−2

2.09 × 10−2

1.67 × 10−2

5.00 × 10−2

7.78 × 10−2

0

390

3.59 × 10−1

9.50 × 10−2

2.09 × 10−2

8.57 × 10−2

5.00 × 10−2

7.78 × 10−2

0

400

3.59 × 10−1

2.36 × 10−1

5.21 × 10−2

1.59 × 100

5.00 × 10−2

7.78 × 10−2

1.17 × 100

A0 , h−1

2.29 × 1015

1.33 × 104

3.01 × 103

1.49 × 1028

3.49 × 105

k 11

k 12

k 13

Félix and Ancheyta (2019b) (5 lumps, 9 constants) 360

9.31 × 10−2 2.06 × 10−2 1.74 × 10−2

0 −1

2.48 × 10−3

1.00 × 10

4.96 × 10−2 8.39 × 10−2 6.68 × 10−3 3.64 × 10−2

390

2.65 × 10−1 9.31 × 10−2 2.06 × 10−2 7.40 × 10−2

4.96 × 10−2 8.39 × 10−2 6.68 × 10−3 1.22 × 10−1

400

2.65 × 10−1 1.31 × 10−1 3.29 × 10−2 2.74 × 100

4.96 × 10−2 8.39 × 10−2 2.03 × 100

2.57 × 10−1

3.08 × 105

1.17 × 1032

A0 , h−1 2.06 × 1013

7.59 × 100

−2

1.93 × 10−2 7.76 × 10−2 0

380

−2

9.31 × 10

−2

2.06 × 10

9.01 × 100

1.74 × 10

3.64 × 1030

3.07 × 10−1 3.52 × 1080

Félix and Ancheyta (2019b) (5 lumps, 12 constants) 360

0

9.56 × 10−2 1.89 × 10−2 0

1.37 × 10−2

7.52 × 10−3 1.97 × 10−2

0

7.51 × 10−2 0

380

1.21 × 10−1 9.56 × 10−2 1.89 × 10−2 0

1.37 × 10−2

7.52 × 10−3 5.29 × 10−2

0

7.51 × 10−2 0

390

3.37 × 10−1 9.56 × 10−2 1.89 × 10−2 2.37 × 10−2 7.87 × 10−2

7.52 × 10−3 5.29 × 10−2

0

7.51 × 10−2 0

400

3.37 × 10−1 2.14 × 10−1 7.89 × 10−2 2.37 × 10−2 1.44 × 100

7.52 × 10−3 5.29 × 10−2

3.19 × 10−2 7.51 × 10−2 1.06 × 100

A0 , h−1 1.72 × 1014

3.35 × 103

2.21 × 106

8.90 × 1028

7.24 × 105

Félix and Ancheyta (2019b) (6 lumps, 13 constants) 360

0

9.38 × 10−2 1.86 × 10−2

1.43 × 10−2

7.53 × 10−3 1.97 × 10−2

0

380

9.65 × 10−2 9.38 × 10−2 1.86 × 10−2

1.43 × 10−2

7.53 × 10−3 5.55 × 10−2

1.49 × 10−2 7.77 × 10−2 3.61 × 10−3 0.035 86

390

2.59 × 10−1 9.38 × 10−2 1.86 × 10−2

6.72 × 10−2

1.39 × 10−2 5.55 × 10−2

1.99 × 10−2 7.77 × 10−2 3.61 × 10−3 0.1195

400

2.59 × 10−1 1.14 × 10−1 5.68 × 10−2

1.59 × 100

1.39 × 10−2 5.55 × 10−2

3.46 × 10−2 7.77 × 10−2 1.17 × 100

5.65 × 1028

8.68 × 102

2.86 × 1010

A0 , h−1 3.45 × 1013

1.22

3.78 × 104

1.68 × 106

7.50 × 10−2 0

1.41 × 10−1 1.81 × 1081

0.002 48

0.249 24 7.00 × 1031 (Continued)

Table 7.6

(Continued)

Temperature (∘ C)

k1

k2

k3

k4

k5

k6

k7

k8

k9

k 10

Pham et al. (2021) 410

2.57 × 10−3

1.30 × 10−3

3.00 × 10−4

1.30 × 10−4

4.81 × 10−2

9.83 × 10−2

2.28 × 10−1

425

5.57 × 10−3

1.89 × 10−3

8.20 × 10−4

5.40 × 10−4

1.42 × 10−1

9.83 × 10−2

2.28 × 10−1

440

1.72 × 10−2

6.26 × 10−3

3.64 × 10−3

2.07 × 10−3

2.62 × 10−1

9.83 × 10−2

2.28 × 10−1

450

4.27 × 10−2

3.02 × 10−2

1.05 × 10−2

6.69 × 10−3

2.62 × 10−1

1.43 × 10−1

2.67 × 10−1

7.10 × 1020

3.20 × 1024

2.70 × 1026

2.80 × 1012

1.73 × 103

2.17 × 100

A0 , h−1

Coronel-García et al. (2021) 310

2.23 × 10−6

2.22 × 10−6

1.62 × 10−7

5.55 × 10−7

1.07 × 10−4

9.34 × 10−6

1.76 × 10−7

1.99 × 10−4

1.39 × 10−4

7.62 × 10−5

330

3.07 × 10−6

2.48 × 10−6

9.03 × 10−7

4.24 × 10−7

1.66 × 10−4

0

3.36 × 10−8

2.53 × 10−4

2.28 × 10−4

9.51 × 10−5

350

0

3.03 × 10−5

7.52 × 10−5

0

1.17 × 10−3

6.43 × 10−4

7.80 × 10−4

2.92 × 10−4

0

1.43 × 10−3

370

0

3.94 × 10−4

6.20 × 10−4

9.36 × 10−5

3.23 × 10−3

1.79 × 10−2

3.37 × 10−3

1.89 × 10−3

8.18 × 10−4

0

A0 , h−1

3.56

43.81

77.48

43.81

29.66

60.03

108.85

12.94

10.28

35.16

Cai et al. (2022) 430

0.1971

0.1047

0.0708

1.43 × 10−15

0.0149

0.0345

1.0927

Alvarez et al. (2019) A0 , h−1

8.00 × 1018

8.22 × 1026

3.73 × 108

1.10 × 108

6.69 × 107

k 11

k 12

k 13

7.6 Results and Discussion

Table 7.7

Kinetic parameter for the continuous lumping kinetic model.

Parameter

Value

Kinetic model parameters kmax (h−1 )

0.7462

𝛂

0.7186

𝛅

7.7672 × 10−6

a0

3.0784

a1

26.1330

Parameters for liquid, gas, and solid fractions kL (h−1 )

26.5992

K L (wt%−1 )

66.4266

−1

K G (wt% )

76.6965

K S (wt%−1 )

151.5629

−1

56.5676 0.569 007

PG

0.4443

PS

0.5557

1.0

1.0

0.8

0.8

0.6

0.6

k2/kR

k1/kR

LL (wt% ) 𝛄

0.4 0.2 0.0 300

0.4 0.2

330

360

390

420

0.0 300

450

330

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 300

360

390

420

450

420

450

Temperature (°C) 0.20 0.15 k4/kR

k3/kR

Temperature (°C)

0.10 0.05

330

360

390

Temperature (°C)

420

450

0.00 300

330

360

390

Temperature (°C)

Figure 7.19 Ratios of individual kinetic parameters (k 1 , k 2 , k 3 , and k 4 ) with the global kinetic parameter for R hydrocracking (k R ) with reaction temperature for the kinetic models of: ( ) Hassanzadeh and Abedi (2010), ( ) Galarraga et al. (2012), ( ) Loria et al. (2011), ( ) Da Silva De Andrade (2014) for bitumen, ( ) Da Silva De Andrade (2014) for pitch, ( ) Da Silva De Andrade (2014) for VR, ( ) Orozco Castillo (2016) for heavy oil, ( ) Orozco Castillo (2016) for VR, ( ) Rodriguez-DeVecchis et al. (2017), ( ) Elahi et al. (2019), ( ) Duran Armas (2021), ( ) Asaee et al. (2014), ( ) T. Huang et al. (2017), ( ) Ortega-Garcia et al. (2017), ( ) Pham et al. (2021), and ( ) Coronel-García et al. (2021).

359

7 Catalytic Mechanism and Kinetics

1.0

k5/kVGO

0.8 0.6 0.4 0.2 0.0 300

330

360 390 Temperature (°C)

420

450

330

360 390 Temperature (°C)

420

450

330

360 390 Temperature (°C)

420

450

0.8

k6/kVGO

0.6 0.4 0.2 0.0 300 0.54 0.45 k7/kVGO

360

0.36 0.27 0.18 0.09 0.00 300

Figure 7.20 Ratios of individual kinetic parameters (k 5 , k 6 , and k 7 ) with the global kinetic parameter for VGO hydrocracking (k VGO ) with reaction temperature for the kinetic models of: ( ) Hassanzadeh and Abedi (2010), ( ) Galarraga et al. (2012), ( ) Loria et al. (2011), ( ) Da Silva De Andrade (2014) for bitumen, ( ) Da Silva De Andrade (2014) for pitch, ( ) Da Silva De Andrade (2014) for VR, ( ) Orozco Castillo (2016) for heavy oil, ( ) Orozco Castillo (2016) for VR, ( ) Rodriguez-DeVecchis et al. (2017), ( ) Elahi et al. (2019), ( ) Duran Armas (2021), ( ) Asaee et al. (2014), ( ) T. Huang et al. (2017), ( ) Ortega-Garcia et al. (2017), ( ) Pham et al. (2021), and ( ) Coronel-García et al. (2021).

7.6 Results and Discussion

1.00

k8/kD

0.80 0.60 0.40 0.20 0.00 300

330

360 390 Temperature (°C)

420

450

330

360 390 Temperature (°C)

420

450

1.00

k9/kD

0.80 0.60 0.40 0.20 0.00 300

Figure 7.21 Ratios of individual kinetic parameters (k 8 and k 9 ) with the global kinetic parameter for D hydrocracking (k D ) with reaction temperature for the kinetic models of: ( ) Hassanzadeh and Abedi (2010), ( ) Galarraga et al. (2012), ( ) Loria et al. (2011), ( ) Da Silva De Andrade (2014) for bitumen, ( ) Da Silva De Andrade (2014) for pitch, ( ) Da Silva De Andrade (2014) for VR, ( ) Orozco Castillo (2016) for heavy oil, ( ) Orozco Castillo (2016) for VR, ( ) Rodriguez-DeVecchis et al. (2017), ( ) Elahi et al. (2019), ( ) Duran Armas (2021), ( ) Asaee et al. (2014), ( ) T. Huang et al. (2017), ( ) Ortega-Garcia et al. (2017), ( ) Pham et al. (2021), and ( ) Coronel-García et al. (2021).

to be estimated since the reaction rate coefficients presented values of zero. Confirming that N and D are not capable of reacting at the evaluated reaction conditions. The activation energy values (136–261 kJ/mol) obtained by Loria et al. (2011) were related to those observed in aforementioned experiments performed at laboratory bench-scale (Galarraga et al. 2012), where reactions in a series are carried out but not the gas production reactions (k7 , k9 , and k10 ), implying that G fraction is exclusively produced from the R fraction. Da Silva De Andrade (2014) reported higher activation energies for the upgrading of Athabasca bitumen (139–633 kJ/mol), the Athabasca pitch (144–487 kJ/mol), and the Athabasca vacuum residue (43–362 kJ/mol). The reactions in the series of R → VGO and VGO → D prevail in the three kinetic models given, and the G production is only from the R fraction, with the N generation from D being the most energy demanding reaction. The activation energies presented lower values during the reaction in the series R → VGO → D → N in both heavy oil (maximum value of 404.65 kJ/mol) and vacuum residue (maximum value of 276.5 kJ/mol) upgrading, reported by Orozco Castillo (2016). Furthermore, as well as Loria et al. (2011) and Da Silva De Andrade (2014), the production of G is only from R. The kinetic model reported by Rodriguez-DeVecchis et al. (2017) gives low values of activation energies (105.7–255.3 kJ/mol), where the lowest activation

361

Table 7.8

Activation energies (kJ/mol) of the kinetic models reviewed for the in situ and ex situ upgrading.

References

k1

k2

k3

k4

k5

Hassanzadeh and Abedi (2010)

141.91

301.82

236.02

320.67

303.83

Galarraga et al. (2012)

172.1

276.7

271.7

303.1

157.0

342.9

242.0

Loria et al. (2011)

136

167

192

261

145

190

146

Da Silva De Andrade (2014) (Bitumen)

139.5

168.3

174.3

272.9

168.3

324.0

633.3

Da Silva De Andrade (2014) (Pitch)

144.4

173.3

179.8

276.0

168.3

487.9

Da Silva De Andrade (2014) (VR)

76.6

43.2

58.1

146.6

70.8

362.7

Orozco Castillo (2016) (Heavy oil)

48.56

382.16

404.65

79.02

k6

7.54

k7

k8

112.88

112.88

175.52

58.38 85.8

k9

81.67

k 10

Orozco Castillo (2016) (VR)

128.1

153.8

276.5

147.1

131.8

Rodriguez-DeVecchis et al. (2017)

192.2

189.7

184.8

105.7

207.8

255.3

Elahi et al. (2019)

138.7

164.9

200.3

78.8

164.7

124.7

96.4

159.7

Duran Armas (2021)

85.9

160.8

114.4

163.5

86.6

98.6

107.3

151.2

99.8

Nguyen et al. (2013)

194.5

165.5

88.5

Asaee et al. (2014)

287.29

245.55

244.97

209.54

319.5

300.9

282.99

245.56

300.0

T. Huang et al. (2017)

117.14

73.44

199.04

67.20

185.78

84.01

128.17

249.39

Ortega García et al. (2017)

27.65

18.08

50.42

58.42

8.65

18.63

64.68

125.31

182.67

263.38

123.09

338.82

Félix and Ancheyta (2019b) (8 coefficients)

202.63

63.35

63.47

367.36

Félix and Ancheyta (2019b) (9 coefficients)

178.11

23.51

32.51

396.64

Félix and Ancheyta (2019b) (12 coefficients)

188.57

55.90

99.29

Félix and Ancheyta (2019b) (13 coefficients)

181.13

Coronel-García et al. (2021)

46.69

k 12

k 13

3.26

1051.86

417.02

188.9

77.5

Félix and Ancheyta (2019a)

Pham et al. (2021)

k 11

27.93

39.16

86.52

13.68

77.61

312.49

367.48

396.71

178.99 188.99

278.40

453.84

286.73

368

115

126

Browning et al. (2019)

289

334

306

76.3

17.58

7.20 1039.55

94 147

419.74

90.88

375.35

261

27.2

87.15 377.86

Álvarez et al. (2019)

19

243.82

61.84 380.53

613.79

95.27

153.85

57.10

12.96

105.77

93.39

219.20

7.6 Results and Discussion

energy is for the cracking of R and the production of G is in agreement with results of previous reaction mechanisms, being generating from R. Elahi et al. (2019) obtained activation energies (78.8–200.3 kJ/mol) with similar results to almost all the abovementioned kinetic models, where the G fraction is only produced from the R fraction, but in this case, G is also produced from N since the reactions of VGO and D to form G are not achieved at the operating conditions considered. The reaction mechanism reported by Duran Armas 2021 neglected only one reaction (D → G) and the activation energies (85.9–163.5 kJ/mol) show that the easier reactions are in the series (R → VGO → D, and N → G) and the production of N from VGO. The reported activation energies (77.5–194.5 kJ/mol) of Nguyen et al. (2013) show that the main reaction carried out at the studied conditions are in the series (VGO → D → N), since these reactions presented the lowest activation energies, while the reaction of N → G is not contemplated in the kinetic model. Asaee et al. (2014) reported activation energies in the range of 209.5–319.5 kJ/mol, where lower values were for the production of light lumps (AGO, N, and G), and the higher values were for the generation of C from both R and VGO in the kinetic scheme. The activation energy values (67.2–249.4 kJ/mol) estimated by S. Huang et al. (2017) indicate that the production of G from heavy lumps (R and VGO) is carried out easily, whereas the reactions VGO → N and N → G are not performed in the reaction network. The study of Ortega-Garcia et al. (2017) reported low values of activation energies (8.6–58.4 kJ/mol) compared with the abovementioned values, denoting that the lower energy-demand reactions in the reaction network are in the series (HGO → LGO → N → G), and the higher values are for the G and N production from R. The reactions VGO to lighter lumps (N and G) and D to G were not achieved at the conditions studied by Pham et al. (2021), and the activation energies reported (12.96–396.71 kJ/mol) indicate that the easier reactions to be performed are in the series (VGO → D → N → G). Another set of activation energies (46.7–613.8 kJ/mol) reported by Coronel-Garcia et al. (2021) displayed a similar pattern than those of Orozco Castillo (2016), since the reactions in the series (R → VGO → D → N) present a low value together with the parallel reaction from D → G, and the highest value was for the VGO → G reaction. The activation energies values (64.6–338.8 kJ/mol) obtained for the four-lump kinetic model reported by Félix and Ancheyta (2019a) based on distillation boiling points suggest that the conversion of R and VGO (heavy fractions) were the easier reactions to be carried out, while the conversion of D demands high activation energy. However, the four-lump kinetic model reported by Félix and Ancheyta (2019b) but based on SARA fractions shows that the production of lighter fractions (Sa and Ar) from As is the easiest reaction performed since they have the lowest values of the activation energies, the higher values of activation energies are for the reaction in the series (As → Re → Ar → Sa), and the production of Sa from Re is neglected at the conditions considered. The condensation reactions (Ar → Re, and Re → As) are achieved only at temperatures above 400 ∘ C (severe conditions). Similar tendencies of the activation energies (7.2–1039.5 kJ/mol) are obtained for the two five-lump kinetic models (based on SARA fractions) described by Félix and Ancheyta (2019b) since the generation of the heavy fractions has the highest values of activation energies while the lowest values are for the production of lighter components, while some reactions (Re → Sa, and Ar → G) are not performed. The six-lump kinetic model of Félix and Ancheyta (2019b) reported that the activation energies (3.2–1051.8 kJ/mol) displayed the lowest values for the hydrocracking reactions of heavy fractions (As and Re) and the parallel reactions from As → G, Re → Sa and Ar → G are not performed. The activation energies (94–369 kJ/mol) obtained for the detailed kinetic model of Alvarez et al. (2019) suggest that the reactions involving the conversion of VGO and D (middle fractions) along with the HDS reactions are carried out easier than the hydrocracking of R. Browning et al. (2019)

363

364

7 Catalytic Mechanism and Kinetics

reported that the activation energies (76.3–334 kJ/mol) showing that the reaction in the series (R → VGO → D) and HDS reaction needs lower energy to be performed than the hydroconversion of R to VGO and D by parallel reactions. The collision or preexponential factors for all the hydrocracking kinetic models are presented in Table 7.6, presenting different values for each kinetic model and fraction. This kinetic parameter provides an estimate of the impacts between pseudocomponents to carry on the reaction subject to several factors, such as temperature, catalyst, type of molecules, whereby there are different values for each kinetic model since they contemplate different pseudocomponents and hydrocracking conditions. 7.6.1.4

Selectivity of Hydrocracking Reactions

Almost all kinetic models based on the lumping approach suggest that the G fraction is only produced from R or other heavy fractions. This could be because the operating condition (reaction temperature mainly) in almost all the studies is moderate, thus hindering the cracking of the middle and low molecular weight fractions. Furthermore, the activation energies in all reaction schemes indicate that the reaction pathways carried out are mainly in the series (R → VGO → D → N) along with R conversion to light products. This behavior is attributed to the low temperature used, where the over-cracking is minimum and light product hydrocracking reactions are hindered to occur like some works (Félix et al. 2019; Félix and Ancheyta 2019a) reported. The most energy demanding reactions in the hydrocracking reaction networks reported are for N production from middle and heavy fractions and the forming of C from heavy fractions (R or VGO). This is due to the operating conditions, particularly reaction temperature, that need to be severe enough to cleavage the C—C bonds and produce free radicals faster than hydrogenation reaction to condensate the heavy molecules to form C. The reaction mechanisms for the kinetic models based on SARA fractions indicate that the reactions Re → Sa, Ar → G, and As → G are not achieved at the conditions used, leaving the formation of G fraction mainly from Re fraction. The condensation reactions (Ar → Re, Re → As, and As → C) need high temperature as C fraction formation since these reactions need C—C bond breakage faster than hydrogenation reactions. The different values of the activation energies reported by the kinetic models using similar reaction pathways can be attributed to the upgrading conditions comprising different reactor configurations relating to hydrogen-liquid mass transference (continuous and noncontinuous reactors), the complexity of feedstock (heavy oil and vacuum residue fractions), catalyst type (liquid or ultradispersed), and in a minor degree the boiling point differences of the lumps. While for in situ applications, the operating conditions are less severe (particularly temperature), for ex situ applications generally used severe reaction conditions because the aim is maximizing the conversion of the residue fraction into lighter hydrocarbons; meanwhile, for in situ propose, the focus is on partial upgrading of heavy and extra-heavy oil. However, the higher temperature used in ex situ upgrading enhances the generation of nondesired products, like coke. In order to confirm the diverse activation energies obtained by kinetic models with similar reaction mechanism, the works of Da Silva De Andrade (2014) and Orozco Castillo (2016) at the same operating conditions but with different feedstocks are compared. Different feeds are used such as vacuum residue obtaining activation energies lower than those estimated using a heavy oil or bitumen as feedstock as can be seen in Figure 7.22, indicating that the heavier fractions need less energy to produce lighter compounds. In addition, the comparison of the similar operating conditions of Galarraga et al. (2012) and Loria et al. (2011) works using different types of reactors, continuous reactor shown lower values of activation energies than batch reactor. Among all the kinetic models reviewed, the lowest values are reported by Ortega-Garcia et al. (2017) using similar operating conditions, feedstock, and reactor types than other studies. Although, only two

7.6 Results and Discussion

14

° API

10 6 2 –2 –6 0

50

100

150

200

250

300

Ea of residue conversion (kJ/mol) Figure 7.22 Activation energies of global residue conversion for different feedstocks (API gravity) using ( ) batch, ( ) CSTR, and ( ) PFR reactors.

temperatures were used in this study to calculate the activation energies, the principal difference compared with other studies seems to be the use of liquid catalysts. Similar mechanisms using different types of catalysts provoke changes in reaction pathways since they can favor some reactions more than others, thus changing the selectivity toward some products and increasing or decreasing reaction rates, obtaining diverse values of reaction rate coefficients and consequently different activation energies. This effect can be noticed even when the selectivity toward some products is drastically changed to the point that it seems that some reaction pathways do not take place because the reaction rate obtained is negligible. One of the causes may be due to the type of catalyst, i.e. liquid or ultradispersed, since this characteristic can favor the diffusion within the heavy crude oil in a different manner, provoking different conversion levels of heavy fractions into lighter products.

7.6.2 7.6.2.1

Aquathermolysis Reaction Order for Aquathermolysis Reaction

The reaction order is based on the number of molecules that interact in a reaction and defines the manner in which the reaction rate varies concerning the concentrations of reactants and products. The estimation or assumptions of these kinetic parameters must be according to the behavior of the reaction system and not supported solely by mathematical adjustment. According to the results of most kinetic models for aquathermolysis reaction, the reaction rate equations used to predict the production of reaction gas as well as the distribution of liquid phase fractions are adequately fitted by assuming a first-order reaction, considering the assumption that the water concentration is in excess and its yield is grouped with the reaction rate coefficients for each reaction. Nonetheless, the studies that justify these assumptions are few. The results of tetrahydrothiophene experiments reported by Lin et al. (2020) showed that during thermal cracking reactions, the H2 S concentration detected was 344mg/m3 , while for aquathermolysis reaction was 3020 mg/m3 . Therefore, it was considered that for aquathermolysis conditions, H2 S is not generated by thermal cracking reactions. Additionally, it was observed that during aquathermolysis experiments, the H2 S concentration increases concerning reaction time and temperature. As described in Section 7.5.2, the integrated batch reactor mass balance equation (Eq. (7.153)) was utilized to determine the reaction order for H2 S production. Figure 7.23a shows

365

Determination coefficient, R2

7 Catalytic Mechanism and Kinetics

1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 0.84 0.82 0.80 0

1n y0 /y

366

0.2

0.4

0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

0.6

0.8 1 1.2 1.4 Reaction order, n (a)

1.6

1.8

2

y = 0.0052x – 0.0139 R2 = 0.9408

y = 0.0022x – 0.0079 R2 = 0.8879

0

10

20

30 40 50 Reaction time (h) (b)

60

70

80

Figure 7.23 Estimation of reaction order for aquathermolysis reaction of (a) tetrahydrothiophene and (b) asphaltenes (Félix et al. 2022a): 220 ∘ C ( ), 250 ∘ C ( ), 260 ∘ C (Δ), 300 ∘ C ( ). Source: (a) Adapted from Lin et al. (2020). (b) Adapted from Tirado et al. (2022b).

the fit of experimental data for different assumed reaction order values, where it was indicated that a value equal to 0.9 presented the best fit concerning the data set of H2 S generation. Otherwise, Hamedi-Shokrlu and Babadagli (2014) reported that the production of H2 S during aquathermolysis reaction of oil sand in the presence of Ni nanoparticles is adjusted to first-order equation using a similar procedure. Zhang et al. (2020) obtained similar results for noncatalyzed and core-catalyzed aquathermolysis of Liaohe Oil field heavy oil. In addition, it was indicated that certain reaction gases such as CO2 , CH4 , and CO presented the best fit with reaction order values of 0.8, 1.2, and 0.8, respectively. While CO and C2+ agreed, the best fit with the experiment data when the reaction order was 1. The results of Félix et al. (2022a) show that the global asphaltenes conversion from Ashal’cha Oil field heavy crude oil fits with a first-order reaction. Figure 7.23b shows the plot of ln(yAsp0 /yAsp ) versus reaction time, where the tendencies for each temperature show a good agreement with experimental data, following a straight line. 7.6.2.2

Reaction Rate Coefficients and Activation Energies

The different approaches and methods used to estimate the kinetic parameters of the models have presented adequate results predicting the behavior of the products and reagents during the

7.6 Results and Discussion

aquathermolysis reaction. These results are analyzed to discuss the diverse reaction pathways performed using diverse feeds and operating conditions. The reaction rate coefficients for the tetrahydrothiophene aquathermolysis reaction (Lin et al. 2020) were estimated considering a reaction order value of 0.9. Figure 7.23a shows the determination of the slopes of fitting curves, which correspond to the rate coefficients values for H2 S generation. Using these results, the activation energy and frequency factor were estimated, obtaining values of 54.55 kJ/mol and 166.47 day−1 , respectively. These activation energy value is approximate to the results obtained in the model developed by Hamedi-Shokrlu and Babadagli (2014), which indicated that the catalytic effect of Ni nanoparticle reduces the activation energy for the H2 S generation from 69 to 38 kJ/mol and the frequency factor from 870 to 1.8 day−1 . However, the nickel nanoparticle presented the maximum catalytic effect at 270 ∘ C since, at higher temperatures, thermal cracking reactions show a predominant role over aquathermolysis reaction (1982a). The adjustment of the kinetic parameters of both the models with experimental data presented values of determination coefficients (r 2 ) is higher than 0.98. The calculated kinetic parameters by Perez-Perez et al. (2011) for H2 S and CO2 are shown in Table 7.3. These results presented a relative error lower than 5% concerning experimental data. Table 7.9 compares the values of kinetic parameters obtained for aquathermolysis kinetic studies that are focused on predicting the H2 S and CO2 generation. It is observed that the activation energies of H2 S generation reported by Lin et al. (2020), Ibatullin et al. (2011), Jia et al. (2016), and Hamedi-Shokrlu and Babadagli (2014) in the absence of catalyst present similar values in the range of 54−69 kJ/mol, while the frequency factor differs significantly by several orders of magnitude. The latter is attributed to the fact that the frequency factor depends on the frequency of intermolecular collisions to cause chemical reactions. On the other hand, the frequency factor values for CO2 generation show a higher approximation even to those reported by Perez-Perez et al. (2011). However, the activation energies values present a higher difference since the CO2 production depends on minerals concentration in the feed. The results of the kinetic model reported by Belgrave et al. (1997) showed that asphaltenes conversion to LO is negligible (a13 = 0), and the thermal cracking of LO does not occur during experiments with the sample from Athabasca cores (sample 2) since it presented a preexponential factor for reaction 4 that equals to 0. The kinetic model presented a good fit with the three Table 7.9

Comparison of kinetic parameters for aquathermolysis.

Kinetic parameters

A0 (day−1 )

E A (kJ/mol)

Lin et al. (2020) H2 S

166.47

54.55

Ibatullin et al. (2011) H2 S

3.395 × 106

63.5

CO2

3.900 × 106

64

Jia et al. (2016) H2 S CO2

9.160 1.667 × 10

58.54 6

11.29

Hamedi-Shokrlu and Babadagli (2014) H2 S

870

69

367

368

7 Catalytic Mechanism and Kinetics

pseudocomponents conversions and hydrogen generation. However, some product gases (CO2 , CH4 , C2 H6 , and C3+ ) showed a poor match. The experimental results were obtained in a temperature range of 360–420 ∘ C, in which the thermal cracking reactions dominate over aquathermolysis reactions (Hyne et al. 1984, 1982a); therefore, the model deviates from the behavior under steam injection conditions. The values of kinetic parameters for the sulfur-based kinetic model reported by Lamoureux-Var and Lorant (2007) are presented in Table 7.10. The model results allow determining the distribution of the sulfur content in NSO, aromatics, and resins fractions, which involve some organic proportions and minerals. On the other hand, the simulations carried out by Barroux et al. (2013) showed deviations concerning experimental data reported by Lamoureux-Var and Lorant (2005). Therefore, the stoichiometric coefficient values of Sa/Ar, H2 S, and COK components as well as the frequency factors had to be tuned to fit the calculated profiles with experimental data. Table 7.10 shows the values of frequency factors used. However, the modifications of the stoichiometry could not be justified. The kinetic model of Ayache et al. (2015) predicts correctly the trends of SARA fractions and H2 S, even though the stoichiometric coefficients are restricted by atomic balances and are not adjusted mathematical parameters as in the case of Barroux et al. (2013). Table 7.10 shows that the values of frequency factors and activation energies reported present similar order of magnitude to those of the other models. The values of kinetic coefficients obtained by Zhang et al. (2020) show fit concerning experimental data of reaction gases. Nevertheless, since the estimation of the kinetic parameters was performed individually, the values obtained do not consider the effect of the reactants conversion and simultaneous production of other gases. Therefore, these values are just approximations of the appropriate kinetic parameters. Table 7.11 summarizes the models results developed by Zhang Table 7.10 Kinetic parameters for sulfur-based aquathermolysis kinetic models. Reaction

A0 (day−1 )

E A (kJ/mol)

Lamoureux-Var (2007) R1 (7.84)

8.64 × 1018

204.32

R2 (7.85)

8.64 × 1018

231.11

R3 (7.86)

8.64 × 10

18

203.06

R4 (7.87)

8.64 × 1018

228.60

Barroux et al. (2013) R1 (7.84)

4.50 × 1018

204.32

R2 (7.85)

4.50 × 10

18

231.11

R3 (7.86)

3.50 × 1018

203.06

R4 (7.87)

18

228.60

4.50 × 10

Ayache et al. (2015)

aSH2 S

aSCOK

204.30

0.369

0

222.22

0.899

0.151

R1 (7.88)

1.00 × 10

19

R2 (7.89)

5.00 × 1018

R3 (7.90)

4.50 × 10

18

201.34

0.651

0.349

R4 (7.91)

2.00 × 1018

216.67

0.960

0.04

R5 (7.92)

3.80 × 1018

209.09

Table 7.11 Reaction

Kinetic parameters determined for aquathermolysis based on reaction scheme reported in the literature. 1

2

3

4

5

6

7

8

4.21 × 109

Heavy oil, model of Zhang et al. (2020) without core (HO–H2 S)

(HO–CO2 )

(HO–CH4 )

(HO–H2 )

(HO–CO)

A0 (day−1 )

1638.60

1.70

24.55

647 57.19

0.20

39.67

EA (kJ/mol)

62.49

20.04

25.40

80.70

32.65

26.50

(HO–H2 S)

(HO–CO2 )

(HO–CH4 )

(HO–H2 )

(HO–CO)

(HO–C2+ )

A0 (day )

86.58

0.22

4.65

4621.62

0.05

4.53

EA (kJ/mol)

46.85

6.35

16.90

67.40

23.98

14.73

(HO–C2+ )

Heavy oil, model of Zhang et al. (2020) with core −1

Bitumen, model of Kapadia et al. (2013) (B–H2 )

(B–CH4 )

(B–CO)

(B–CO2 )

(B–H2 S)

(B–HMWG)

A0 (day−1 )

1.47 × 102

3.34 × 10−1

4.85 × 10−4

1.87 × 10−4

1.35 × 100

8.81 × 105

5.57 × 107

EA (kJ/mol)

81.23

55.45

21.67

4.54

54.71

116.2

149

190

(Res-Sat)

(Res–Aro)

(Res–H2 )

(Res–CH4 )

(Res–CO)a)

(Res–CO2 )

(Res–H2 S)

(Res–HMWG)

Resins, model of S. Huang et al. (2017) A0 (day−1 )

5.10 × 10−1

3.29 × 100

3.45 × 10−4

1.67 × 10−3

—

4.90 × 10−3

1.96 × 10−3

3.54 × 10−3

EA (kJ/mol)

48.99

72.62

92.77

58.12

—

41.92

121.5

55.23

Asphaltenes, model of S. Huang et al. (2017) (Asp–Sat)

(Asp–Aro)

(Asp–H2 )

(Asp–CH4 )

(Asp–CO)a)

(Asp–CO2 )

(Asp–H2 S)

(Asp–HMWG)

A0 (day−1 )

2.48 × 10−1

1.59 × 100

1.58 × 10−4

8.09 × 10−4

—

2.37 × 10−3

9.47 × 10−4

1.72 × 10−3

EA (kJ/mol)

39.07

62.7

63.59

48.2

—

32

111.6

45.31

a) The pre-exponential factor and the activation energy was not calculated because of lack of experiment data.

370

7 Catalytic Mechanism and Kinetics

et al. (2020), Kapadia et al. (2013), and S. Huang et al. (2017). The kinetic parameters obtained by Zhang et al. (2020) indicate that the activation energies of the different reaction pathways are reduced in the presence of core samples, corroborating that the chemical and mineralogical content of the rock presents catalytic effects on aquathermolysis reactions. Furthermore, the activation energy values for H2 S production obtained by Zhang et al. (2020) and Kapadia et al. (2013) are in the range of those previously reported with the different models focused on predicting gas production. On the other hand, the value of the frequency factor calculated by Kapadia et al. (2013) for the H2 S generation is closer to the value reported by Hamedi-Shokrlu and Babadagli (2014) during the catalytic aquathermolysis reaction (1.8 day−1 ). The Kapadia et al. (2013) model adequately predicts the gas generation below 300 ∘ C, where it was attributed that thermal cracking reactions predominate at higher temperatures, presenting a deviation due to the higher amount of gas produced. The kinetic model of S. Huang et al. (2017) showed a good fit with experimental data obtained during the aquathermolysis of heavy oils with different viscosities. Activation energy values for H2 S production from asphaltenes and resins are higher than previously reported. However, those values obtained for CO2 production are in the reported range. The Kapadia et al. (2013) kinetic model was adjusted to the experimental data used by Huang et al. (2019) in order to compare the predictability of both models concerning the decay terms. Figure 7.24 shows that the Gaussian rate model presents a fitter, predicting gas generation because it can describe the reaction in two stages. First stage depicts the reaction rate increases to a maximum value. Thus, the second stage depicts the reduction until it reaches equilibrium. Therefore, the exponential rate model is suitable for high-sulfur heavy oil, whereas the Gaussian rate model is more suitable for low-sulfur heavy oil (Huang et al. 2019). The results of the analysis developed by Tirado et al. (2022a) on the use of stoichiometric coefficients during kinetic modeling indicated that the models that group these coefficients into kinetic pseudocoefficients can adequately predict the yields of reaction products and reactants, obtaining adequately results of kinetic parameters concerning experimental data and the kinetic model used. Therefore, the use of each approach can be based on the available experimental data to calculate the different parameters and the complexity level required by the model. The kinetic parameters of those models that consider the yield distribution of SARA fractions and gas production are summarized in Table 7.12. The optimized kinetic coefficients reported by Tirado et al. (2022c) based on the reaction scheme shown in Figure 7.8d indicate that during the catalytic aquathermolysis of Liaohe heavy crude oil using NiO, the saturates and gas compounds are produced mainly by asphaltenes and resins conversion. A similar trend was reported by Félix et al. (2022a) for noncatalytic aquathermolysis of Ashal’cha heavy crude oil. Qu et al. (2021) reported that the activation energy of the different reaction pathways considered for thermal cracking and aquathermolysis experiments of Mackay River oil sand presented lower values during aquathermolysis reaction than thermal cracking process. It was reported that all the kinetic parameters presented a coefficient of determination higher than 0.96. The asphaltenes conversion into resins (As → Re) presents considerable values of kinetic coefficients in the model of Tirado et al. (2022c), Qu et al. (2021), and Félix et al. (2022a), where the difference between the values of these models is attributed to the diverse reactivity of the compounds contained in the unconventional oils because of their different compositions and structures. In addition, the generation of aromatic compounds from asphaltenes (As → Ar) showed significant yields in the models of Qu et al. (2021) and Félix et al. (2022a). On the other hand, the results of the models of Qu et al. (2021) and Tirado et al. (2022c) indicated that condensation reactions (Re → As) are carried out under the operating conditions studied. Both reaction pathways have been reported in the experimental studies.

7.6 Results and Discussion

160 H2 (μmol)

(a)

120 80 40 0 500

CH4 (μmol)

(b)

400 300 200 100

0 1000 CO (μmol)

(c)

800 600 400 200 0 8 H2S (μmol)

(d)

6 4 2 0

500 HMWG (μmol)

(e)

400 300 200 100 0 0

10

20

30

40 50 60 Reaction time (h)

70

80

90

100

Figure 7.24 Comparison of gas production of ( ) Gaussian rate model (Kapadia et al. 2013) and ( ) Exponential rate model (S. Huang et al. 2017) with (o) experimental data: (Fan 2002) (a) H2 , (b) CH4 , (c) CO, (d) H2 S, and (e) HMWG.

According to the reaction rate constants reported by Félix et al. (2022a), the reactions performed are mainly in the series (As → Re → Ar → Sa) as well as some in parallel, such as asphaltenes conversion to aromatics, saturates, and gases compounds. The results of the models presented values of average absolute error lower than 5%, indicating a good fit of the estimated parameters with the experimental data.

371

Table 7.12

Kinetic parameter reported for the aquathermolysis reaction.

Reaction As–Re

As–Ar

As–Sa

As–Gas

Re–As

Re–Ar

Re–Sa

Re–Gas

Ar–As Ar–Re

Ar–Sa

Ar–Gas

Sa–Re Sa–Ar Sa–Gas

Catalytic aquathermolysis of heavy oil, Tirado et al. (2022c) (Figure 7.8d) 200

1.992 × 10−3

4.862 × 10−3 4.764 × 10−6 1.218 × 10−3 1.286 × 10−2 3.739 × 10−7 1.912 × 10−6

1.549 × 10−2 1.689 × 10−7 1.231 × 10−10

1.256 × 10−10

220

1.382 × 10−2

8.637 × 10−3 2.846 × 10−5 7.409 × 10−3 1.583 × 10−2 3.078 × 10−6 3.851 × 10−6

1.684 × 10−2 1.890 × 10−6 2.268 × 10−7

2.713 × 10−8

240

7.295 × 10−2

1.565 × 10−2 1.279 × 10−4 2.898 × 10−2 2.861 × 10−2 9.896 × 10−6 3.864 × 10−6

2.419 × 10−2 3.913 × 10−6 2.310 × 10−7

3.176 × 10−8

260

7.459 × 10−2

2.183 × 10−2 1.942 × 10−4 2.900 × 10−2 3.543 × 10−2 1.252 × 10−5 3.884 × 10−6

2.824 × 10−2 4.788 × 10−6 2.394 × 10−7

3.980 × 10−8

A

1.300 × 10+12

4.130 × 10+3

3.210 × 10+9

9.410 × 10+9

1.920 × 10+2

2.810 × 10+7

7.770 × 10−4

4.620 × 10+0

1.290 × 10+6

9.460 × 10+17

2.260 × 10+11

EA

132.810

53.591

133.337

115.360

38.010

124.073

22.871

22.599

114.530

244.380

200

2.480 × 10−4 9.530E−5

2.400E−5

0

8.980E−5

1.670E−4

5.130E−5

0

2.810E−7 5.340E−5

1.590E−4

0

1.750E−6 1.400E−7 0

186.882

Noncatalytic aquathermolysis of oil sand model of Qu et al. (2021) (Figure 7.8e)

250

1.120 × 10−3 2.010E−4

2.020E−4

7.660E−10

2.050E−4

5.510E−4

3.070E−4

3.780E−10

1.900E−6 1.360E−4

4.870E−4

3.340E-10

1.310E−5 1.230E−6 3.800E−10

300

4.260 × 10−3 2.540E−4

1.540E−3

1.280E−8

4.350E−4

1.760E−3

1.450E−3

5.270E−9

1.090E−5 4.710E−4

1.340E−3

4.030E−9

1.470E−4 1.540E−5 7.820E−9

350

1.050 × 10−2 9.210E−4

9.880E−3

6.080E−8

1.460E−3

4.710E−3

3.580E−3

2.730E−8

3.080E−5 1.470E−3

3.610E−3

2.090E−8

9.700E−4 1.030E−4 3.750E−8

A

1.700 × 10+3 1.210E+0

1.500E+6

7.060E+2

5.890E+0

1.740E+2

3.280E+3

1.770E+2

1.190E+2 4.600E+1

6.040E+1

6.140E+1

4.740E+5 1.320E+5 1.440E+3

EA

61.840

98.180

119.190

44.130

54.690

70.450

116.430

77.990

50.730

112.440

104.280 109.070 125.230

250

7.360 × 10−4 6.010 × 10−4 5.500 × 10−4 3.540 × 10−4

2.480 × 10−3 7.970 × 10−4 1.440 × 10−4

5.700 × 10−4 8.700 × 10−5

1.750 × 10−4

300

1.630 × 10−3 1.390 × 10−3 1.050 × 10−3 1.260 × 10−3

3.800 × 10−3 1.100 × 10−3 4.050 × 10−4

1.080 × 10−3 9.600 × 10−5

1.850 × 10−4

A

6.840 × 10+0 8.910 × 10+0 8.560 × 10−1 7.630 × 10+2

3.310 × 10−1 3.260 × 10−2 2.000 × 10+1

8.620 × 10−1 2.790 × 10−4

3.430 × 10−4

EA

39.740

21.290

31.840

2.940

37.400

54.360

Noncatalytic aquathermolysis of heavy oil, model of Félix et al. (2022a) (Figure 7.8f)

A (h−1 ), EA (kJ/mol)

41.780

31.970

63.440

16.140

51.490

5.080

300 250 200 150 100 50 0

Sa – Sa Gas – Sa Ar – A R r– e A Ga r– s A Sa r– A Re r Re – A – s Re Ga – s Re Sa – Re A r A –A s– s A Ga s– s A Sa s– A A s– r Re

Activation energy (kJ/mol)

7.6 Results and Discussion

Reaction pathway Figure 7.25 Activation energy values reported for the aquathermolysis reaction model: S. Huang (2017) and Huang et al. (2019) ( ), Tirado et al. (2022c) ( ), Qu et al. (2021) ( ), and Félix et al. (2022a) ( ).

Figure 7.25 presents the activation energy values with respect to the diverse reaction pathways considered during the aquathermolysis process of unconventional oils. The activation energy values of Tirado et al. (2022c) and Qu et al. (2021) models display higher values compared with those reported by S. Huang et al. (2017) and Félix et al. (2022a), which can be attributed, in part, to the high asphaltenes content in oils used in the former studies (>18%), while the asphaltenes content in oils used in the latter is lower than 5%. 7.6.2.3

Discussion on Kinetic Studies and Modeling for Aquathermolysis Reaction

Most of the kinetic studies are focused on initial temperatures of 220–240 ∘ C because the generation of H2 S and CO2 is low at temperatures below 235 ∘ C (Jia et al. 2016; Kapadia et al. 2012). The increase in temperature favors the generation of H2 S as well as the asphaltenes and resins conversion to produce lighter fractions (Kapadia et al. 2011; Thimm 2007). However, thermal cracking reactions predominate around 300 ∘ C, resulting in the conversion of heavier molecules into coke by polymerization reactions and increased gas production (Hyne et al. 1982a; Hyne 1986). The experimental studies focused on aquathermolysis reactions are carried out in inert atmosphere of nitrogen or hellion to simulate the reserves conditions. In addition, the use of quartz, gold, or glass tubes to avoid direct contact of oil samples with reactor walls is usually made. In this way, catalytic effect of metals present on the inner walls of the reactor can be avoided. Nonetheless, the equilibrium of the aquathermolysis reactions is not affected by catalytic effects of minerals. The reaction time varies depending on the oil composition, operating conditions, and catalytic effect. However, kinetic studies should be focused on analyzing the advance of the reaction adequately until reaching the equilibrium state under the determined conditions. Otherwise, the model could present erroneous tendencies by exceeding the experimental time. Although the reaction times used in experimental studies are short compared to those used during the steam injection process, these were justified by the difference in volumes treated under laboratory scale and in the field. The equilibrium-produced gas concentration does not change under determined operating conditions. Based on these observations, it is concluded that different reaction pathways can occur during the aquathermolysis reaction of heavy oils depending on feedstock properties and composition, operating conditions, and type of catalyst. The development of new reaction schemes and kinetic models must be supported on experimental data and proper parameter estimation. So far, there is still a lack of understanding of the different chemical reactions that take place in the aquathermolysis reaction of heavy crude oil, and more detailed studies are necessary.

373

374

7 Catalytic Mechanism and Kinetics

Experimental studies have reported the effect of hydrogen donors, such as tetralin and decalin, which in the presence of catalysts can significantly reduce the viscosity of heavy oil. However, these effects have not been examined in kinetic studies that are critical for aquathermolysis in situ oil upgrading and recovery (Jiang et al. 2005; Ovalles et al. 2001; Zhang et al. 2012). Some kinetic models present a complex number of kinetic parameters such as activation energies, frequency factors, stoichiometric coefficients, and even correction factors. The determination of these parameters requires a high number of experimental data and proper methods to assure that the set of calculated parameters provides the best approximation with measured values. The use of statistical techniques capable of determining the optimal values of parameters of a kinetic model is highly recommended (Alcázar and Ancheyta 2007). Approaches such as nonlinear optimization, sensitivity, and residual analysis have clearly demonstrated that some literature works, which do not take this into consideration, have reported erroneous values of kinetic parameters (Tirado and Ancheyta 2020).

7.7

Conclusion

7.7.1

Hydrocracking Kinetic Models

In this chapter, the kinetic models reported in the literature for in situ and ex situ upgrading of heavy crude oil or residue with liquid or ultradispersed nanocatalyst in a hydrogen atmosphere were described and analyzed in terms of the type of kinetic model, number of lumps, reaction pathways, and calculated values of reaction rate coefficients and activation energies. The most used technique by almost all the reported kinetic studies to model the reaction mechanism is the lumping approach due to its ease to estimate the kinetic parameter estimation. In addition, the nonlinear regression is the most employed method since it is agree that this gives better results when optimizing kinetic parameters than linear regression. However, the methods to calculate the reaction rate coefficients reported by almost all the studies do not ensure the optimal values for the kinetic parameters obtained. Therefore, the development of a robust algorithm is needed to assure that the obtained parameter values achieve the global minimum (and not local minima) of the objective function. Some lumping kinetic models reported reaction rate coefficients inaccurately estimated because their values do not follow the Arrhenius equation law (increasing with temperature), thus the values of activation energy are incorrectly calculated. The values of activation energies reported range between 7 and 633 kJ/mol, and the principal tendency in almost all reaction pathways analyzed agree that the easier reaction pathways are mainly in series. The difference in the activation energies depends on the feedstock, catalyst, or reactor type employed. The order of reaction in almost all kinetic models reviewed during the hydrocracking of heavy crude oil using moderate conditions is first order of reaction, while second order of reaction is observed for severe operating conditions.

7.7.2

Aquathermolysis Kinetic Models

Different aspects during the development of kinetic models for the aquathermolysis reaction were analyzed. The operating conditions and setup used for experimental tests influence the yield of reaction products due to aquathermolysis reactions that predominate under an established temperature range, and the experimental results can be influenced by different factors. Those reported models based on the distribution of sulfur compounds in the whole composition of heavy oils make use of some assumptions to have better accuracy but become more empirical in nature.

References

The different reaction pathways carried out are influenced by the composition and structure of the compounds in oils as well as the catalytic effects involved. The development of complex kinetic models highly depends on the number of experimental data available to predict the reaction product distribution. More studies with model compounds and real heavy oils are still necessary for a better understanding of catalytic and noncatalytic reaction mechanisms. Furthermore, the effect of hydrogen donors has not been considered in kinetic models for the aquathermolysis reaction.

References Al-Attas, T.A., Ali, S.A., Zahir, M.H. et al. (2019). Recent advances in heavy oil upgrading using dispersed catalysts. Energy and Fuels 33: 7917–7949. Alcázar, L.A. and Ancheyta, J. (2007). Sensitivity analysis based methodology to estimate the best set of parameters for heterogeneous kinetic models. Chemical Engineering Journal 128: 85–93. Aliev, F.A., Mukhamatdinov, I.I., Sitnov, S.A. et al. (2021). In-situ heavy oil aquathermolysis in the presence of nanodispersed catalysts based on transition metals. Processes 9: 1–22. Al-Marshed, A., Hart, A., Leeke, G. et al. (2015). Effectiveness of different transition metal dispersed catalysts for in situ heavy oil upgrading. Industrial and Engineering Chemistry Research 54: 10645–10655. Al-Muntaser, A.A., Varfolomeev, M.A., Suwaid, M.A. et al. (2021). Hydrogen donating capacity of water in catalytic and non-catalytic aquathermolysis of extra-heavy oil: deuterium tracing study. Fuel 283: 118957. Al-Muntaser, A.A., Varfolomeev, M.A., Suwaid, M.A. et al. (2022). Effect of decalin as hydrogen-donor for in-situ upgrading of heavy crude oil in presence of nickel-based catalyst. Fuel 313: 122652. Álvarez, P., Browning, B., Jansen, T. et al. (2019). Modeling of atmospheric and vacuum petroleum residue hydroconversion in a slurry semi-batch reactor: study of hydrogen consumption. Fuel Processing Technology 185: 68–78. Ancheyta, J. (2013). Modeling of Processes and Reactors for Upgrading of Heavy Petroleum, 1e. New York: CRC Press. Ancheyta, J. (2020). Experimental Methods for Evaluation of Hydrotreating Catalysts, 1e. Hoboken, NJ: Wiley. Ancheyta, J. and Sotelo-Boyás, R. (2000). Estimation of kinetic constants of a five-lump model for fluid catalytic cracking process using simpler sub-models. Energy and Fuels 14: 1226–1231. Ancheyta, J. and Speight, J.G. (2007). Hydroprocessing of Heavy Oils and Residua. CRC Press. Ancheyta, J., Sánchez, S., and Rodríguez, M.A. (2005). Kinetic modeling of hydrocracking of heavy oil fractions: a review. Catalysis Today 109: 76–92. Ancheyta, J., Ramírez, S., Leyva, C., Quitian, A. (2018). Hydroconversion process to upgrade the transport properties of heavy and extra-heavy crude oils at low severity conditions using dispersed-phase catalyst. U.S. Patent. US9994779B2. Angeles, M.J., Leyva, C., Ancheyta, J., and Ramírez, S. (2014). A review of experimental procedures for heavy oil hydrocracking with dispersed catalyst. Catalysis Today 220–222: 274–294. Asaee, S.D.S., Vafajoo, L., and Khorasheh, F. (2014). A new approach to estimate parameters of a lumped kinetic model for hydroconversion of heavy residue. Fuel 134: 343–353. Ashoori, S., Sharifi, M., Masoumi, M., and Mohammad Salehi, M. (2017). The relationship between SARA fractions and crude oil stability. Egyptian Journal of Petroleum 26: 209–213. Ayache, S. V., Lamoureux-Var, V., Michel, P., and Preux, C. (2015). Reservoir simulation of H2 S production during a SAGD process using a new sulfur-based-compositional kinetic model. Society of Petroleum Engineers – SPE Canada Heavy Oil Technical Conference 2015, CHOC 2015. 1265–1281.

375

376

7 Catalytic Mechanism and Kinetics

Ayache, S.V., Lamoureux-Var, V., Michel, P., and Preux, C. (2017). Reservoir simulation of hydrogen sulfide production during a steam-assisted-gravity-drainage process by use of a new sulfur-based compositional kinetic model. SPE Journal 22: 080–093. Balasubramanian, P. and Pushpavanam, S. (2008). Model discrimination in hydrocracking of vacuum gas oil using discrete lumped kinetics. Fuel 87: 1660–1672. Barroux, C., Flauraud, E., and Nouvelles, I.F.P.E. (2013). Forecasting of H2 S Production due to Aquathermolysis Reactions. SPE Middle East Oil and Gas Show and Conference, Manama, Bahrain (10–13 March). Becker, P.J., Serrand, N., Celse, B. et al. (2016). Comparing hydrocracking models: continuous lumping vs single events. Fuel 165: 306–315. Belgrave, J.D.M., Moore, R.G., and Ursenbach, M.G. (1997). Comprehensive kinetic models for the aquathermolysis of heavy oils. Journal of Canadian Petroleum Technology 36: 38–44. Bellussi, G., Rispoli, G., Landoni, A. et al. (2013). Hydroconversion of heavy residues in slurry reactors: developments and perspectives. Journal of Catalysis 308: 189–200. del Bianco, A., Panariti, N., di Carlo, S. et al. (1994). New developments in deep hydroconversion of heavy oil residues with dispersed catalysts. 2. Kinetic aspects of reaction. Energy & Fuels 8: 593–597. Browning, B., Afanasiev, P., Pitault, I. et al. (2016). Detailed kinetic modelling of vacuum gas oil hydrocracking using bifunctional catalyst: a distribution approach. Chemical Engineering Journal 284: 270–284. Browning, B., Pitault, I., Couenne, F. et al. (2019). Distributed lump kinetic modeling for slurry phase vacuum residue hydroconversion. Chemical Engineering Journal 377: 119811. Bueno Zapata, N., Morales Mora, O.A., and Mejía Cárdenas, J.M. (2019). Practical kinetic coupling to multi-component and multi-phase flow transport during in-situ heavy oil upgrading processes using an equation of state-based numerical reservoir simulation. Society of Petroleum Engineers. Cai, Z., Ma, Y., Zhang, J. et al. (2022). Tunable ionic liquids as oil-soluble precursors of dispersed catalysts for suspended-bed hydrocracking of heavy residues. Fuel 313: 122664. Castañeda, L.C., Muñoz Arroyo, J.A.D., and Ancheyta, J. (2014). Current situation of emerging technologies for upgrading of heavy oils. Catalysis Today 220–222: 248–273. Clark, P.D., Hyne, J.B., and Tyrer, J.D. (1983). Chemistry of organosulphur compound types occurring in heavy oil sands. Fuel 62: 959–962. Clark, P.D., Clarke, R.A., Hyne, J.B., and Lesage, K.L. (1990). Studies on the chemical reactions of heavy oils under steam stimulation conditions. AOSTRA Journal of Research 6: 29–39. Coronel-García, M.A., Reyes de la Torre, A.I., Domínguez-Esquivel, J.M. et al. (2021). Heavy oil hydrocracking kinetics with nano-nickel dispersed in PEG300 as slurry phase catalyst using batch reactor. Fuel 283: 118930. Da Silva De Andrade, F.J. (2014). Kinetic modeling of catalytic in situ upgrading for Athabasca bitumen, deasphalting pitch and vacuum residue (PhD. Thesis). University of Calgary. Du, H., Li, M., Liu, D. et al. (2015). Slurry-phase hydrocracking of heavy oil and model reactant: effect of dispersed Mo catalyst. Applied Petrochemical Research 5: 89–98. Duran Armas, J.L. (2021). Application of In-situ Upgrading in Naturally Fractured Reservoirs. Elahi, S.M., Scott, C.E., Chen, Z., and Pereira-Almao, P. (2019). In-situ upgrading and enhanced recovery of heavy oil from carbonate reservoirs using nano-catalysts: upgrading reactions analysis. Fuel 252: 262–271. Elahi, S.M., Ahmadi Khoshooei, M., Carbognani Ortega, L. et al. (2020). Chemical insight into nano-catalytic in-situ upgrading and recovery of heavy oil. Fuel 278: 118270. Elizalde, I. and Ancheyta, J. (2014). Modeling catalyst deactivation during hydrocracking of atmospheric residue by using the continuous kinetic lumping model. Fuel Processing Technology 123: 114–121.

References

Elizalde, I., Rodríguez, M.A., and Ancheyta, J. (2009). Application of continuous kinetic lumping modeling to moderate hydrocracking of heavy oil. Applied Catalysis A: General 365: 237–242. Elizalde, I., Trejo, F., Muñoz Arroyo, J.A.D. et al. (2016). Dynamic modeling and simulation of a bench-scale reactor for the hydrocracking of heavy oil by using the continuous kinetic lumping approach. Reaction Kinetics, Mechanisms and Catalysis 118: 299–311. Fan, H.F. (2002). Study and Application of Aquathermolysis for Liaohe Heavy Oils. Heilongjiang: Northeast Petroleum University. Fan, H. (2003). The effects of reservoir minerals on the composition changes of heavy oil during steam stimulation. Journal of Canadian Petroleum Technology 42: 11–14. Fan, H., Zhang, Y., and Lin, Y. (2004). The catalytic effects of minerals on aquathermolysis of heavy oils. Fuel 83: 2035–2039. Félix, G. and Ancheyta, J. (2019a). Using separate kinetic models to predict liquid, gas, and coke yields in heavy oil hydrocracking. Industrial & Engineering Chemistry Research 58: 7973–7979. Félix, G. and Ancheyta, J. (2019b). Comparison of hydrocracking kinetic models based on SARA fractions obtained in slurry-phase reactor. Fuel 241: 495–505. Félix, G., Quitian, A., Rodríguez, E. et al. (2017). Methods to calculate hydrogen consumption during hydrocracking experiments in batch reactors. Energy & Fuels 31: 11690–11697. Félix, G., Ancheyta, J., and Trejo, F. (2019). Sensitivity analysis of kinetic parameters for heavy oil hydrocracking. Fuel 241: 836–844. Félix, G., Tirado, A., Chengdong, Y. et al. (2022a). Analysis of kinetic models for hydrocracking of heavy oils for in-situ and ex-situ applications. Fuel 323: 124322. Félix, G., Tirado, A., Al-Muntaser, A. et al. (2022b). SARA-based kinetic model for non-catalytic aquathermolysis of heavy crude oil. Journal of Petroleum Science and Engineering 216 (2): 110845. Fogler, H.S. (2020). Elements of Chemical Reaction Engineering, 6e. Boston: Pearson. Fukuyama, H. and Terai, S. (2007). Kinetic study on the hydrocracking reaction of vacuum residue using a lumping model. Petroleum Science and Technology 25: 277–287. Galarraga, C.E., Scott, C., Loria, H., and Pereira-Almao, P. (2012). Kinetic models for upgrading athabasca bitumen using unsupported NiWMo catalysts at low severity conditions. Industrial and Engineering Chemistry Research 51: 140–146. Gauthier, S., Keane, C.J., Niemela, J.J. et al. (2006). Optimization algorithms and weighting factors for analysis of dynamic PET studies. Physics in Medicine & Biology 51: 4217. Gillis, K.A., Palmgren, C., and Thimm, H.F. (2000). Simulation of Gas Production in SAGD. In: SPE/CIM International Conference on Horizontal Well Technology. SPE. Goicetty, J.L.B. (2010). Estimation of the H2 S formation under steam injection conditions for the Orinoco Oil Belt. In: Proceedings – SPE Annual Technical Conference and Exhibition 7, 5870–5874. SPE. Gray, M.R., Khorasheh, F., and Wanke, S.E. (1992). Role of catalyst in hydrocracking of residues from Alberta bitumens. Energy & Fuels 6: 478–485. Guntermann, K., Gudenau, H.W., and Mohtadi, M. (1982). Mathematical modelling of the in situ coal gasification. In: Process Proceedings of the Eighth Underground Coal Conversion Symposium, 297–306. Hajdo, L.E., Hallam, R.J., and Vorndran, L.D.L. (1985). Hydrogen generation during in-situ combustion. In: SPE California Regional Meeting, 675–689. Society of Petroleum Engineers. Hamedi-Shokrlu, Y. and Babadagli, T. (2014). Kinetics of the in-situ upgrading of heavy oil by nickel nanoparticle catalysts and its effect on cyclic-steam-stimulation recovery factor. SPE Reservoir Evaluation and Engineering 17: 355–364. Hashemi, R., Nassar, N.N., and Pereira Almao, P. (2014). Nanoparticle technology for heavy oil in-situ upgrading and recovery enhancement: opportunities and challenges. Applied Energy 133: 374–387.

377

378

7 Catalytic Mechanism and Kinetics

Hassanzadeh, H. and Abedi, J. (2010). Modelling and parameter estimation of ultra-dispersed in situ catalytic upgrading experiments in a batch reactor. Fuel 89: 2822–2828. Hinkle, A. and Batzle, M. (2006). Heavy oils: a worldwide overview. The Leading Edge 25: 742–749. Hongfu, F., Yongjian, L., Liying, Z., and Xiaofei, Z. (2002). The study on composition changes of heavy oils during steam stimulation processes. Fuel 81: 1733–1738. Huang, S., Huang, Q., Liu, H. et al. (2017). A modified model for aquathermolysis and its application in numerical simulation. Fuel 207: 568–578. Huang, T., Liu, B., Wang, Z., and Guo, X. (2017). Kinetic model for hydrocracking of Iranian heavy crude with dispersed catalysts in slurry-phase. Petroleum Science and Technology 35: 1846–1851. Huang, S., Cao, M., Huang, Q. et al. (2019). Study on reaction equations of heavy oil aquathermolysis with superheated steam. International Journal of Environmental Science and Technology 16: 5023–5032. Hyne, J. B. (1986). Aquathermolysis. A synopsis of work on the chemical reaction between water (steam) and heavy oil sands during simulated steam simulation. Alberta. Hyne, J.B., Clark, P.D., Clarke, R.A., and Koo, J. (1982a). Aquathermolysis-the reaction of organo-sulphur compounds during steaming of heavy oil. In: Proceedings, 2nd International Conference on Heavy Crude and Tar Sands, 1, Chapter 9. Caracas, Venezuela: Caracas. Hyne, J.B., Clark, P.D., Clarke, R.A. et al. (1982b). Aquathermolysis of heavy oils. Rev. Tec. INTEVEP; (Venezuela) 2:2: EDB-85-062549. Hyne, J.B., Greidanus, J.W., Tyrer, J.D. et al. (1984). Aquathermolysis of heavy oils. Revista Técnica INTEVEP; (Venezuela) 404–411. Ibatullin, T.R., Yang, T., Petersen, E.B. et al. (2011). Simulation of hydrogen sulfide and carbon dioxide production during thermal recovery of bitumen. In: Society of Petroleum Engineers – SPE Reservoir Characterisation and Simulation Conference and Exhibition 2011, RCSC 2011, 276–286. OnePetro. Jia, N., Zhao, H., Yang, T. et al. (2016). Experimental measurements of Bitumen–Water aquathermolysis during a steam-injection process. Energy & Fuels 30: 5291–5299. Jiang, S., Liu, X., Oilfield, L. et al. (2005). In situ upgrading heavy oil by aquathermolytic treatment under steam injection conditions. In: Proceedings – SPE International Symposium on Oilfield Chemistry, 1–8. SPE. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2010). A comprehensive kinetic theory to model thermolysis, aquathermolysis, gasification, combustion, and oxidation of Athabasca bitumen. In: SPE – DOE Improved Oil Recovery Symposium Proceedings 1, 263–293. SPE. Kapadia, P.R., Wang, J., Kallos, M.S., and Gates, I.D. (2011). Reactive thermal reservoir simulation: hydrogen sulphide production in SAGD. In: All Days, 2173–2183. SPE. Kapadia, P.R., Wang, J., Kallos, M.S., and Gates, I.D. (2012). New thermal-reactive reservoir engineering model predicts hydrogen sulfide generation in Steam Assisted Gravity Drainage. Journal of Petroleum Science and Engineering 94–95: 100–111. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2013). A new reaction model for aquathermolysis of Athabasca bitumen. Canadian Journal of Chemical Engineering 91: 475–482. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2015). A review of pyrolysis, aquathermolysis, and oxidation of Athabasca bitumen. Fuel Processing Technology 131: 270–289. Katritzky, A.R., Allin, S.M., and Siskin, M. (1996). Aquathermolysis: reactions of organic compounds with superheated water. Accounts of Chemical Research 29: 399–406. Kim, S.-H., Kim, K.-D., and Lee, Y.-K. (2017). Effects of dispersed MoS2 catalysts and reaction conditions on slurry phase hydrocracking of vacuum residue. Journal of Catalysis 347: 127–137. Lamoureux-Var, V. and, Lorant, F (2007). Method for constructing a kinetic model allowing the mass of hydrogen sulfide produced by aquathermolysis to be estimated. US 2007/0197807 a1.

References

Lamoureux-Var, V. and Lorant, F. (2005). H2 S artificial formation as a result of steam injection for EOR: a compositional kinetic approach. SPE/PS-CIM/CHOA International Thermal Operations and Heavy Oil Symposium Proceedings 2005. Laxminarasimhan, C.S., Verma, R.P., and Ramachandran, P.A. (1996). Continuous lumping model for simulation of hydrocracking. AIChE Journal 42: 2645–2653. Li, K., Hou, B., Wang, L., and Cui, Y. (2014). Application of carbon nanocatalysts in upgrading heavy crude oil assisted with microwave heating. Nano Letters 14: 3002–3008. Li, C., Huang, W., Zhou, C., and Chen, Y. (2019). Advances on the transition-metal based catalysts for aquathermolysis upgrading of heavy crude oil. Fuel 257: 115779. Lin, R., Chen, K., Miao, M. et al. (2020). Reaction mechanism of H2 S generation during tetrahydrothiophene aquathermolysis reaction. Energy and Fuels 34: 2781–2789. Liu, D., Kong, X., Li, M., and Que, G. (2009). Study on a water-soluble catalyst for slurry-phase hydrocracking of an atmospheric residue. Energy and Fuels 23: 958–961. Loria, H., Trujillo-Ferrer, G., Sosa-Stull, C., and Pereira-Almao, P. (2011). Kinetic modeling of bitumen hydroprocessing at in-reservoir conditions employing ultradispersed catalysts. Energy and Fuels 25: 1364–1372. Ma, Y., Zhang, J., Wu, W. et al. (2022). Trialkylmethylammonium molybdate ionic liquids as novel oil-soluble precursors of dispersed metal catalysts for slurry-phase hydrocracking of heavy oils. Chemical Engineering Science 253: 117516. Machefer, S., Falk, L., and de Panthou, F. (2013). Intensification principle of a new three-phase catalytic slurry reactor. Part II: Eco-efficiency and techno-economic performances. Chemical Engineering and Processing: Process Intensification 70: 267–276. Maity, S.K., Ancheyta, J., and Marroquín, G. (2010). Catalytic aquathermolysis used for viscosity reduction of heavy crude oils: a review. Energy and Fuels 24: 2809–2816. Marafi, A., Stanislaus, A., and Furimsky, E. (2010). Kinetics and modeling of petroleum residues hydroprocessing. Catalysis Reviews – Science and Engineering 52: 204–324. Marquardt, D.W. (1963). An algorithm for least-squares estimation of nonlinear parameters. Journal of the Society for Industrial and Applied Mathematics 11: 431–441. Marques, J., Guillaume, D., Merdrignac, I. et al. (2011). Effect of catalysts acidity on residues hydrotreatment. Applied Catalysis B: Environmental 101: 727–737. Martínez, J. and Ancheyta, J. (2014). Modeling the kinetics of parallel thermal and catalytic hydrotreating of heavy oil. Fuel 138: 27–36. Martinez-Grimaldo, H.J., Chavarria-Hernandez, J.C., Ramírez, J. et al. (2011). Prediction of sulfur content, API gravity, and viscosity using a continuous mixture kinetic model for maya crude oil hydrocracking in a slurry-phase reactor. Energy and Fuels 25: 3605–3614. Matsumura, A., Sato, S., Kondo, T. et al. (2005). Hydrocracking Marlim vacuum residue with natural limonite. Part 2: Experimental cracking in a slurry-type continuous reactor. Fuel 84: 417–421. Miki, Y., Yamadaya, S., Oba, M., and Sugimoto, Y. (1983). Role of catalyst in hydrocracking of heavy oil. Journal of Catalysis 83: 371–383. Muraza, O. and Galadima, A. (2015). Aquathermolysis of heavy oil: a review and perspective on catalyst development. Fuel 157: 219–231. Nguyen, T.S., Tayakout-Fayolle, M., Ropars, M., and Geantet, C. (2013). Hydroconversion of an atmospheric residue with a dispersed catalyst in a batch reactor: kinetic modeling including vapor-liquid equilibrium. Chemical Engineering Science 94: 214–223. Nguyen, M.T., Nguyen, N.T., Cho, J. et al. (2016). A review on the oil-soluble dispersed catalyst for slurry-phase hydrocracking of heavy oil. Journal of Industrial and Engineering Chemistry 43: 1–12.

379

380

7 Catalytic Mechanism and Kinetics

Nguyen, N., Chen, Z., Pereira Almao, P. et al. (2017). Reservoir simulation and production optimization of bitumen/heavy oil via nanocatalytic in situ upgrading. Industrial and Engineering Chemistry Research 56: 14214–14230. Orozco Castillo, C. (2016). In Situ Heavy Oil Upgrading Through Ultra-disperse Nano-catalyst Injection in Naturally Fracture Reservoirs (PhD. Thesis). University of Calgary. Ortega García, F.J., Muñoz Arroyo, J.A.D., Flores Sánchez, P. et al. (2017). Hydrocracking kinetics of a heavy crude oil on a liquid catalyst. Energy & Fuels 31: 6794–6799. Ovalles, C., Vallejos, C., Vasquez, T. et al. (2001). Extra-Heavy crude oil downhole upgrading process using hydrogen donors under steam injection conditions. In: All Days. SPE. Perego, C. and Peratello, S. (1999). Experimental methods in catalytic kinetics. Catalysis Today 52: 133–145. Perez-Perez, A., Kamp, A.M., Soleimani, H., and Darche, G. (2011). Numerical simulation of H2 S and CO2 generation during SAGD. In: 2011 World Heavy Oil Congress, 1–16. Canada: Alberta. Pham, H.H., Thuy Nguyen, N., Go, K.S. et al. (2020). Kinetic study of thermal and catalytic hydrocracking of asphaltene. Catalysis Today 353: 112–118. Pham, H.H., Kim, K.H., Go, K.S. et al. (2021). Hydrocracking and hydrotreating reaction kinetics of heavy oil in CSTR using a dispersed catalyst. Journal of Petroleum Science and Engineering 197: 107997. Puron, H., Chin, K.K., Pinilla, J.L. et al. (2014). Kinetic analysis of vacuum residue hydrocracking in early reaction stages. Fuel 117: 408–414. Qu, X., Li, Y., Li, S. et al. (2021). Thermal cracking, aquathermolysis, and their upgrading effects of Mackay River oil sand. Journal of Petroleum Science and Engineering 201: 920–4105. Quitian, A., Leyva, C., Ramírez, S., and Ancheyta, J. (2015). Exploratory study for the upgrading of transport properties of heavy oil by slurry-phase hydrocracking. Energy and Fuels 29: 9–15. Rodriguez-DeVecchis, V.M., Carbognani Ortega, L., Scott, C.E., and Pereira-Almao, P. (2017). Thermal upgrading of Athabasca bitumen in porous media: limitations and kinetic modelling. Fuel 208: 566–575. Sadighi, S. and Zahedi, G.R. (2013). Comparison of kinetic-based and artificial neural network modeling methods for a pilot scale vacuum gas oil hydrocracking reactor. Bulletin of Chemical Reaction Engineering and Catalysis 8: 125–136. Sámano, V., Tirado, A., Félix, G., and Ancheyta, J. (2020). Revisiting the importance of appropriate parameter estimation based on sensitivity analysis for developing kinetic models. Fuel 267: 117113. Sánchez, S., Rodríguez, M.A., and Ancheyta, J. (2005). Kinetic model for moderate hydrocracking of heavy oils. Industrial & Engineering Chemistry Research 44: 9409–9413. Scherzer, J. and Gruia, A.J. (1996). Hydrocracking science and technology. In: Hydrocracking Science and Technology, 1e. Boca Raton: CRC Press. Scott, C.E., Carbognani-Ortega, L., and Pereira-Almao, P. (2019). In situ upgrading via hot fluid and nanocatalyst injection. In: Advanced Catalytic Materials: Current Status and Future Progress, 129–149. Springer International Publishing. Song, F.M., Liu, C.G., Zhou, G.S., and Yang, G.J. (2004). Thermal hydrocracking kinetics of Chinese gudao vacuum residue. Petroleum Science and Technology 22: 689–708. Speight, J.G. (2013). Heavy and Extra-heavy Oil Upgrading Technologies. Elsevier Science. Suwaid, M.A., Varfolomeev, M.A., Al-muntaser, A.A. et al. (2020). In-situ catalytic upgrading of heavy oil using oil-soluble transition metal-based catalysts. Fuel 281: 118753. Thimm, H.F. (2007). Prediction of hydrogen sulphide production in SAGD projects. Canadian International Petroleum Conference 2007, CIPC 2007 0–2. Tirado, A. and Ancheyta, J. (2020). Modeling of a bench-scale fixed-bed reactor for catalytic hydrotreating of vegetable oil. Renewable Energy 148: 790–797.

References

Tirado, A., Félix, G., Kwofie, M. et al. (2022a). Kinetics of heavy oil non-catalytic aquathermolysis with and without stoichiometric coefficients. Fuel 323: 124365. Tirado, A., Yuan, C., Varfolomeev, M.A., and Ancheyta, J. (2022b). Kinetic modeling of aquathermolysis for upgrading of heavy oils. Fuel 310: 122286. Tirado, A., Félix, G., Varfolomeev, M.A. et al. (2022c). Defining reaction pathways for catalytic aquathermolysis of Liohe crude oil. Fuel 333: https://doi.org/10.1016/j.fuel.2022.126345. Tumanyan, B.P., Petrukhina, N.N., Kayukova, G.P. et al. (2015). Aquathermolysis of crude oils and natural bitumen: chemistry, catalysts and prospects for industrial implementation. Russian Chemical Reviews 84: 1145–1175. Vakhin, A.V., Aliev, F.A., Kudryashov, S.I. et al. (2018). Aquathermolysis of heavy oil in reservoir conditions with the use of oil-soluble catalysts: part I–changes in composition of saturated hydrocarbons. Petroleum Science and Technology 36: 1829–1836. Vakhin, A.V., Aliev, F.A., Mukhamatdinov, I.I. et al. (2020). Catalytic aquathermolysis of boca de jaruco heavy oil with nickel-based oil-soluble catalyst. Processes 8, 532. Weitkamp, J. (2012). Catalytic hydrocracking-mechanisms and versatility of the process. ChemCatChem 4: 292–306. Zhang, S., Liu, D., Deng, W., and Que, G. (2007). A review of slurry-phase hydrocracking heavy oil technology. Energy and Fuels 21: 3057–3062. Zhang, Z., Barrufet, M., Lane, R., and Mamora, D. (2012). Experimental Study of In-Situ Upgrading for Heavy Oil Using Hydrogen Donors and Catalyst under Steam Injection Condition. All Days, 1610–1616. SPE. Zhang, J., Han, F., Yang, Z. et al. (2020). Significance of aquathermolysis reaction on heavy oil recovery during the steam-assisted gravity drainage process. Energy & Fuels 34: 5426–5435. Zhang, X., Che, H., and Liu, Y. (2021). Enhanced aquathermolysis of extra-heavy oil by application of transition metal oxides submicro-particles in relation to steam injection processes. Journal of Petroleum Exploration and Production Technology 11: 4019–4028.

381

382

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms of In Situ Upgrading Nail Khafizov 1 , Vadim Neklyudov 1,2 , Anastasiya Mikhailova 1 , and Oleg Kadkin 1 1

Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Haifa 32000, Israel

2 Technion,

8.1

Introduction

Despite the global problem of carbon dioxide emission and the rapid advancement of the so-called “green” power sources, fossil fuels will still cover an overwhelming part of the world’s energy demand in the near future for a number of reasons, such as: their low cost compared to the anticipated alternatives; the already existing developed infrastructure for their processing, storage, and transportation; vast (though nonrenewable) reserves; and the near-term irreplaceability of fossil fuels in many sectors of world’s economy. However, easily recovered fossil fuel deposits come to depletion. Therefore, the development of technologies for the recovery and processing of fossil fuels that were previously considered too difficult and expensive to extract and to process (in particular, these refer to heavy oil, bitumen, and oil distillation residues) is a priority of the global oil and gas industry. In situ upgrading is a promising approach to increase the productivity of oil wells through improving the rheological properties of heavy oil fractions. The downhole upgrading processes in various technology versions may include partial combustion, partial oxidation, thermal cracking, hydrogenation, hydrocracking, steam reforming, and aquathermolysis (Clark and Kirk 1994; Greaves et al. 2000; Wu et al. 2010; Maity et al. 2010; Shah et al. 2010; Chao et al. 2012; Shokrlu et al. 2013; Kapadia et al. 2013; Hart 2014; Galukhin et al. 2015; Tumanyan et al. 2015; Kayukova et al. 2016; Hart. et al. 2017; Mironenko et al. 2017). Some of these processes require extremely high temperatures and pressures, while the others can proceed under mild conditions. Moreover, some of the in situ thermal methods can be further enhanced by addition of catalysts. High temperatures necessary for in situ upgrading of heavy oil fractions in the reservoir are produced on account of either the energy of combustion and oxidation reactions or the thermal energy of injected superheated steam (Santos et al. 2014). As can be concluded from the foregoing analysis, the existing in situ upgrading technologies involve a very broad range of various physical, physicochemical, and chemical processes. Because of complexity and diversity of chemical reactions that take place in heavy oil reservoirs under high pressures and temperatures during in situ upgrading, the theoretical treatment of these reactions meets certain difficulties. However, the current level of development of quantum chemical methods and computational powers allows one to simulate typical heavy oil molecules and

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

8.2 A General View of In Situ Upgrading Processes from the Standpoint of Physical Chemistry

to find the ways to theoretically process the thermochemistry and kinetics of their fragmentation under conditions of in situ upgrading. In addition, quantum chemical calculations can serve as a valuable tool in the development of catalysts for different chemical transformations occurring in oil reservoirs under extreme conditions. Below we describe the general methodology, calculation techniques, and preliminary results of applying quantum chemistry methods for studying complex physicochemical phenomena that accompany the in situ upgrading processes.

8.2 A General View of In Situ Upgrading Processes from the Standpoint of Physical Chemistry In view of the diverse chemical composition of heavy oil components and, consequently, a wide range of chemical reactions involved, it is critical to generalize and systematize the in situ upgrading processes when attempting to apply the theoretical calculation tools to their development. It seems that relatively common approaches can be developed on the basis of distinguishing the chemical nature of oil components, different types of chemical bonds, and different chemical reaction mechanisms involved in bond cleavage reactions. It should be admitted that the necessity of supplying large amounts of energy (first of all) and reagents is a common attribute of all in situ upgrading techniques, since the heavy oil deposits are chemical systems that are thermodynamically equilibrated for many thousands of years, even if they are composed of energy saturated components. The last circumstance—i.e. the presence of high-energy components—is used in various versions of in situ partial combustion technologies, in which high temperatures are produced inside the oil reservoir on account of the energy of oxidation reactions. This makes it possible to trigger certain thermochemical processes that largely involve free radicals. The free radicals are generated both in the combustion process and during thermal degradation of components that are not engaged in the oxidation reactions. In the elementary steps, the thermal degradation implies the thermolysis of C—H, C—C, and other bonds, including the heteroatomic ones. The generated free radicals may be further stabilized in either addition reactions or elimination reactions. In the latter case, the radicals may be eventually stabilized via either one-step or multistep processes leading to further fragmentation, which results in a release of low-molecular-weight reaction products (for example, hydrogen, the compounds of heteroatoms with hydrogen, low hydrocarbons, and other volatiles). The cyclization and aromatization reactions may also play an essential role in stabilizing the generated free radicals. Addition reactions are not always favorable for ennobling heavy oil components – perhaps, excluding the case of producing branched hydrocarbons with intermediate molecular weights, since the branching of lower hydrocarbons (in contrast to the branching of high-molecular-weight hydrocarbons) leads to lower viscosity in comparison with linear homologues (Haynes 2014). Dilution with volatile reaction products (this can be achieved by controlling the oxygen flow) and, probably, even the injection of free radical scavengers (some of them might be sufficiently cheap to achieve economic efficiency) are apparent ways that come first to mind when we think about restraining the additional reactions. In any case, the polymerization processes cannot be completely suppressed, and some coke formation is inevitable. In all the partial oxidation technologies, which are normally associated with the development of rather high pressures and temperatures (300 ∘ C and much higher) in heavy oil reservoirs, it seems that the search for the proper balance between temperature regimes, production of unnecessary volatile oxides (mainly CO and CO2 ), and inescapable coke formation is a challenging problem that faces the research community working in the field.

383

384

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

In contrast to partial combustion technologies, steam injection technologies are associated with milder temperature regimes (steam is considered superheated after reaching 375 ∘ C, so it is very cost ineffective to develop temperatures higher than 300 ∘ C in heavy oil reservoirs even with superheated steam), selectivity of the chemical processes, and chemical reaction conditions that involve the generation of ionic intermediates in a medium with a high dielectric constant. Moreover, water may also serve as a reagent in the hydrolysis reactions or may serve as an amphotheric catalyst in different ionic processes. At any rate, all possible chemical reactions in which water may take part as a medium (solvent), a reagent, a catalyst, or a catalyst promotor are classified into the so-called “aquathermolysis” processes. Indeed, the flooding of heavy oil reservoirs with hot water is not equal to the creation of strictly homogeneous reaction environment, but the miscibility of incompatible oil and water can be improved by means of ionic or nonionic surfactants offered by modern chemical industry. From the above considerations, the aquathermolysis reactions seem to be largely governed by ionic mechanisms, especially at lower temperatures. Nevertheless, the very broad diversity of heavy oil components in chemical composition does not exclude the occurrence of free radical processes as well, which are more intense, the higher the temperature (in a broad sense, these free radical processes are caused by thermal energy provided by steam and rather should be treated as mere pyrolysis reactions not much different in their nature from those considered in the partial combustion methods). From the point of view of selectivity of aquathermolysis processes and prevention of polymerization reactions (coke formation), the lower temperatures are more preferable. However, to run aquathermolysis processes with reasonable rates at low temperatures, the development of efficient catalysts is absolutely necessary. The potential of surfactants in enhancing the aquathermolysis reactions also should not be downplayed. After all, surfactants not only may help to increase the homogeneity of water–oil emulsions, but can also be considered as phase transfer catalysts for delivering reagents (oil components) to the reaction medium (water), in which specifically the ionic aquathermal reactions take place. Indeed, there is no strong division between possible free radical and ionic reactions under the real aquathermal treatment conditions, but it should be kept in mind that the former would prefer, presumably, the oil phase as a reaction medium, while the latter would apparently proceed in an aqueous medium. From the general standpoint, the immiscibility of the organic and aqueous phases implies that all the reactions that follow ionic mechanisms actually must be heterogeneous reactions that take place, to a larger extent, at the interface between organic and aqueous phases (as said above, these kinds of processes can be greatly enhanced by phase transfer catalysts, i.e. surfactants). Nonetheless, we associate the catalytic processes in this case with homogeneous catalysis, since the catalyst itself and reaction intermediates (organic ions) are distributed directly in the reaction medium in which the elementary reactions take place, i.e. in the aqueous medium. As a matter of principle, the ionic reactions can also take place on the surface of solid catalysts insoluble in both of the fluid phases, but the rates of these kind of processes would be limited by diffusion of reagents to the surface (very slow process), unless the reaction is hypothetically catalyzed by the surface of rocks during the filtration of the oil–water mixture through them. In any case, we took an option of homogeneous catalysis in our theoretical treatment of ionic aquathermolysis reactions, since they were aimed at developing the theoretical tools for evaluating the efficiency of injected catalysts. As for the free radical aquathermolysis reactions—which dominate, in our opinion, in the high temperature regions—they are likely be of the homogeneous nature as taking place directly in the oil phase with the minimum presence of water molecules. In this case, homogeneous catalysis also should be a preferable option, since it will allow one to avoid the diffusion control of the kinetics, given that the chemical reactions in the oil reservoir are conducted in the unstirred medium, in contrast to conventional chemical reactors. However, the heterogeneity of catalysis

8.3 Quantum Chemical Approaches to the Calculation of Thermochemical and Kinetic Parameters

in homogeneous free radical reactions is a less-critical issue, especially keeping in mind that some processes that take place at high temperatures—for example, the destruction of aromatic thiophene heterocycles—may require specifically heterogeneous catalysts. Many aromatic heterocycles based on oxygen, nitrogen, and sulfur heteroatoms are incorporated into the polycondensed aromatic systems of asphalthenes and cannot be readily destroyed under aquathermolysis conditions. To date, a large amount of data have been accumulated from laboratory and field experiments devoted to studying the in situ upgrading processes. As already noted above, the aquathermolysis reactions at low temperatures seem to follow ionic mechanisms, and it is natural to consider, first of all, polar heteroatomic bonds when we start to search for likely candidates that need to be destroyed in these processes. Moreover, it is heteroatomic bonds that are generally the weakest links compared to relatively strong C—C, C=C, and C=Carom bonds in a thermodynamically equilibrated chemical system, such as the oil reservoir. Hence, the polarity and lower thermodynamic strength of heteroatomic bonds make it possible to conduct aquathermolysis processes at lower temperatures. In low temperature regimes, the heterolytic cleavage of nonpolar carbon bonds is unlikely, and vice versa, homolytic cleavage of nonpolar C—H and ordinary C—C bonds, and other processes connected with the destruction of stronger aromatic heterocycles should be considered first when we theoretically analyze the thermal degradation processes at higher temperatures. Based on the above general considerations, we have developed some preliminary approaches to the calculations of thermochemistry and kinetics of in situ upgrading processes, which entail the differentiation of free radical and ionic reaction mechanisms, the consideration of catalytic mechanisms, and separate theoretical treatment of homogeneous and heterogeneous catalytic processes.

8.3 Quantum Chemical Approaches to the Calculation of Thermochemical and Kinetic Parameters of In Situ Upgrading Processes 8.3.1

Choice of Model Compounds for Simulating the In Situ Processes

Despite extensive practice of using catalytic aquathermolysis, the theory of degradation of heavy oil components during in situ upgrading is less developed. The first attempts to apply quantum chemical ab initio methods to studying the bond cleavage reaction during aquathermolysis have been made in the Kazan Federal University in an original series of studies by Aminova et al. (Lysogorskii and Aminova 2015; Lysogorskii et al. 2016). Using cyclohexyl phenyl sulfide (CPS) as a model compound, they numerically analyzed the mechanisms of C—S bond cleavage under aquathermolysis conditions on the basis of calculated energy parameters of possible elementary reactions. As can be concluded from their results, the heterolytic cleavage of C—S bonds, i.e. the ionic mechanism of reaction, is a more thermodynamically favorable option. As mentioned above, general considerations imply that heteroatomic bonds should be a subject of study first of all when we theoretically treat in situ upgrading processes at low temperatures, which largely follow ionic mechanisms. Typical heteroatomic bonds can be simulated by the appropriate selection of relatively simple model compounds that can substitute real heavy oil components in quantum chemical calculations. Any crude oil is a continuum of many saturated, unsaturated, and aromatic hydrocarbons, among which heteroatomic organic compounds represent a significant portion. The total identification of oil components is barely possible in practice. The oil composition can be roughly characterized by the contents of main elemental constituents, which are as follows for typical oil: carbon (83.0–87.0%), hydrogen (10.0–14.0%), nitrogen (0.1–2.0%), oxygen (0.05–1.5%), sulfur

385

386

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

(0.05–6.0%), and metals (Ni and V, Co2+ ≈ Ni2+ > Fe2+ . The kinetic patterns of catalytic cleavage of CPS are more clearly illustrated by corresponding kinetic constants calculated using Eq. (8.1). Based on the calculated standard rate constants (Table 8.5), all the investigated transition metal catalytic systems substantially accelerate the reaction rate at 500–600 K. This temperature range is usually achieved in reservoirs during steam injection process for heavy oil recovery, which means that these metals have potential in catalyzing aquathermolysis processes in a real steam injection project in the oilfield. In the case of Cu2+ , the relatively high kinetic constant is obtained even at T = 400 K (see Table 8.5). Figure 8.4 shows the logarithmic temperature dependences of the half-life of the reaction of heterolytic cleavage of CPS bonds for different catalysts, which are determined under the assumption of first-order decay by the following expression: t1/2 = ln2/k∘ . As noted above, activation parameters Δ#G∘ calculated by the DFT/B3LYP method are most likely understated. For this reason, the half-life of the reaction corresponding to t1/2 = 1 hour achieved with this functional seems to be improbable in the temperature range of about 225–300 K for the different catalytic systems, let alone a half-time of t1/2 = 1 second that can be obtained for different transition metals in the temperature range of 260–340 K (see Figure 8.4a). A half-life of 1 hour and 1 seconds with the ωB97X-D functional can be achieved for different catalytic systems in the temperature range of 310–375 K and 350–430 K, respectively (Figure 8.4b). These simulation results obtained with the ωB97X-D functional seem to be more reliable. In comparison to the ωB97X-D functional,

397

Table 8.4 Gibbs free energy changes Δ# G∘ (kJ/mol) of the second rate-limiting step of the reaction sequence given in Scheme 8.5 with different catalysts and at different temperatures for CPS (calculated using the DFT method with corresponding functionals in the GAUSSIAN 09 software package). Fe2+

Co2+

Cu2+

Ni2+

T, K

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

0

152.90

193.71

204.20

146.32

185.65

198.25

114.93

159.59

184.59

147.36

192.38

201.95

100

134.59

173.71

181.58

127.84

168.17

180.97

97.56

140.25

165.44

129.11

174.39

184.19

200

114.13

152.21

160.85

107.01

148.18

162.51

78.76

117.69

143.04

108.24

153.99

163.78

300

93.57

130.97

139.86

86.00

127.80

144.30

60.33

94.74

119.95

87.12

133.40

142.80

400

73.04

110.00

118.81

64.98

107.34

126.37

42.34

71.78

96.53

66.05

112.81

121.57

500

52.59

89.29

97.77

44.03

86.90

108.71

24.79

48.93

72.95

45.02

92.28

100.19

600

32.25

68.84

76.81

23.19

66.54

91.31

7.65

26.26

49.29

24.10

71.85

78.75

700

12.03

48.63

55.95

2.46

46.26

74.17

−9.12

3.79

25.59

3.30

51.53

57.29

Table 8.5 Standard rate constants k∘ (s−1 ) of the second step (Scheme 8.5) at different temperatures for different catalysts in the aquathermolysis of CPS, which are determined using the standard Gibbs energies of activation calculated by the DFT method with different functionals. Fe2+

Co2+

Cu2+

Ni2+

T, K

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

0

0

0

0

0

0

0

0

0

0

100

1 × 10−58

4 × 10−79

3 × 10−83

4 × 10−55

3 × 10−76

6 × 10−83

200

−18

7 × 10

7 × 10

−28

−30

−16

−27

−30

300

3 × 10−4

1 × 10−10

3 × 10−12

400

2 × 103

4 × 10−2

500

3 × 107

600 700

4 × 10

5 × 10

2 × 10−39 1 × 10

−8

8 × 10

2 × 10

7 × 10−3

4 × 10−10

5 × 10−13

2 × 102

3 × 10−3

3 × 104

8 × 10−2

3 × 10−4

5 × 103

6 × 102

3 × 108

9 × 103

2 × 1010

1 × 107

3 × 106

1 × 1011

2 × 1012

3 × 109

1 × 109

1 × 1013

B3LYP

0

𝛚B97X-D

M06-2X

0

0

1 × 10−61

8 × 10−75

8 × 10−56

2 × 10−79

1 × 10−84

−19

−25

−16

−28

7 × 10−31

8 × 10

2 × 10

2 × 10

3 × 10

2 × 10−4

8 × 10−9

4 × 10−3

4 × 10−11

9 × 10−13

3 × 107

4 × 103

2.07

2 × 104

2 × 10−2

1 × 10−3

45.8

3 × 1010

8 × 107

2 × 105

2 × 108

2 × 103

4 × 102

2 × 107

1 × 105

3 × 1012

6 × 1010

6 × 108

1 × 1011

7 × 106

2 × 106

5 × 109

4 × 107

7 × 1013

8 × 1012

2 × 1011

8 × 1012

2 × 109

8 × 108

400

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms lgt1/2, h

lgt1/2, h 30

30 Ni(II) Co(II) Fe(II) Cu(II)

20

10

20

10

1h

0

Ni(II) Co(II) Fe(II) Cu(II)

1h 1s

0 1s –10

–10

–20

–20 100

200

300

400

500

600

100

700 T, K

200

300

400

500

600

700 T, K

(b)

(a) lgt1/2, h 30 Ni(II) Co(II) Fe(II) Cu(II)

20 10 1h

0

1s –10 –20 100

200

300

400

500

600

700 T, K

(c) Figure 8.4 Logarithmic temperature dependences of the half-life of the reaction of heterolytic cleavage of CPS with different catalysts determined from the appropriate standard Gibbs energies calculated using the DFT method with the (a) B3LYP, (b) ωB97X-D, and (c) M06-2X functionals (dashed lines show reaction halftime levels corresponding to t 1/2 = 1 hour and t 1/2 = 1 second).

the M06-2X functional gives somewhat higher values (t1/2 = 1 hour and t1/2 = 1 second in the temperature ranges of 340–410 and 400–470 K, respectively) (see Figure 8.4c). To comprehend the reasons for a substantial difference in the catalytic activity of Cu2+ , a Mulliken population analysis of charge densities on relevant atoms of the prereaction complexes is performed, as well as bond lengths and HOMO–LUMO gaps of these complexes are compared (Table 8.6). The analysis shows that Cu2+ , compared to other studied ions, has significant differences in the charge distribution, M—S bond length, and HOMO–LUMO gap. At the same time, there is no direct relationship between the reference data on the ionic radii and the considered parameters for transition metals under study. As can be seen from Table 8.6, the effect of increased catalytic activity of Cu2+ in comparison with the other metals is determined by a shorter M—S bond length, a substantial distribution of the positive charge to S, and a larger HOMO–LUMO gap in the prereaction complex. The Mulliken population analysis shows a significant positive charge on the S atom and a less negative charge on the C atom of the cyclohexyl ring in the prereaction complex with Cu2+ compared to the other metals, which should facilitate the elimination of the cyclohexyl cation in the cleavage reaction. There is no substantial differences in the C—S bond lengths of the

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

Table 8.6 Ionic radii of metals under study and calculated parameters of prereaction complexes of CPS with these metals (see Scheme 8.5).

M2+

Ionic radius, Å

S—M bond length, Å

S—C bond length, Å

Mulliken atomic charge on S

Mulliken atomic charge on M

Fe2+

0.78

2.624

1.879

0.010

0.943

−0.176

5.712

Co2+

0.745

2.561

1.880

0.010

0.956

−0.194

5.719

Ni

0.69

2.506

1.880

0.017

0.918

−0.199

5.728

Cu2+

0.73

2.420

1.884

0.188

0.689

−0.003

5.882

2+

Mulliken atomic charge, C

HOMO–LUMO gap, eV

Notes: Ionic radii are taken from Shannon 1976; C is the carbon atom of the cyclohexyl ring in the C—S bond; the B3LYP/TZVP method was used in calculations.

prereaction complexes with different metals; however, this bond in the starting CPS was noticeably shorter in length (1.845 Å). The found HOMO–LUMO gaps of the prereaction complexes require further analysis beyond the scope of this study, but allow us to make a preliminary conclusion that lower gaps for complexes with Ni2+ , Co2+ , and Fe2+ give rise to their greater stability through delocalization of the energy when compared to the complex with Cu2+ . The less stability of the prereaction complex with Cu2+ explains its higher reactivity and enhanced catalytic activity of Cu2+ in the cleavage of C—S bonds.

8.4.4 Calculation of Thermodynamic and Kinetic Parameters of Aquathermal Decay of Cyclohexyl Phenyl Ether Table 8.7 shows the changes in the standard electronic energy and Gibbs free energy in each step of the reaction mechanism sequence and in the total reaction (Scheme 8.5) for CPE, which is calculated using the B3LYP/TZVP method. One can see from Table 7.8 that the second step, in which the heterolytic breakage of C—O bonds occurs, is again the most energy-consuming step of the reaction sequence (see Figure 8.5 for the typical energy profile of the reaction sequence). Therefore, a further analysis of the calculated thermochemical parameters of the reaction mechanism was performed in the same way as in the case of CPS. Table 8.8 shows estimates of the standard rate constants for the rate-limiting step of aquathermal cleavage of CPE at room temperature (298.15 K), from which it is clear that activation Energy, kJ/mol

o

+

250 Cat

200

+

Cat

+

CH3

+ H3O

150 100 50

o

Cat

+ Cat

+

CH3

+ H2O Cat +

0 –50

Pre-reaction complex

Figure 8.5

Typical energy profile of the aquathermolysis of CPE.

OH

+

CH3

+ H2O

401

Table 8.7 Changes in the electronic energy and Gibbs free energy calculated by the DFT method at T = 298.15 K with different functionals (GAUSSIAN 09 software) for elementary steps and the total reaction given in Scheme 8.5 in the aquathermolysis of CPE with use of different catalysts. Electronic energy and Gibbs free energy changes (indices correspond to numbers assigned to elementary reaction steps shown in Scheme 8.5), kJ/mol DFT method

Catalyst +

H3 O B3LYP/ TZVP

ωB97X-D/ TZVP

M06-2X/ TZVP

𝚫E 1

𝚫G 1

𝚫E 2

𝚫G 2

301.79

251.33

Fe2+

34.92

86.85

103.63

51.81

Ni2+

34.93

97.85

96.13

37.14

Co2+

37.24

90.06

102.77

46.18

−72.13 −36.95

𝚫E 3

𝚫G 3

𝚫E 4

𝚫G 4

−46.64

−49.70

52.95

50.93

𝚫E 5

Energy changes for total reaction, kJ/mol 𝚫G 5

−115.03 −171.98

3.72

77.85

16.03

−13.45

−56.73

−74.37 −36.33

115.49

53.02

19.91

−8.17

Fe2+

−10.31

48.99

165.41

105.11

Ni2+

−12.59

45.19

155.60

105.56

Co2+

10.78

67.99

137.20

85.73

Cu2+

−72.49 −16.41

133.76

78.79

−0.25

−53.87

H3 O+

−74.54 −35.90

124.61

65.97

20.32

−11.76

173.35

128.19

Ni2+

−17.72

40.02

172.38

118.12

Co2+

−32.00

42.28

172.38

125.17

Cu2+

−72.31 −23.23

145.96

113.34

−35.76

55.30

−2.19

59.45

5.35

−123.97 −173.23

−48.36

41.69

22.35

−122.51 −175.66

Cu2+

−7.38

𝚫G tot

−213.62 −251.38

H3 O+

Fe2+

𝚫E tot

−94.08 −145.59 −56.76

−59.67

51.04

48.97

−81.98 −142.23 −86.96 −145.21

−95.58 −151.58 −48.71

−52.68

37.77

39.73

−84.27 −139.83 −69.99 −149.14 −3.27

−71.81

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

Table 8.8 Theoretical standard rate constants reduced to the 𝜘 factor for the rate-limiting step (see Scheme 8.5 for the reaction sequence) according to the results of calculations for CPE with different functionals at T = 298.15 K.

Catalyst

+

H

Fe2+

2+

Ni

Co2+

2+

Cu

Calculation method

Reduced rate constant k∘ / 𝝒, s−1

B3LYP/TZVP

5.774 × 10−32

ωB97X-D/TZVP

3.200 × 103

M06-2X/TZVP

17.242

B3LYP/TZVP

5.208 × 103

ωB97X-D/TZVP

2.390 × 10−6

M06-2X/TZVP

2.163 × 10−10

B3LYP/TZVP

1.938 × 106

ωB97X-D/TZVP

1.999 × 10−6

M06-2X/TZVP

1.259 × 10−8

B3LYP/TZVP

5.045 × 104

ωB97X-D/TZVP

5.935 × 10−3

M06-2X/TZVP

7.323 × 10−10

B3LYP/TZVP

9.674 × 109

ωB97X-D/TZVP

9.784 × 10−2

M06-2X/TZVP

8.663 × 10−8

energy parameters in the case of transition metals are strongly underestimated with the B3LYP functional. In the calculations with the ωB97X-D and M06-2X functional, the estimated standard rate constants look more realistic in absolute values, especially with the latter functional. On the contrary, the B3LYP/TZVP method gives more reliable results for the proton-catalyzed reaction. As can be seen from Table 8.8, the calculated rate constants of the heterolytic cleavage of CPE at 298.15 K are considerably increased in the case of catalysis with transition metals when compared to the proton-catalyzed reaction. The relative catalytic activities of the transition metals under study can be arranged in the following decreasing orders: for the calculation with the B3LYP functional, Cu2+ > Ni2+ > Co2+ > Fe2+ ; for the calculation with the ωB97X-D functional, Cu2+ ≈ Co2+ > Ni2+ ≈ Fe2+ ; and for the calculation with the M06-2X functional, Cu2+ > Ni2+ ≈ Co2+ ≈ Fe2+ . Thus, the Cu2+ catalyst may show, according to the results of theoretical calculations, superior catalytic activity in comparison with the case of uncatalyzed aquathermolysis and with the other transition metal catalysts under study. Nevertheless, the standard rate constant under conditions of catalysis with Ni2+ is close to that of Cu2+ from the result of simulation with the ωB97X-D functional. The obtained rate constant estimates at 298.15 K under conditions of catalysis with transition metals are still low, especially considering the reducing transmission coefficient 𝜘. Therefore, we have calculated the Gibbs free energies of the rate-limiting second step at different temperatures for different catalytic systems with different DFT functional. According to the results of calculations by the DFT/ωB97X-D and DFT/M06-2X methods, which are more reliable methods for the treatment of systems with transition metal atoms, the Gibbs energy of activation substantially decreases with an increase in the temperature from ambient temperatures (300 K) to 700 K. The dependences of

403

404

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms ln(k/𝜘), s–1

ln(k/𝜘), s–1 DFT/B3LYP

50

DFT/ωB97X-D

50

0

0 0.002

0.004

0.006

0.008

0.01

1/T, K–1

0.002

–50

0.004

0.006

0.008

0.01

1/T, K–1

–50 Fe(II) Cu(II)

–100

Ni(II) Co(II)

–150 –200

Fe(II) Cu(II)

–100

Ni(II) Co(II)

–150 –200

(a)

(b) ln(k/𝜘), s–1 50

DFT/M06-2X

0 0.002

0.004

0.006

0.008

0.01

1/T, K–1

–50 Fe(II)

–100

Cu(II) Ni(II)

–150

Co(II)

–200

(c) Figure 8.6 Logarithmic temperature dependences of the standard rate constants k∘ (s−1 ) reduced to the 𝜘 factor calculated for CPE in the rate-limiting step approximation for different catalysts with use of the appropriate standard Gibbs energies calculated by the DFT method with the (a) B3LYP, (b) ωB97X-D, and (c) M06-2X functionals.

the logarithms of the calculated rate constants on the inverse temperature with different DFT methods more clearly illustrate the kinetic patterns of catalytic cleavage of CPE at different temperatures (Figure 8.6), which follow well the classic van Hoff–Arrhenius law. At lower temperatures, the calculated rate constants are generally different for different transition metal ions, but the differences in the case of calculation with the DFT/ωB97X-D and DFT/M06-2X methods become insignificant upon reaching 400 K and higher temperatures. Thus, the corresponding kinetic constants calculated by Eq. (8.1) with the use of the DFT/ ωB97X-D and DFT/M06-2X methods show that all investigated transition metal catalytic systems substantially accelerate the reaction rate at 500–600 K. This temperature range is a normal operating range when performing steam injection in reservoirs with heavy oil, which means that the studied transition metal ions are potentially capable of catalyzing aquathermolysis processes in real oilfields. Catalytic systems based on nickel (Vakhin et al. 2021), copper (Chao et al. 2012), iron (Chen et al. 2008), and molybdenum (Wen et al. 2007) were tested during steam injection treatments of different heavy oil reservoirs and proved their efficiency for enhancing the well productivity. Table 8.9 shows the reaction half-lives for heterolytic cleavage of C—O bonds of CPE with different catalysts, which are determined under limiting-rate approximation from the Gibbs energies of the second rate-limiting step with use of different DFT functionals. As noted above, activation parameters Δ#G∘ calculated by the DFT/B3LYP method are substantially underestimated for the transition metals. Therefore, the half-lives of the reaction corresponding to this functional are highly improbable, since they are in the range of 10−6 –10−13 minutes for different catalytic systems, i.e. correspond to very short reaction times even at such a low temperature as 300 K. On the other

Table 8.9 Theoretical half reaction times (min) multiplied by the 𝜘 factor according to the results of calculation of rate constants by Formula (8.1) in the rate-limiting step approximation for the aquathermolysis of CPE at different temperatures with different catalysts. Fe2+

Co2+

Cu2+

Ni2+

T, K

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

B3LYP

𝛚B97X-D

M06-2X

0

∞

∞

∞

∞

∞

∞

∞

∞

∞

∞

∞

∞

100

7.2 × 1029

2.2 × 1060

4.2 × 1067

3.3 × 1028

1.9 × 1047

7.5 × 1066

2.0 × 1014

9.2 × 1044

1.7 × 1056

3.0 × 1024

3.0 × 1057

3.2 × 1065

200

2.2 × 103

2.6 × 1018

5.4 × 1022

187.6

1.1 × 1012

1.1 × 1022

6.2 × 10−6

4.4 × 1010

6.9 × 1017

1.28

1.7 × 1017

7.5 × 1020

300

1.7 × 10−6

4.0 × 103

3.4 × 107

1.8 × 10−7

1.374

1.0 × 107

9.7 × 10−13

0.084

9.1 × 104

4.7 × 10−9

3.9 × 103

5.9 × 105

400

3.6 × 10−11

9.8 × 10−5

0.275

4.1 × 10−12

9.7 × 10−7

0.227

2.9 × 10−16

8.7 × 10−8

3.0 × 10−2

2.1 × 10−13

4.6 × 10−4

1.3 × 10−2

500

4.9 × 10−14

2.1 × 10−9

1.5 × 10−5

5.8 × 10−15

1.6 × 10−10

5.2 × 10−6

1.9 × 10−18

1.9 × 10−11

3.8 × 10−6

4.3 × 10−16

2.8 × 10−8

2.8 × 10−7

600

5.5 × 10−16

1.4 × 10−12

1.1 × 10−8

6.7 × 10−17

4.2 × 10−13

4.0 × 10−9

6.0 × 10−20

6.3 × 10−14

9.2 × 10−9

6.4 × 10−18

3.9 × 10−11

2.0 × 10−10

700

2.1 × 10−17

6.7 × 10−15

5.8 × 10−11

2.6 × 10−18

5.5 × 10−15

2.3 × 10−11

4.8 × 10−21

1.0 × 10−15

1.2 × 10−10

3.0 × 10−19

3.4 × 10−13

1.1 × 10−12

406

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

hand, the ωB97X-D functional and, especially, the M06-2X functional give plausible values for the reaction half-lives with different catalysts in the temperature range of 300–400 K. Nevertheless, all these DFT methods allow one to conduct a comparative analysis of the catalytic activity of different metals in the process of CPE aquathermolysis, according to which the studied metals can be arranged in the following order: Cu2+ > Ni2+ > Co2+ > Fe2+ . As can be seen from Figure 8.6 and Table 8.9, the catalytic system based on Cu2+ substantially outperforms the other catalytic systems. The observed substantial difference in the catalytic activity of Cu2+ can be explained based on a Mulliken population analysis of charge densities distributed over relevant atoms of the prereaction complexes, and the corresponding bond lengths and HOMO–LUMO gaps (Table 8.10). One can see that there are substantial differences in the charge distribution, HOMO–LUMO gap, and M—O bond length for Cu2+ compared to other ions. However, the calculated parameters of prereaction complexes for transition metals under study are not directly correlated to the published data on the ionic radii (Shannon 1976). As can be seen from Table 8.10, the manifestation of increased catalytic activity by Cu2+ compared to that of the other metals can be explained by a shorter M—O bond length, a longer O—C bond length, and stronger distribution of the negative charge to C. The Mulliken population analysis shows a considerable negative charge on the C atom of the cyclohexyl ring in the prereaction complex of Cu2+ and a lower positive charge on the M atom compared to that of the other metals under study. An analysis of the energies of frontier orbitals in the prereaction complexes shows that the HOMO–LUMO gap values correlate with the catalytic activities of the corresponding metal ions. As can be seen from Table 8.10, the lowest HOMO–LUMO gap in the case of Cu2+ corresponds to the highest catalytic activity of Cu2+ ions in the cleavage of C—O bonds. The order in the catalytic activity series precisely corresponds to the order in which they can be arranged according to increasing HOMO–LUMO gap values. The found sequence in the catalytic activity series to some extent can be explained by the fact that electronic excitation, including thermally induced excitation of the pre-reaction complexes, proceeds easier with low HOMO–LUMO gap values. In turn, the elimination of a cyclohexyl carbocation is a rate-limiting step in the reaction of CPE aquathermolysis, and the carbocation is more likely to be eliminated from the excited states. However, specific transition states in the reaction of carbocation elimination basically cannot be located by the DFT methods through PES scanning. This makes it difficult to give a more detailed explanation for the found dependence.

Table 8.10 Ionic radii of metals under study and some parameters of prereaction complexes of CPE with these metals calculated (see Scheme 8.5). HOMO–LUMO gap value in pre-reaction complex, eV

M2+

Ionic radiusa), nm

O—M bond length, nm

O—Cb) bond length, nm

Fe2+

0.078

0.2144

0.1463

−0.451

1.191

0.011

0.38533

0.35309

Co2+

0.0745

0.2130

0.1467

−0.444

1.135

−0.006

0.38012

0.33317

Ni2+

0.069

0.2113

0.1468

−0.430

1.071

0.000

0.37911

0.32714

Cu2+

0.073

0.1990

0.1482

−0.451

0.951

−0.114

0.37762

0.30940

Mulliken atomic charge on O

Mulliken atomic charge on M

Notes: a) Ionic radii are taken from Shannon 1976. b) C is the carbon atom of the cyclohexyl ring in the C—O bond.

Mulliken atomic charge on C

𝛚B97X-D

M06-2X

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

8.4.5 Calculation of Thermodynamic and Kinetic Parameters of Aquathermal Decay of Cyclohexyl Phenyl Amine Table 8.11 shows the calculated values of the electronic energy and Gibbs free energy changes for each step of the reaction sequence given in Scheme 8.5 and for the total reaction of aquathermal decay of CPA. Different catalysts lower to different extents the energy of the second step which limits the kinetic rate of the total reaction. However, copper(II) stands out and sets apart from other catalysts by a large positive value of the Gibbs free energy of the fifth step associated with the elimination of final product aniline from the catalytic complex. Figure 8.7 clearly shows this difference in the behavior of the copper(II) catalyst. This means that the strength of aniline coordination with copper(II) is so high that the catalyst cannot be regenerated in the final step of the reaction sequence, i.e. the self-poisoning of the catalyst by the reaction product occurs. Hence, copper(II) cannot be considered as a good working catalyst for CPA aquathermolysis, and further kinetic calculations and comparative analysis are performed only for H3 O+ , Co(II), Ni(II), and Fe(II). In DFT calculations involving transition metals, it should be noted that functionals specially developed for studying the kinetics, thermodynamics, and properties of transition metal compounds are more preferred to use instead of the B3LYP functional that gives excellent results with purely organic compounds and poorer results with metal complexes. In addition to the B3LYP functional, the ωB97X-D (Chai and Head-Gordon 2008) and M06-2X (Zhao and Truhlar 2008) functionals were used to calculate the energy parameters of the reaction participants. Table 8.12 shows the calculated rate constants of CPA aquathermolysis by different DFT methods for the H3 O+ , Co(II), Ni(II), and Fe(II) catalysts. As can be seen from Table 8.12, the standard rate constants of CPA aquathermolysis at room temperature in the case of calculations with the B3LYP functional are substantially higher than in the case of calculations with the ωB97X-D and M06-2X functionals. In contrast with aquathermolysis of organic sulfides (Al-Hajji et al. 2008) and ethers (Yamamoto et al. 1991), the catalytic activity of

ΔG, kJ/mol 400 3

300 200

2

4 5

100 CPA 0

1

Products

–100 –200 –300 –400

noncatalytic Fe(II) B3LYP Co(II) B3LYP Ni(II) B3LYP H3O+ B3LYP Cu(II) B3LYP

–500 Figure 8.7 Free energy profiles of the total reaction of CPA aquathermolysis with different catalysts according to DFT calculations with the B3LYP functional for the sequence of elementary reactions given in Scheme 8.5.

407

Table 8.11 Electronic energy and Gibbs free energy changes calculated by the DFT/TZVP method with use of different functionals in the GAUSSIAN 09 software package for CPA aquathermolysis under standard conditions (T = 298.15 K) with different catalysts. Electronic energy and Gibbs free energy changes of elementary reaction steps (Scheme 8.5), kJ/mol DFT functional

Catalyst

H3 O+ B3LYP

ωB97X-D

M06-2X

𝚫E 1

𝚫G 1

−151.67 −117.52

𝚫E 2

𝚫G 2

168.50

110.02

Fe2+

−60.74

−25.21

180.18

127.21

Ni2+

−107.29

−71.03

152.27

99.63

Co2+

−76.20

−34.29

156.24

𝚫E 3

𝚫G 3

𝚫E 4

Energy of total reaction, kJ/mol 𝚫G 4

𝚫E 5

3.23

𝚫G 5

50.93

−58.53

−59.98

−89.45

−46.64

−49.69

Cu2+

−435.48 −403.28

51.81

−3.75

403.72

377.11

H3 O+

−159.23 −127.45

226.82

170.42

6.98

−20.13

Fe2+

−45.54

−11.76

288.89

235.15

Ni2+

−92.94

−56.85

268.13

213.60

Co2+

−68.27

−29.65

225.83

164.36

Cu2+

−402.40 −376.24

111.78

60.41

365.19

338.67

H3 O+

−151.27 −113.99

221.15

160.28

6.24

−22.07

−53.36

−15.92

243.24

178.49

−31.72

4.87

188.67

129.81

Co2+

−46.78

−11.85

211.99

154.69

−347.17 −320.74

119.91

67.49

Cu2+

26.37

−28.64

68.85

12.14

71.29

13.93

−99.38 −131.93 52.95

93.81

Fe2+

𝚫G tot

−22.43

−24.92

Ni2+

𝚫E tot

−168.78 −200.55 −56.76

−59.67

51.04

48.97

−100.62 −133.91 −82.98 −111.87

−113.75 −138.35 −44.56

−48.07

37.77

39.73

−80.83 −110.45 −89.08 −118.62 303.38

277.47

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

Table 8.12 Standard rate constants (T = 298.15 K) calculated in the rate-limiting step approximation by Formula (8.1) for the reaction of CPA aquathermolysis with different catalysts (the rate constant values normalized to the 𝜘 factor are given; standard Gibbs free energies Δ# G∘ of the rate-limiting second step of the reaction sequence given in Scheme 8.5 are calculated using different DFT functionals).

Catalyst

+

H3 O

Fe2+

2+

Ni

Co2+

DFT method

Reduced rate constant k∘ / 𝝒, s−1

B3LYP/TZVP

3.30 × 10−7

ωB97X-D/TZVP

8.66 × 10−18

M06-2X/TZVP

5.18 × 10−16

B3LYP/TZVP

3.21 × 10−10

ωB97X-D/TZVP

3.95 × 10−29

M06-2X/TZVP

3.34 × 10−19

B3LYP/TZVP

2.18 × 10−5

ωB97X-D/TZVP

2.36 × 10−25

M06-2X/TZVP

1.13 × 10−10

B3LYP/TZVP

2.28 × 10−4

ωB97X-D/TZVP

9.98 × 10−17

M06-2X/TZVP

4.92 × 10−15

the hydroxonium ion in CPA aquathermolysis is not lower and comparable to the catalytic activities of transition metal ions, that is, CPA aquathermolysis can be catalyzed even with hydroxonium ions from water. In the case of DFT calculations performed with the use of the B3LYP functional, the relative catalytic activities of the studied catalytic systems are as follows: Co2+ > Ni2+ > H3 O+ > Fe2+ . The catalytic activity order in the case of using the ωB97X-D functional is Co2+ > H3 O+ > Ni2+ Fe2+ . The M06-2X functional gives a sequence of Ni2+ > Co2+ > H3 O+ > Fe2+ . Despite the fact that DFT calculations show that CPA aquathermolysis may not require transition metal catalysts, the latter are still useful for cleavage of C—N heteroatomic bonds during complex in situ treatment of oil reservoirs when they are added with the purpose of breaking other heteroatomic bonds. The free energy of the total reaction of CPA aquathermolysis at 298.15 K is near zero (slightly negative or slightly positive, depending on the DFT functional used for the thermochemistry calculation). For further analysis of the reaction kinetics, the Gibbs free energies of the rate-limiting step, i.e. the second step in Scheme 8.5, and the corresponding reaction rate constants at different temperatures in the range of 2–700 K were calculated using different DFT functionals. The DFT calculation results show that not only the Gibbs free energy of the total reaction (see Scheme 8.5) but also the Gibbs energy of the rate-limiting step (which is considered an equivalent of the Gibbs energy of activation) under conditions of catalysis substantially decreases with an increase in the temperature from ambient temperatures (300 K) to 700 K. Figure 8.8 shows the logarithmic dependences of the rate constants, which are calculated by Eq. (8.1) based on the activation of Gibbs free energy determined by different DFT functional, on the inverse temperature. The dependences strongly follow the classic van Hoff–Arrhenius law. As can be seen from Figure 8.8, the calculated rate constants for different transition metals by the DFT/B3LYP differ insignificantly from each other, especially at high temperatures (low 1/T values).

409

410

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms ln(k/𝜘), s–1

ln(k/𝜘), s–1 B3LYP

100

100

50

50

0 0.002

0.004

0.006

0.008

–50

0.002

0.004

–50

0.006

0.008

0.01 1/T, K–1

–100 –150

Fe(II)

–200

Co(II)

–250

Ni(II)

–300

0

0.01 1/T, K–1

–100 –150

WB97XD

Fe(II) Co(II)

–200 –250

H3O+

Ni(II) H3O+

–300 ln(k/𝜘), s–1

MO62X

100 50 0 0.002

–50

0.004

0.006

0.008

0.01 1/T, K–1

–100 –150 –200

Fe(II) Co(II) Ni(II)

–250 –300

H3O+

Figure 8.8 Dependences of standard rate constant logarithms ln(k∘ /𝜘) on the inverse temperature for CPA aquathermolysis with different catalysts. The calculations were performed on the basis of data obtained by the DFT method with the B3LYP, ωB97X-D, and M06-2X functionals.

Based on the DFT/B3LYP method, Co(II) and Ni(II) exhibit practically the same catalytic activities. The ωB97X-D and M06-2X functionals give the controversial data for Co(II) and Ni(II) even though both of them have shown themselves as reliable functionals for the thermochemistry and kinetics calculations with transition metal compounds. Thus, the DFT theory does not allow one to make an unambiguous conclusion about the preference of one of these transition metal catalysts. According to the calculation results, all the studied catalytic systems excluding Cu(II) ions start to show rather high kinetic constants (actually, one should consider uncertain reducing factor 𝜘). The typical temperature range of in situ treatment of underground oil reservoirs is around 500–600 K, i.e. acids and transition metal ions Ni(II), Co(II), and Fe(II) can serve as potential catalysts for cleavage of C—N bonds in amines during in situ aquathermal treatment of heavy oil. The following sequence of catalytic activity of the studied catalytic systems in the reaction of CPA aquathermolysis can be arranged in the decreasing order: Ni2+ ≈ Co2+ > H3 O+ > Fe2+ . Table 8.13 shows the natural bond orbital (NBO) isosurfaces of frontier orbitals and some structural parameters for prereaction complexes with different transition metal ions. The NBO isosurfaces show that substantial mixing of d orbitals of transition metal ions with p orbitals of nitrogen occur in LUMOs, thereby pointing to possibility of effective donor–acceptor interaction between metals and nitrogen. The CNC angle in CPA and the bond lengths related to nitrogen atom also substantially change with the formation of prereaction complexes. It can be noted that Ni2+ and

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

Table 8.13 Natural bond orbital (NBO) isosurfaces of frontier orbitals with isovalues 0.02 and some structural parameters for prereaction complexes of CPA with different transition metal ions as a catalyst (simulated by the M06-2X/TZVP method, polarizable continuum model, solvent water). Transition metal ion in prereaction complex Structural parameter

2+

Fe

Co2+

Cu2+

Ni2+

114.258∘ (122.323∘ )a)

112.789∘

124.258∘

112.911∘

HOMO

LUMO

Angle CNC N—Calif bond length, Å

1.525 77 (1.458 04)

1.513 23

1.485 90

1.513 80

N—Carom bond length, Å

1.463 55 (1.386 88)

1.456 74

1.358 74

1.458 04

N—H bond length, Å

1.021 74 (1.011 84)

1.022 61

1.023 34

1.022 54

N—M bond length, Å

2.141 39

2.115 79

2.233 36

2.071 93

a) Structural parameters for initial CPA are given in parentheses.

Co2+ bring the largest disturbances to the CNC angle of CPA by reducing it, which agrees with their higher catalytic activity. Being a less-active catalyst, Fe2+ decreases the CNC angle to a lesser extent. In contrast to the counterparts, Cu2+ slightly increases the CNC angle when compared to that in the initial CPA. The N—M bond lengths are also arranged in the same order as catalytic activities, i.e. Ni2+ and Co2+ are characterized by the shortest N—–M lengths, the Cu—N bond is the longest one, and the Fe—N bond length is intermediate between them. The N—Calif and N—Carom bond lengths do not directly correlate with the catalytic activities of the metal ions. However, it should be noted that Cu2+ again barely disturbs aforesaid bond lengths in comparison with those in the initial CPA. There are substantial differences in the distribution of charge on the N and M atoms in the prereaction complex of Cu2+ with CPA when comparing it with prereaction complexes of the other transition metals under study (Table 8.14). A less negative charge is distributed on the N atom in the former case, which should facilitate the elimination of the cyclohexyl carbocation because of a weaker electrostatic attraction force. At the same time, a substantially lower charge separation between the M and N atoms in the prereaction complex of Cu2+ with CPA indicate a more covalent nature of the M—N bond. This explains the high stability of the copper complexes with reaction product aniline, which is substantially higher than the stability of the complexes with the other metals under study. The HOMO–LUMO gaps are also in line with the relative stabilities of the prereaction complexes of the studied metal ions with CPA; the smaller the HOMO–LUMO gap, the better the delocalization of the electronic energy in the complexes. According to the obtained calculation results, the behavior of the amine of this study in the aquathermolysis processes is slightly different compared to previously studied model heteroatomic compounds. First, the coordination ability of the substrate to the catalytic site strongly varies depending on the chemical nature of the heteroatom. In particular, copper(II) was the most effective catalyst for cleaving C—S and C—O bonds among the other studied transition metal ions.

411

412

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

Table 8.14 Ionic radii of the transition metals under study and the calculated characteristics of C—N and N—M bonds in prereaction complexes of CPA with these metals.

M2+

Ionic radius, Å

Fe2+

0.78

−0.539

1.677

−0.100

0.150

0.745

−0.444

1.622

−0.059

0.125

0.69

−0.487

1.545

−0.103

0.097

0.73

−0.337

0.868

−0.123

0.071

Co

2+

Ni2+ 2+

Cu

Mulliken atomic charge on N

Mulliken atomic charge on M

Mulliken atomic charge on C

HOMO–LUMO gap, eV

Notes: Ionic radii are taken from the published source Shannon 1976; the C—N bond is given for carbon atom C of the cyclohexyl ring; the calculations were performed using the M06-2X/TZVP method.

In the case of CPA, copper(II) is shown to be the least efficient catalysts due to the self-poisoning by the amine. Second, this study has shown the similarity in the catalytic mechanisms of aquathermal cleavage of C—N bonds in the presence of transition metal ions with the mechanisms of cleavage of C—O and C—S bonds on the quantitative basis. Studies of transition metal catalysts on the aquathermal cleavage of different heteroatomic bonds with use of model compounds are important for effective selection and molecular design of catalysts for different types of heavy oil with different contents of C—N, C—O, and C—S bonds. As a concluding remark, this theoretical study shows that the benefits of using transition metal catalysts in the case of amines is uncertain, since even protons that are always present in water can potentially catalyze the cleavage of amines to no lesser extent than transition metals do. On the contrary, transition metal catalysts can certainly be recommended for in situ aquathermal cleavage of C—S and C—O heteroatomic links, since they show superior catalytic activity compared to the proton catalyst.

8.4.6

Aquathermolysis Reactions of Dibenzyl Sulfide Under Conditions of Pyrolysis

According to the GC–MS data in the autoclave experiments, the heating of the mixture of water with dibenzyl sulfide in an autoclave to 573 K under a pressure of 72 bar leads to the formation of dibenzyl (DB) and trans-stilbene (SB) as dominant products in 44% and 16% yields, respectively. Substantial amounts (up to 10%) of tetrabenzyl thiophene and about 3.5% of 3-phenyl benzo[b]thiophene can also be formed. The last two compounds may be the result of reaction paths by the radical mechanisms, which are highly likely at such high reaction temperatures. The decomposition of dibenzyl sulfide with formation of DB and SB (desulfurization) not only may involve water molecules but can also proceed without direct participation of water molecules when considering the possible reaction paths. In theoretical studies of aquathermal degradation of DBS by the quantum chemical calculation methods, the conformational analysis of dibenzyl sulfide was performed first. According to the conformational analysis, benzyl substituents rotate relatively easily around the C—S bond. The maximum energy barrier of rotation does not exceed 10 kJ/mol, which equals about 4RT. Considering the high reaction temperatures, it can be stated that the dibenzyl sulfide molecule has a high conformational mobility under conditions of autoclave reaction. To simulate the reaction paths, we chose the most stable conformer corresponding to the minimum value of the total electron energy. Figure 8.9 shows the possible reaction path taken for calculations with the formation of trans-stilbene from dibenzyl sulfide in the absence of a catalyst under conditions closest to those used in an autoclave (P = 72 bar, T = 573 K). The effect of a reaction medium was taken into

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

500 ΔGex, ΔH (kJ/mol) 400

TS3*

300 TS1*

100

230

TS3

200

TS1

0

345

TS2*

208

Figure 8.9 Calculated pathway of the formation of trans-stilbene from dibenzyl sulfide without involving the water molecule under conditions simulating those used in the experiment (P = 72 bar, T = 573 K). The main energy barriers of the intermediate stages are shown by arrows.

TS2 P

R

H2S+SB

Reaction pathway

account using the IEFPCM polarizable continuum model with the values of parameters typical for aqueous media. The calculated reaction pathway is characterized by the presence of three main activation energy barriers. The first two barriers relate to the sequential transfer of two hydrogen atoms from the CH2 group to the sulfur atom and are 208 and 230 kJ/mol for the transfer of the first and second hydrogen atom, respectively. A third energy barrier of 345 kJ/mol characterizes the transition from the TS2 to TS3* state and the subsequent elimination of the hydrogen sulfide molecule. The dissociation energy of the intermolecular complex of trans-stilbene with hydrogen sulfide is −77 kJ/mol. The entropy factor has a weak effect on each of the transformation stages; the total reaction has a weak endothermic effect. Table 8.15 shows the main activation barrier values depending on the simulation conditions, which make it possible to establish the influence of temperature and pressure, as well as the effect of a medium on the activation energies. As can be seen from Table 8.15, the temperature and pressure barely influence the heights of all three main activation barriers of the reaction. As a rule, the dielectric continuum stabilizes the transition states; in this case, it can be stated that the medium hardly has an effect on the energy barrier heights. Figure 8.10 shows the pathway of aquathermolysis of dibenzyl sulfide to dibenzyl, including the oxidation of sulfide to sulfoxide and sulfone.

Table 8.15 Effects of the temperature, pressure, and reaction medium on the main activation barriers of the reaction of conversion of dibenzyl sulfide to trans-stilbene (the Δ# H∘ and Δ# G∘ values in kJ/mol and reaction rate constant values in s−1 are given in respective order through slashes). Conditions

E [R–TS1*], kJ/mol

E [TS1–TS2*], kJ/mol

E [TS2–TS3*], kJ/mol

T = 300 K, P = 1 bar, reactant molecules are simulated in vacuum

217/211/1.4 × 10−24

243/243/3.8 × 10−30

331/336/2.7 × 10−46

T = 573 K, P = 72 bar, reactant molecules are simulated in vacuum

218/206/1.9 × 10−23

242/243/7.3 × 10−30

333/339/1.5 × 10−46

T = 573 K, P = 72 bar, reactant molecules are simulated using IEFPCM in an aqueous medium

221/208/8.8 × 10−24

214/230/1.3 × 10−27

365/345/1.4 × 10−47

413

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

500 400

TS3*

300

TS2* 415

TS1*

200

283

100 0

Figure 8.10 Calculated reaction paths of dibenzyl formation from dibenzyl sulfide through the oxidation stage with the direct involvement of water molecules under conditions simulating the experimental ones (P = 72 bar, T = 573 K). The main energy barriers of the intermediate stages are shown by arrows.

ΔGex, ΔH (kJ/mol)

218

414

TS3 TS1

TS4

R1 H2O+DBS

R2

TS2 DB+2H2+SO2

Reaction pathway

At the first stage of the aquathermolysis process, the formation of a complex of dibenzyl sulfide with a water molecule occurs, which is characterized by a Gibbs energy of 39 kJ/mol. The first barrier equaled to 218 kJ/mol corresponds to the reaction of sulfoxide formation. If we assume that the reaction stops at this stage, then the energy of desorption of the resulting hydrogen molecule is −29 kJ/mol, i.e. proceeds spontaneously. Next, the second water molecule is added with a sorption energy of −85 kJ/mol. The second activation barrier equal to 283 kJ/mol corresponds to further oxidation of the sulfoxide to sulfone. Successive removal of two hydrogen molecules from the sulfone molecule (TS2) is accompanied by energy releases of −2 and −58 kJ/mol, respectively. Thus, the resulting hydrogen is easily removed from both the sulfone and sulfoxide molecules. Similar to the case of the reaction of stilbene formation, the highest activation barrier equal to 415 kJ/mol is observed for the last stage. In this case, it is associated with the elimination of a sulfur dioxide molecule and the formation of dibenzyl. The removal of the resulting gases (hydrogen and sulfur dioxide) proceeds with a decrease in energy by −115 kJ/mol. In contrast to the reaction path shown in Figure 8.9, the entropy factor plays an important role in the scheme given in Figure 8.10. It should be noted that we are dealing with excess entropy, since the reactions takes place in the condensed phase. Table 8.16 shows the main activation energy barriers depending on the simulation conditions. In contrast to the reaction leading to trans-stilbene formation, a substantial increase in the activation barrier corresponding to the TS1–TS2* transition is observed with an increase in the temperature. Consideration of the medium in the calculations also leads to an increase in the activation barrier Table 8.16 Effects of the temperature, pressure, and reaction medium on the main activation barriers of the reaction of conversion of dibenzyl sulfide to dibenzyl (the Δ# H∘ and Δ# G∘ values in kJ/mol and reaction rate constant values in s−1 are given in respective order through slashes). Conditions

E[R–TS1*], kJ/mol

E[TS1–TS2*], kJ/mol

E[TS2–TS3*], kJ/mol

T = 300 K, P = 1 bar, reactant molecules are simulated in vacuum

189/193/1.8 × 10−21

164/100/2.6 × 10−5

385/401/1.4 × 10−57

T = 573 K, P = 71 bar, reactant molecules are simulated in vacuum

188/196/1.1 × 10−21

66/247/1.5 × 10−30

384/415/9.6 × 10−60

T = 573 K, P = 71 bar, reactant molecules are simulated using IEFPCM in an aqueous medium

196/218/1.6 × 10−25

105/283/8.2 × 10−37

406/415/9.6 × 10−60

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

of the process. However, this does not lead to conceptual changes in the model, since the limiting stage of the process is determined by the TS2–TS3* barrier. It can be assumed that both considered processes proceed simultaneously under conditions of autoclave experiments. Despite the fact that the rate-limiting activation barrier for the formation of dibenzyl is higher than that for the formation of trans-stilbene by about 60 kJ/mol, the experimental GC–MS data indicate that dibenzyl is a dominant product in the mixture. It is likely that the processes of direct cleavage of C—S bonds with formation of intermediate carbocations can be involved in the aquathermal degradation of DBS, which require further quantum chemical studies.

8.4.7 Molecular Structure of Metal Stearate Catalysts and Simulation of Their Supramolecular Arrangement by MD Methods Metal stearates, especially those including aluminum, copper, cobalt, zinc, calcium, and magnesium, have been used since long time in different industries such as lubricants, acid scavengers, catalysts, stabilizers, drying agents for paints, finishing coatings, cosmetic ingredients (Lower 1981, 1982a,b, 1990a,b,c; Arend 1961). As was shown in recent studies (Yuan et al. 2018, 2021; Varfolomeev et al. 2021), copper(II) stearate can serve as an efficient catalyst for the oxidative upgrading of heavy oil in the in situ combustion (ISC) processes. Other metal stearates—in particular, nickel stearate and iron stearate—had an insignificant catalytic effect on the oxidation of heavy oil (Yuan et al. 2018). As was found from the accelerated rate calorimetry (ARC) experiments, the presence of copper(II) stearate substantially reduces the induction period and the ignition temperature of the combustion process in heavy oil (Yuan et al. 2021). Experiments in the porous medium thermoeffect cell and combustion tube with copper(II) stearate indicated a great potential of this catalyst for its application for in situ upgrading of heavy oil deposits in real fields (Varfolomeev et al. 2021). In contrast to the injection of transition metal ionic salts, the good solubility of metal stearates in oil fractions allows one to better distribute the catalyst over heavy oil deposits. Under the conditions of in situ combustion or other thermal methods of heavy oil processing, metal stearates are capable of forming metal oxide nanoparticles as a result of hydrolysis and thermal degradation. Thermal degradation of metal stearates makes it possible to effectively synthesize finely dispersed metal oxide nanoparticles (Dou and Ng 2016). This opens the way to use metal stearates for in situ catalysis of thermal reactions in real fields through the formation of metal oxide nanoparticles that are well distributed over the oil reservoir. For example, mixed iron oxide nanoparticles synthesized in situ from the iron(III) acetylacetonate complex are shown to be efficient catalysts for in situ upgrading of heavy oil (Mehrabi-Kalajahi et al. 2020; Morelos-Santos et al. 2022). Stabilized suspensions of nickel nanoparticles showed an exceptional performance in reducing viscosity of heavy oil fractions in the aquathermal processes taking place during the in situ heavy oil upgrading (Shokrlu and Babadagli 2013). Comparison of the catalytic activities of nickel nanoparticles and iron oxide nanoparticles showed better performance of the former in aquathermolysis reactions (Yi et al. 2020). The exact chemical composition and chemical structure of metal stearates in the absence of donor ligands other than water is not always determined with certainty, since water molecules can serve as ligands competing for the metal coordination with the carboxylic acid group. Furthermore, the carboxylate groups can serve either as a monodentate or bidentate ligand, and the bidentate carboxylate can be either a chelating or bridging ligand. In some cases, several structures with different chemical compositions may present in the same sample depending on the water content. The solubility, distribution, and subsequent thermal degradation of transition metal stearates in oil fractions under conditions of in situ upgrading may greatly vary depending on the structure of the starting

415

416

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

material. Realizing that the efficiency of using metal stearates as catalysts for in situ upgrading of heavy oil to a certain extent depends on their structural organization, we have performed molecular mechanics and molecular dynamics simulations of copper, nickel, cobalt, and iron stearates that are potential catalysts for in situ upgrading of heavy oil in various thermal degradation processes. The simulation results are intended to provide preliminary knowledge about molecular arrangement of these stearates, their stability at different temperatures, and possibility of solubilization in heavy oil fractions. The model structures in this study were constructed largely based on published data on the structure of transition metal stearates (Catterick and Thomton 1977; Nelson and Taylor 2014). Generally, the structure of the coordination site in transition metal stearates presumably repeats the structure of the homologous transition metal acetates. In accordance with this conjecture, the coordination site of copper(II) stearate should have a binuclear structure with four bridging carboxylates and two water molecules in the square pyramidal apical position (the so-called “Chinese lantern structure”). This kind of structure for copper(II) acetate monohydrate has been established long time ago (Van Niekerk and Schoening 1953a). However, copper stearate binuclear species under anhydrous conditions may form polymeric structures with a coordination of oxygen atoms of neighboring molecules in the apical position (Nordin et al. 2015). Using the same kind of analogy with nickel(II) acetate tetrahydrate (Van Niekerk and Schoening 1953b), it can be concluded that nickel(II) stearate with sufficient access to water molecules should form a complex with four coordinated water molecules and two stearate ions with monodentate coordination. This kind of mononuclear complex is additionally stabilized by hydrogen bonding between coordinated water molecules and carbonyl oxygen of the carboxylate group. According to the X-ray structure of cobalt(II) acetate tetrahydrate (Van Niekerk and Schoening 1953b), this compound is isostructural to nickel(II) acetate tetrahydrate with monodentate coordination of the carboxylate group. However, there is a strong evidence that cobalt(II) acetate with the less number of coordinated water molecules, i.e. Co(OAc)2 •2H2 O, is an infinite coordination polymer linked with carboxylate bridges (Zhang et al. 2010). Furthermore, the liquid crystalline behavior was described in the literature for the anhydrous cobalt(II) stearate (Van Hecke et al. 2003). The mesomorphic phase of either the smectic A or nematic type exists in the temperature range from 380.9 to 400.4 K. As was noted above, metal stearates can serve as versatile precursors for the synthesis of metal oxide nanoparticles (Dou and Ng 2016). Iron stearate structures for nanoparticles design are synthesized and investigated in detail in a recently published study of researchers from Strasbourg University (Perton et al. 2021). Iron stearates derived from Fe2+ and Fe3+ ions were synthesized. In particular, it was shown that Fe2+ ions in the initial iron(II) stearate are partially oxidized to Fe3+ and μ3 -oxo bridged polynuclear cations with general structure [Fe3 -(μ3 -O)St6 ⋅xH2 O]+ Cl− are formed as a result. Polycations with different structures are also formed from the initial iron(III) stearate. Bigger polynuclear cations and then mixed oxide nanoparticles are subsequently assembled through a number of consecutive thermal hydrolysis/decarboxylation and condensation steps with departure of stearic acid molecules. It should be noted that full metal–ligand structures of iron stearates have been barely analyzed in such a detail prior to this published study Perton et al. (2021) despite the fact that iron stearates are large-scale commercial products that are widely used as catalysts in the chemical manufacturing. Based on the above discussed published data, the following structures of copper(II), nickel(II), cobalt(II), and mixed iron(II)/iron(III) stearates were taken in this study as models for molecular mechanics simulation (Figure 8.11). The initial structures of the molecules shown in the above scheme were simulated using the simple molecular mechanics algorithms embedded in the Hyperchem 7.1 software package. The

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results H2O R R

H

O Cu

HC

H

O

O O O

CH

R

R

O

C

O O

HC

O

CH

Cu

R

H

R

O Co

H2O OH2

O

O

C R

O

OH2

C R

O

H2O

O

O

OH2

C

O

O

Co

Co O O

O

H2O

OH2 O

Co O

C

+ C

C C

C

C

O

O

O O

O

O

C R

A–

H2O Fe

O C R

H2O

O Fe

Fe

R

R

O

H

R

R

R

R

C

O

H2O

R

O

Ni

OH2

O

O

H2O

O

O O

C R

R = C17H35

Figure 8.11

Structures of copper(II), nickel(II), cobalt(II), and mixed iron(II)/iron(III) stearates.

obtained molecular models are shown in Figure 8.12. In the ground state, stearate chains in all the simulated molecules tend to trans-conformation arrangement. The carbonyl oxygen atoms of the stearate groups in the Ni(II) complex are oriented in such a way so that they can form hydrogen bonds with the coordinated water molecule ligands to additionally stabilize the structure. The iron(II)/iron(III) μ3 -oxo bridged stearate complex was simulated in the initial trinuclear cluster form with six stearate chains. The stearate chains were oriented by hand in the position most favorable for lateral hydrophobic interactions. This kind of orientation fully conforms to experimental data on the lamellar type of structure for iron(II)/iron(III) stearate complexes (Perton et al. 2021). Simulation of the packing of copper(II) stearate molecules into clusters containing about 50 dimer molecules allowed us to predict well-organized layered structures in this complex with an interlayer periodicity of about 47–48 Å. In addition, periodic layers with spacing distances of 4.28 and 4.67 Å between hydrocarbon chains are most likely present in the molecular organization within the layers. The simulated cluster is shown in Figure 8.13. Metal atoms are arranged approximately at the same distances between each other as hydrocarbon chains. In addition to hydrophobic lateral interactions of long hydrocarbon chains, the ordered structures are retained by strong hydrogen bonding between water molecules coordinated to copper(II) ions in the layer comprised of hydrophilic parts of the complex, which may also involve oxygen atoms of the carboxylic groups. The results of constant temperature MD simulation at 300 K (room temperature region) shows that ordered structures are preserved almost unchanged. Constant temperature molecular dynamics simulation at 700 K reveals complete disordering of hydrocarbon chains and only weak interactions between hydrophilic parts of the dimer molecules, which hold molecules together in the form of flocculent aggregates. Monomolecular nickel(II) stearate tetrahydrate also forms layered structures similar to that of copper(II) stearate complex (Figure 8.14). But close packing requirements in the simulated molecular clusters of nickel(II) stearate lead to substantial tilting of the molecules in the layers. As a result, the interlayer spacing in the small-angle region can be estimated as close to 39 Å. Also, reflections corresponding to periodicities of 4.25 and 4.34 Å due to ordering of hydrocarbon chains within the lamellar layers can be predicted. In addition, metal atoms may be arranged in the periodic

417

418

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

H

Figure 8.12

C

O Cu Ni

Fe

Co

Simulated molecular models of copper, nickel, iron, and cobalt stearates.

structures with an interlayer spacing of 5.30 Å. Molecular dynamics behavior of nickel(II) stearate is not much different from the molecular dynamics behavior of copper(II) stearate complex. Figure 8.15 shows the packing of iron(II)/iron(III) stearate complex into molecular clusters. The simulated molecular cluster of iron(II)/iron(III) stearate cationic complex with general formula [Fe3 -(μ3 -O)St6 •3H2 O]+ is arranged into layers with molecules held together by hydrophobic lateral interactions of hydrocarbon chains and hydrogen bonding between coordinated water ligands and neighboring oxygen atoms of carboxylate groups. It is remarkable that the calculated thickness of the simulated lamellar layer (49.5 Å) practically corresponds to the experimental value reported for the iron stearate complex with this structure (50.7 Å according to the published small angle X-ray diffraction data) (Perton et al. 2021). The periodic arrangement of hydrocarbon chains in the simulated clusters with a spacing of 4.14 Å also conforms quite well to the published value of 4.12 Å for lateral arrangement of aliphatic chains into a two-dimensional hexagonal array within the lamellar layers (Perton et al. 2021). Constant temperature molecular dynamics simulation runs also show the preservation of the molecular ordering at ambient temperatures and strong disordering at 700 K.

8.4 Mechanisms of Aquathermal Cleavage of Carbon–Heteroatom Bonds in Maltene Fractions and Calculation Results

Different views of ground state molecular packing

700 K

Hydrogen bonding

300 K

Figure 8.13 Molecular packing of Cu(II) stearate optimized by the molecular mechanics method and molecular dynamics simulation of the molecular clusters at 300 and 700 K. Different views of ground state molecular packing

700 K

Hydrogen bonding

300 K

Figure 8.14 Molecular packing of Ni(II) stearate optimized by the molecular mechanics method and molecular dynamics simulation of the molecular clusters at 300 and 700 K.

Because of the polymeric structure of cobalt(II) stearate dihydrate with strong bridging coordination between monomer units, this complex exhibits strongly ordered molecular arrangement (Figure 8.16). According to the molecular mechanics simulation results, it looks like the coordination polymer chains are tilted within the lamellar layers periodically repeating at an

419

420

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

Different views of ground state molecular packing

Hydrogen bonding

700 K

300 K

Figure 8.15 Molecular packing of Fe(II)/Fe(III) stearate optimized by the molecular mechanics method and molecular dynamics simulation of the molecular clusters at 300 and 700 K. Different views of ground state molecular packing

Hydrogen bonding

300 K

700 K

Figure 8.16 Molecular packing of Co(II) stearate optimized by the molecular mechanics method and molecular dynamics simulation of the molecular clusters at 300 and 700 K.

8.5 Mechanisms of Aquathermal Pyrolysis of Asphalthene Fractions and Calculation Results

every distance of 41–42 Å. Additional periodicities of 4.15 and 4.35 Å may appear for aliphatic chains and metal atom sites, respectively. Molecular dynamics simulations at 300 and 700 K in this case show that the coordination polymer chains of cobalt(II) stearate complex provide stronger molecular ordering in comparison with the other stearate complexes under consideration, though hydrocarbon chains obviously exhibit the conformational flexibility at 700 K to the same extent. Thus, the obtained simulated structures of the transition metal stearate complexes may serve as a starting point for a further study of the solubility and distribution of these compounds in hydrocarbons by molecular dynamics methods. In addition, the mechanism of nanoparticle formation from these metal stearate complexes under conditions of various in situ thermal treatments of oil reservoirs can be studied by the molecular mechanics methods.

8.5 Mechanisms of Aquathermal Pyrolysis of Asphalthene Fractions and Calculation Results 8.5.1 General Approaches to Quantum Chemical Calculations of Aquathermal Pyrolysis of Polycondensed Aromatic Compounds Asphaltene fractions are largely composed of purely carbon polycondensed aromatics. The heteroatoms in this case are mainly present in the composition of heteroaromatic compounds (for example, in the form of five-membered thiophene, furane, and pyrrole hetrocycles, as well as variety of six-membered aromatic heterocycles, embedded in the polycondensed aromatic system). The destruction of polycondensed aromatic structures for lowering their molecular weight and improving their rheological properties is a separate problem that is restricted not only to the cleavage of exclusively carbon–heteroatom bonds but also to the fragmentation of purely carbon aromatic structures. Because of thermodynamic stability of polycondensed aromatics with a unified π-electron system in asphaltene fractions of heavy oil, the absolute requirements for their destruction seem to be as follows: harsh conditions, a good catalyst (presumably, a heterogeneous catalyst with suitable active sites for weakening the integrity of the unified π-electron system), and additional reagents (for example, oxygen, hydrogen, or water molecules as a source of hydrogen and oxygen atoms) capable of forming new bonds with carbon atoms of the polycondensed aromatic system. The newly formed bonds will lead to further fragmentation with subsequent thermal desorption of intermediate particles, which may be in the form of ions, neutral molecules, and radicals, and remaining lower-molecular-weight fragments of the previous polycondensed aromatic system from the catalyst surface. The viability of this approach was evaluated using the oxidation of pure pyrene as an example. The main reaction pathway of pyrene oxidation, which will be discussed in more detail in the next sections, was separately processed at the B3LYP/6-311G(d,p) level of theory. The combined semiempitrical approach for geometry optimization and single-point calculation by the DFT method for energy parameters was used to simulate heterogeneous catalytic processes that involved the surface of copper(II) oxide. The crystal structure of CuO was obtained from X-ray structural analysis (Asbrink and Waskowska 1991), and the plane with the (010) Miller indices was chosen. The CuO surface was simulated as a slab containing eight elementary cells (two in the z direction, and four along x axis) that corresponds to 22 oxygen atoms and 32 copper atoms. The thickness of the slab was 3.4 Å. The surface area of the selected fragment of the crystal structure of CuO exceeded the dimensions of the pyrene molecule by almost three times, which was placed

421

422

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

in the center of the slab, excluding the role of edge effects. Implicit solvation was added within the polarizable continuum model with the dielectric constant of pyrene 𝜀 = 3.14 (Ishii et al. 1973). In particular, the above approach to quantum chemical simulations was used to theoretically investigate the process of pyrene oxidation. Pyrene is the simplest representative of condensed aromatic compounds in terms of structure, and the process of its oxidation, including the oxidation process in the presence of catalysts, have been comprehensively studied recently by Rodionov et al. (2022). Pyrene is capable of forming free radicals at high temperatures and its oxidation process can proceed by both radical and molecular mechanisms. The radical processes may lead to polymerization and formation of coke during in situ treatment of heavy oil, so it is necessary to methodically adjust the oxidation condition and to find an efficient catalytic system to defeat this problem. Quantum chemical calculations can be applied to better understand the ongoing processes at the molecular level during oxidation of polycondensed aromatic compounds.

8.5.2 Elucidation of the Mechanisms of Pyrene Oxidation on the Surface of Copper(II) Oxide Nanoparticles by Quantum Chemical Calculation Methods Based on the Results of Laboratory Experiments Next, we consider the process of pyrene oxidation in the absence and presence of a catalyst, in this case, the catalyst is oil-soluble copper(II) stearate. As is known, metal soaps undergo decomposition to form oxides of the corresponding metal at temperatures above 450–450 K. The resulting particles typically have sizes of about several hundred nanometers. Thus, depending on the temperature regime, the system may be homogeneous or heterogeneous. The different mechanisms of hydrocarbon oxidation have been actively studied by both theoretical and experimental methods (Dagaut et al. 2001; Curran et al. 2002; Battin-Leclerc et al. 2010; Bugler et al. 2015; Ponduru et al. 2018; Nayebzadeh et al. 2020; Zhang et al. 2020). Sequential oxidation of molecular polycyclic aromatic hydrocarbons up to CO was simulated using as oxidation reagent the atomic oxygen (Yönder and Hättig 2019). The high temperature (above 1000 K) process of oxidation of the pyrenyl radical by molecular oxygen was also simulated. However, the proposed schemes do not reflect a mechanism at low temperatures. Moreover, molecular pyrene is also present in sufficient quantities at low temperatures along with pyrenyl. Next, we present the process of pyrene oxidation in the temperature range of 473–623 K and a pressure of 50 bar and discuss both experimental and theoretical results. The temperature regime in autoclave experiments was chosen in such a way that the melting temperature of pyrene was reached (T m = 418–421 K) (Kawauchi et al. 2011), but not boiling point (T b = 677 K) (Yuan et al. 2018). Insoluble and soluble in CH2 Cl2 parts were observed after heating of pyrene in autoclave. The insoluble reaction products contained coke. The soluble part was analyzed using the matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) technique, and high molecular weight compounds with masses of about 400, 600, and 800 a.u. were observed, which corresponds to two, three, and four condensed pyrene fragments, respectively. The coke formation maybe associated with a local scarcity of oxygen, which causes polymerization of pyrene due to formation of free radicals. When pyrene is heated to 413 K, free radicals were formed as evidenced by electron paramagnetic resonance (EPR) data (Figure 8.17). The EPR signal intensity increases with an increase in the temperature; hence, the concentration of free radicals also increases. It should be noted that the free radicals of pyrene are stable enough and remained for a long time, which enables their detection after the sample cooled to room temperature. Copper(II) stearate reduced the formation of coke and products with a molecular weight above 400 a.u. to a considerable extent (Gall et al. 2017). Based on the thermogravimetry of copper-based

8.5 Mechanisms of Aquathermal Pyrolysis of Asphalthene Fractions and Calculation Results

180 °C 140 °C 80 °C 25 °C

1000 750

Intensity (abs.u)

500 250 0 –250 –500 –750 –1000 330

Figure 8.17

331

332

333 334 335 336 Magnetic field (mT)

337

338

339

EPR data for pyrene samples after exposure to specific temperatures.

catalysts, the decomposition process begins at temperatures of about 420 K. Hence, copper(II) stearate itself can undergo thermal decomposition under certain experimental conditions (Yuan et al. 2018). During the autoclave experiment, copper(II) stearate decomposed to copper(II) oxide which was established by X-ray powder diffraction (XRD) technique. Further experiments using scanning electron microscopy (SEM) revealed the spherical nanoparticles of CuO with sizes in the range of 20–60 nm. According to the obtained experimental data, we can deduce that polymerization processes with the involvement of pyrenyl also occur along with oxidation. Therefore, the oxidation of pyrenyl also needs to be considered alongside the oxidation of molecular pyrene. We have considered possible pathways for the oxidation of pyrene by molecular oxygen in the absence of the catalyst and on the copper(II) oxide surface. The pathways of the oxidation reaction of the pyrenyl radical through the formation of the phenanthryl radical were considered according to the scheme proposed by Raj et al. (2012), but considering implicit solvation and conditions in autoclave experiments (T = 300 ∘ C, P = 50 bar). Sequential oxidation of pyrenyl with activation barriers and transition state structure is given in Figure 8.18. The addition of one oxygen molecule to pyrenyl proceeds without barrier, as does the stage of elimination of one oxygen atom from the product designated as P4_1. The limiting stage in the pyrenyl oxidation was reorganization of six-membered to five-membered cycle (P2 to P3, Figure 8.18) with an activation energy of 268 kJ/mol. The result of a set of sequential and parallel chemical reactions was the phenanthrene radical. At the same time, the considered sequence reaction pathway of pyrene oxidation that involves fewer stages leads to phenalene core (see Figure 8.18). Next, possible pathways for the oxidation of molecular pyrene are considered. In the most stable pyrene–oxygen complex, the oxygen molecule was located close to the center of a pyrene molecule at a distance of 3 Å. The molecular complex of the oxygen molecule with pyrene undergoes a further transformation with the formation of surface peroxide (Figure 8.19). The negative Gibbs energy of the reaction (ΔGr < 0) was observed only in cases when oxygen was attached to two adjacent carbon atoms at the edge of the pyrene molecule. The most favorable product in terms of thermodynamic control (Δ# G∘ = 179 kJ/mol, ΔGr = −92 kJ/mol) was 4,5-peroxide (E_P1). At the same time, the

423

424

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

Pyrenyl

+O2 Without barrier

P1

116

P2

268

P3

–O 86

52 P4_3a

P4_4a

+O2

P4_2a

Without barrier

P4_5a

23 P4_1

–O

47(57) –CO2

P4

–CO

52 –CO

17

136

–CO2

45

P4_2b P4_4b_b1

69

20(27)

61

P4_4b

47

P4_4b_b2

7(–9) 137

P4_4b_a3

90

P4_3b

P4_4b_a2 P4_4b_a1

Figure 8.18 The calculated reaction pathways of the pyrenyl oxidation through the formation of the phenanthryl radical. Energy barriers (kJ/mol) are indicated next to the transition states. The values in parentheses show the activation energy.

smallest energy barrier (Δ# G∘ = 71 kJ/mol, ΔGr = 2 kJ/mol) in for kinetic control had the pathway corresponding to the formation of 1,3a-peroxo-pyrene (A_P1 in Figure 8.19). The smallest barrier and the most preferred product according to the kinetic control has the pathway of formation of the A_P2_a compound which is highlighted in Figure 8.19 by the yellow curve. Thus, the most likely oxidation product according to the simulation results was 9-oxo-phenaleno-1-carbaldehyde (A_P2_a in Figure 8.20). The further cyclization is due to the attack of the radical carbon center of the formyl group on the neighboring ketone group, which is followed by the elimination of a carbon dioxide molecule and the formation of a byphenalenyl radical. Thus, the most favorable reaction pathway of pyrene oxidation has two stages. The first one is the formation of peroxide with an energy barrier of 71 kJ/mol, and the second stage is the elimination of the acetylene molecule (Δ# G∘ = 227 kJ/mol) with a formation of the final product. The cyclization process (see Figure 8.20) has a low activation barrier of 83 kJ/mol, and this stage is not a limiting step. The efficiency of the catalyst can be determined at the stage of sorption of the substrate on its surface. Change in the electronic structure of the adsorbed substrate compared to the nonadsorbed state leads to changes in the geometric structure of the sorbate and, consequently, the weakening of some bonds. Weakened chemical bonds are excellent targets for the attack of oxygen in this case, which suggests both the reaction path and a rough estimate of the effectiveness of the catalyst. We illustrate this idea using the example of the established pathway for the oxidation of molecular pyrene. Despite the fact that copper(II) stearate decomposes to copper(II) oxide that probably plays

Thermodynamic control (ΔGr298 )

Kinetic control (ΔG*) A_P1

ΔG# = 179

#

ΔG = 71 ΔG# = 105

A_P2_a ΔG# = 227

ΔGr298 = 2

ΔG# = 145

ΔG#

= 145

D_P1

C_P1

B_P1

ΔGr298 = –92

ΔGr298 = 166 ΔGr298 = 73

C_P3b

ΔG# = 174

ΔG# = 173 D_P2

ΔGr298 = –23 ΔG# = 145

C_P4b

E_P1

C_P2b

ΔG# = 325 C_P2a

#

ΔG = 148

Figure 8.19 The possible pathways of formation of pyrene peroxides with thermodynamic parameters (kJ/mol) depending on the position of attachment of the oxygen molecule and different directions of the initial stages of the oxidation of molecular pyrene, as a continuation of the peroxide formation.

Intensity × 104.a.u.

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms 13.51

(a)

Biphenalenyl 208 radical Bicharged ion

(b)

100

O

O

50

13.43

163

104 45 12.50

700 600

13.00

13.50 14.00 Time, min

14.50

15.00

Gibbs energy TdS1

0

70

140

350

420

490

O

–5 –10 –15

A_R1 A_P2_a

300

280 O M/z

0

500 400

210

5

(c)

A_P1

–20 –25

200

TdS, kJ/mol

Gibbs energy, kJ/mol

426

–H+ # 3 ΔG 8

kJ/mo

l

–30 Reaction path

Figure 8.20 Analysis of the oxidized products of pyrene (gas chromatography–mass spectrometry results) after experiments in an autoclave and comparing them with the results of the calculations, establishing the dominant product.

the role of a catalyst as a result of heating under experimental conditions, we first considered the sorption of pyrene on the surface of copper(II) stearate. It has been established that the pyrene molecule is oriented near the complex, occupying one of the axial positions of the coordination sphere of copper ion Cu2+ at a distance of about 2.5 Å with an association energy of 26.8 kJ/mol. It should be noted that the mentioned energy value refers to the transfer of the model pyrene molecule from the vacuum to the associate. If we additionally take into account the energy required to extract the molecule of the model compound from the medium in which it is located (for example, if we consider a pyrene melt and then the energy of pyrene solvation in pyrene), then the association energies become either slightly higher than RT (c. 2.5 kJ/mol) or positive, so the association will be already thermodynamically unfavorable. With such a weak interaction between the components in the associate, one should not expect significant changes in the geometric parameters and the electron density distribution. Changes for all systems do not exceed 2, 4, and 4% in bond lengths, angles, and charges on atoms, respectively. It was assumed that sorption (in the case of heterogeneous catalysis) or association (in the case of homogeneous catalysis) is the first stage of the catalytic process, and the catalytic efficiency depends on the ability of the catalyst to influence the strength of some bonds in the substrate molecule. With such weak changes in pyrene molecules, it is not necessary to expect that the copper(II) stearate complex will exhibit high catalytic activity. In ternary systems that additionally include an oxygen molecule, the latter is located above the model compound. The sorption energy of an oxygen molecule on the associate consisting of a stearate complex and pyrene is also low and does not exceed 6 kJ/mol. The change in the length of the oxygen–oxygen bond as a result of sorption for the considered system is less than 2%. At the same time, the adsorption stage of pyrene on the copper(II) oxide slab has been accompanied by serious changes in the sorbate structure. The bond length between the carbon atoms of pyrene and the oxygen atoms of the surface is approximately 1.45 Å, which indicates the covalent nature of the interaction. Pyrene oxidation processes in the presence of copper-containing systems were studied based on the dominant reaction path according to calculations and experimental data (see the highlighted pathway in Figure 8.19). In the presence of CuO slab (Figure 8.21), the number of elementary stages

8.6 Conclusions A_R1

A_P2_a

A_P1

71

227

178

172 47

A_P1_sorbed

TS1_sorbed

TS2_sorbed

A_P2_a_sorbed

Figure 8.21 Comparison of the energy parameters of the dominant pathway of the pyrene oxidation reaction under autoclave conditions in the case of a pure system (upper diagram) and in the presence of the copper(II) oxide surface (lower diagram). The values above the arrows show the barriers for each stage (kJ/mol).

increases from 2 to 3, without taking into consideration the sorption of reagents and the desorption of products. The first stage is the cleavage of a bond between two carbon atoms, one of which is associated with oxygen, and subsequent cleavage of the peroxide bond and the formation, consequently, of aldehyde and epoxy fragments (Δ# G∘ = 178 kJ/mol). After that, the C—C bond in the epoxy group is destroyed with a barrier of 47 kJ/mol, and the oxygen atom is incorporated between carbon atoms. The final stage is the cleavage of the C—O bond (Δ# G∘ = 172 kJ/mol) in such a way that the oxygen atom remains in the oxidized product. The separated fragment is also bound with the substrate. Comparing the processes occurring in the absence and in the presence of a copper catalyst, one can notice that the largest barrier in the case of pure pyrene oxidation was 223 kJ/mol, while the maximum activation energy on the substrate was 178 kJ/mol, which is 25% lower. As was found because of simulation, copper(II) oxide has a catalytic effect on the oxidation of pyrene, which correlates with experimental data. Comparison of the available experimental and the calculated data suggest that the given model adequately reflects the oxidation process.

8.6

Conclusions

In maltene fractions of heavy oil, the heteroatomic bonds are the weakest ones that are most likely to break under condition of aquathermolysis. The results of theoretical calculations of the thermochemistry and kinetics of the processes involved in the aquathermolysis of model compounds that mimic maltene fractions of heavy oil by the DFT methods show that there are several thermodynamically viable routes for cleaving heteroatomic bonds. Even if the unselective processes involving the generation of free radicals are not excluded under aquathermolysis conditions, the ionic mechanisms are more preferable for selective cleavage of heteroatomic bonds, and the selectivity increases with a decrease in the temperature of the processes. To conduct aquathermolysis processes at low temperatures, the presence of catalysts—for example, transition metal ions, such as Cu2+ , Fe2+ , Co2+ , and Ni2+ —is absolutely necessary. According to the DFT calculation results, the aquathermolysis reaction mechanisms involving the heterolytic cleavage of carbon–heteroatom bonds with the formation of the cyclohexyl carbocation has a

427

428

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

substantially lower kinetic barrier compared to the hydrolysis reactions (if we consider possible ways for destruction of heteroatomic bonds by ionic mechanisms). Moreover, the calculation results show that the hydrolysis is not a preferred process both kinetically and thermodynamically. In accordance with the DFT calculation results obtained with the B3LYP, ωB97X-D, and M06-2X functional, transition metal ions substantially reduce the kinetic barrier of the heterolytic thermal cleavage of model compounds (in comparison with the proton-catalyzed process as well). The ωB97X-D and M06-2X functionals give more reliable energy parameters of the processes that involve intermediate transition metal compounds in comparison with the B3LYP functional. In the case of sulfur- and oxygen-containing models of heteroatomic compounds, the investigated transition metal ions can be arranged in sequence Cu2+ ≫ Ni2+ ≈ Fe2+ ≈ Co2+ > H+ in the order of decreasing catalytic activity, in which the Cu2+ -based catalytic system theoretically outperforms the proton-catalyzed system by more than five orders of magnitude. The energy profiles of the aquathermolysis reaction of model compound CPA (which mimics carbon–nitrogen bonds in maltene fractions) with catalytic systems under study show that the Cu2+ ions form a very strong complex with reaction products (amines), thereby leaving the catalytic cycle, i.e. the self-poisoning of the Cu(II) catalyst with reaction products occurs. Using classical formulas of kinetics and thermodynamics, the kinetic constants of aquathermolysis can be calculated on the basis of the quantum chemical simulation results. The approach used in the calculation of kinetic constants is based on the rate-limiting step approximation and on the assumption that the reagents behave as ideal gases. The latter is substantiated by the notion that reagents are distributed in an inert medium (surrounding molecules do not take part in reactions and are considered to be an analog of vacuum in the ideal gas model, which is taken into account by introducing a transmission coefficient [fudge factor] into the kinetic constant formula). Estimates based on the results of quantum chemical simulation show that relatively high reaction rates can be achieved with transition metal catalysts in the temperature range of 400–500 K. Hence, it is theoretically confirmed that transition metal ions can serve as effective catalysts for the downhole upgrading of maltene fractions of heavy oil under relatively mild temperature regimes achievable by superheated steam injection. Oil-soluble transition metal stearates can be considered as a suitable catalytic form convenient for the effective distribution of a catalyst over the heavy oil deposit. Under aquathermolysis conditions, these kinds of catalytic forms can serve both as a source of metal ions upon thermochemical destruction of the initial structure and as a molecular catalyst with active metal sites. Moreover, nanoparticles (for example, metal oxide nanoparticles) can be assembled from metal stearates upon their thermal degradation. They have diverse structures with different numbers of water molecules in the chemical composition depending on the nature of metal atoms (despite their close chemical nature). The variety of possible structures is determined by the fact that the carboxylate groups can serve as both monodentate and bidentate ligands, and also can be either chelating or bridging ligands in the latter case. An analysis of published structural data suggests that the coordination site in the copper(II) stearate monohydrate should have the so-called “Chinese lantern” structure with two metal atoms in a dimeric molecule, while the nickel(II) stearate tetrahydrate complex is stabilized into a monomeric molecule with monodentate coordination of the carboxylate groups. The cobalt(II) stearate tetrahydrate complex is isostructural to the analogous nickel(II) complex, and the cobalt(II) stearate dihydrate complex most likely forms polymer chains via bridged coordination of carboxylate groups. The iron stearate complex derived from Fe2+ ions is partially oxidized to Fe3+ by air oxygen and forms μ3 -oxobridged trinuclear cationic species with six stearic acid ligands and counterions. At ambient temperatures, all these complexes are arranged into lamellar layers with additional organization of alkyl chains and metal ions in two-dimensional periodic arrays within the layers. The parameters of the simulated structure of a mixed iron(II)/iron(III) trinuclear cation

8.6 Conclusions

are in good agreement with published experimental data on the structure of these kinds of iron stearate complexes. The results of molecular dynamics simulation show that layered structures of the studied complexes become disordered at high temperatures. In principle, the mechanisms of nanoparticle formation upon chemical degradation under aquathermolysis conditions (especially for iron(II)/iron(III) stearate complexes) can be studied using quantum chemistry tools. In addition, reactions of dibenzyl sulfide without a catalyst under aquathermolysis conditions are identified. The experiments in an autoclave and a subsequent GC–MS analysis show that the two major products of aquathermolysis of DBS are dibenzyl and trans-stilbene. The calculations carried out at the PBE/6-31G(d,p) level of theory showed that dibenzyl is formed by a direct reaction of DBS with a water molecule. At the same time, water molecules are not involved in the reaction of trans-stilbene formation. The limiting stage of the reaction pathway has an activation barrier of 345 kJ/mol for trans-stilbene formation and 415 kJ/mol for dibenzyl formation. However, the results of the GC–MS analysis indicate that dibenzyl is present in the final mixture in a much larger amount (44%) than trans-stilbene (16%). The calculated reaction path for the formation of dibenzyl from dibenzyl sulfide indicates that hydrogen is also released along with dibenzyl formation, which is able to reduce trans-stilbene to dibenzyl under autoclave conditions, thereby increasing the content of dibenzyl in the final mixture. The effect of the aqueous medium, which is taken into account using the polarizable continuum model (IEFPCM), practically does not affect the energy barriers of both reaction pathways and the changes do not exceed of 15 kJ/mol. The modern scientific approach requires the comprehensive understanding of ongoing processes and physicochemical phenomena when solving practical problems. It is no secret that often the experiment does not reveal the essential details of the observed phenomena. As a consequence, even experimentalists cannot ignore the rapidly developing methods of computational chemistry in order to support their experimental research with theoretical generalizations. Using pyrene oxidation with molecular oxygen in the presence and in the absence of copper(II) stearate, we have demonstrated the application of quantum chemical approaches to solving the problems of degradation of asphalthenes during in situ treatment of heavy oil. It is experimentally determined that pure pyrene oxidation by oxygen leads to the formation of both low and high molecular weight products. The addition of copper(II) stearate substantially reduces the formation of high molecular weight products, while copper(II) stearate itself undergoes thermal decomposition to copper(II) oxide nanoparticles at the autoclave conditions (T = 573 K, P = 50 bar). The dominant oxidized product with and without catalyst is 9-oxo-phenaleno-1-carbaldehyde. Based on our simulations, it is found that the formed copper(II) oxide nanoparticles rather than stearates play the role of catalysts for pyrene oxidation. The most probable reaction pathway of pure pyrene oxidation leading to the formation of 9-oxo-phenaleno-1-carbaldehyde is proposed. The oxidation pathway of pure pyrene includes two elementary stages with the limiting Gibbs energy barrier of 227 kJ/mol. Copper(II) oxide increases the number of stages to three and reduces the limiting Gibbs energy barrier by 20%. The proposed qualitative model adequately describes the pyrene oxidation with and without the catalyst in autoclave experiments. These new insights into the initial oxidation of pyrene and its catalysis provide theoretical basis for the degradation of condensed polycyclic aromatic hydrocarbons in asphalthene fractions. It is shown that the semiempirical PM6 method reproduces the geometric parameters of both the ground and transition states with sufficient accuracy and can be used to quickly search for reaction paths in a system containing copper(II) oxide in the bulk state. Clarifying the energy characteristics by the B3LYP/6-311G(d,p) method in single-point calculations makes it possible to obtain the activation barriers for each elementary stage of the oxidation pathway, without a substantial loss in accuracy. This approach is especially important in the case of simulating reactions on the catalyst surface, the model structure of which contains a lot of heavy

429

430

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

atoms, and the use of ab initio methods in this case for optimization of structures on the catalyst surface or search for transition states requires enormous computational costs. We would also like to draw attention to another important methodological aspect, namely, the sorption of the initial pyrene on the surface of copper(II) oxide and copper(II) stearate. In fact, already at the sorption stage, a number of important assumptions about the catalyst efficiency can be made based on the analysis of changes in the geometric parameters of the adsorbed form of the compound compared to its free state. Thus, no substantial changes in geometric parameters and redistribution of charges in the adsorbed pyrene compared to its free state are observed in the case of copper(II) stearate (the differences are less than 5∘ in angles and 0.05 Å in bonds), and as a result, copper(II) stearate does not exhibit catalytic activity. On the contrary, substantial changes in the structure of pyrene are observed on the surface of copper(II) oxide, which leads to the weakening of some bonds that become targets for the oxygen molecule to attack. The simulation of sorption requires much less computational effort than the simulation of possible reaction paths and allows one to immediately exclude compounds that do not exhibit catalytic activity. Moreover, such an approach can be automated to search for potential catalysts from a large sample using machine-learning methods that have been rapidly developing in recent years. We hope that the presented results will not only expand the understanding of aquathermal cleavage of heteroatomic bonds in maltene fractions of heavy oil and oxidative degradation processes of condensed polycyclic aromatic hydrocarbons in asphalthene fractions but will also contribute to the methodology of search and design of catalytic systems.

References Al-Hajji, A.A., Muller, H., and Koseoglu, O.R. (2008). Characterization of nitrogen and sulfur compounds in hydrocracking feedstocks by Fourier transform ion cyclotron mass spectrometry. Oil & Gas Science and Technology – Revue IFP 63 (1): 115–128. Arend, A.G. (1961). Researches on zinc stearate. Industrial Lubrication and Tribology 13: 38–38. Asbrink, S. and Waskowska, A. (1991). CuO: X-ray single-crystal structure determination at 196 K and room temperature. Journal of Physics. Condensed Matter 3: 8173–8180. Battin-Leclerc, F., Herbinet, O., Glaude, P.-A. et al. (2010). Experimental confirmation of the low-temperature oxidation scheme of alkanes. Angewandte Chemie 122: 3237–3240. Bugler, J., Somers, K.P., Silke, E.J., and Curran, H.J. (2015). Revisiting the kinetics and thermodynamics of the low-temperature oxidation pathways of alkanes: a case study of the three pentane isomers. The Journal of Physical Chemistry. A 119: 7510–7527. Catterick, J. and Thornton, P. (1977). Structures and physical properties of polynuclear carboxylates. Advances in Inorganic Chemistry and Radiochemistry 20: 291–362. Chai, J.-D. and Head-Gordon, M. (2008). Long-range corrected hybrid density functionals with damped atom–atom dispersion corrections. Physical Chemistry Chemical Physics 10: 6615–6620. Chao, K., Chen, Y., Liu, H. et al. (2012). Laboratory experiments and field test of a difunctional catalyst for catalytic aquathermolysis of heavy oil. Energy & Fuels 26: 1152–1159. Chen, Y., Xia, F., Wang, Y., and Wu, C. (2008). Laboratory experiments and field tests of an amphiphilic metallic chelate for catalytic aquathermolysis of heavy oil. Energy & Fuels 22: 1502–1508. Cho, Y., Ahmed, A., and Kim, S. (2013). Application of atmospheric pressure photo ionization hydrogen/deuterium exchange high-resolution mass spectrometry for the molecular level speciation of nitrogen compounds in heavy crude oils. Analytical Chemistry 85: 9758–9763. Clark, P.D. and Kirk, M.J. (1994). Studies on the upgrading of bituminous oils with water and transition metal catalysts. Energy & Fuels 8: 380–387.

References

Curran, H.J., Gaffuri, P., Pitz, W.J., and Westbrook, C.K. (2002). A comprehensive modeling study of iso-octane oxidation. Combustion and Flame 129: 253–280. Dagaut, P., Ristori, A., and Cathonnet, M. (2001). The oxidation of n-hexadecane: experimental and detailed kinetic modeling. Combustion and Flame 125: 1128–1137. Dou, Q. and Ng, K.M. (2016). Synthesis of various metal stearates and the corresponding monodisperse metal oxide nanoparticles. Powder Technology 301: 949–958. Frisch, M.J., Trucks, G.W., Schlegel, H.B. et al. (2010). Gaussian 09, Revision B.01. Wallingford, CT: Gaussian, Inc. Gall, D., Pushp, M., Davidsson, K.O., and Pettersson, J.B.C. (2017). Online measurements of alkali and heavy tar components in biomass gasification. Energy & Fuels 31: 8152–8161. Galukhin, A.V., Erokhin, A.A., and Nurgaliev, D.K. (2015). Effect of catalytic aquathermolysis on high-molecular-weight components of heavy oil in the Ashal’cha field. Chemistry and Technology of Fuels and Oils 50: 555–560. Greaves, M., Young, T.J., El-Usta, S. et al. (2000). Air injection into light and medium heavy oil reservoirs: combustion tube studies on west of Shetlands Clair oil and light Australian oil. Chemical Engineering Research and Design 78: 721–730. Hart, A. (2014). The novel THAI–CAPRI technology and its comparison to other thermal methods for heavy oil recovery and upgrading. Journal of Petroleum Exploration and Production Technologies 4: 427–437. Hart, A., Wood, J., and Greaves, M. (2017). In situ catalytic upgrading of heavy oil using a pelletized Ni–Mo/Al2 O3 catalyst in the THAI process. Journal of Petroleum Science and Engineering 156: 958–965. Haynes, W.M. (2014). CRC Handbook of Chemistry and Physics. Boca Raton, Florida: CRC Press 2704 p. Ishii, K., Kinoshita, M., and Kuroda, H. (1973). Dielectric constant measurement on organic crystalline powder. Bulletin of the Chemical Society of Japan 46: 3385–3391. Javadli, R. and de Klerk, A. (2012). Desulfurization of heavy oil. Applied Petrochemical Research 1: 3–19. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2013). A new reaction model for aquathermolysis of Athabasca bitumen. Canadian Journal of Chemical Engineering 91: 475–482. Kawauchi, T., Kawauchi, M., Kodama, Y., and Takeichi, T. (2011). Formation of the inclusion complex of helical syndiotactic poly(methyl methacrylate) and polycyclic aromatic hydrocarbons. Macromolecules 44: 3452–3457. Kayukova, G.P., Gubaidullin, A.T., Petrov, S.M. et al. (2016). Changes of Asphaltenes’ structural phase characteristics in the process of conversion of heavy oil in the hydrothermal catalytic system. Energy & Fuels 30: 773–783. Khafizov, N.R., Madzhidov, T.I., Varfolomeev, M.A. et al. (2022a). Theoretical insight into the catalytic effect of transition-metal ions on the Aquathermal degradation of heavy oil: a DFT study of cyclohexyl phenyl ether cleavage. Fuel 311: 122595. Khafizov, N.R., Madzhidov, T.I., Yuan, C. et al. (2022b). Theoretical insight into the catalytic effect of transition metal ions on the Aquathermal degradation of heavy oil: a DFT study of cyclohexyl phenyl amine cleavage. Fuel 312: 123002. Kurniawaty, I., Yulizar, Y., Gamar I., et al. (2017). Nitrogen compounds characterization on Dumai heavy kerosene oil from a blend of Indonesian crude. Proceedings of the 3rd International Symposium on Current Progress in Mathematics and Sciences 2017 (ISCPMS2017), AIP Conf. Proc. 020087-1–020087-4. Laidler, K.J. (1996). A glossary of terms used in chemical kinetics, including reaction dynamics (IUPAC recommendations 1996). Pure and Applied Chemistry 68: 149–192.

431

432

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

Laikov, D.N. and Ustynyuk, Y.A. (2005). PRIRODA-04: a quantum-chemical program suite. New possibilities in the study of molecular systems with the application of parallel computing. Russian Chemical Bulletin 54: 820–826. Laredo, G.C., Leyva, S., Alvarez, R. et al. (2002). Nitrogen compounds characterization in atmospheric gas oil and light cycle oil from a blend of Mexican crudes. Fuel 81: 1341–1350. Le Maître, J., Hubert-Roux, M., Paupy, B. et al. (2019). Structural analysis of heavy oil fractions after hydrodenitrogenation by high resolution tandem mass spectrometry and ion mobility spectrometry. Faraday Discussions 218: 417–430. Lower, E.S. (1981). Magnesium stearate: a review of its uses in paints, plastics, adhesives and related industries. Pigment & Resin Technology 10: 7–25. Lower, E.S. (1982a). The properties of aluminium stearate and its uses in the coatings and allied industries. Pigment & Resin Technology 11: 13–18. Lower, E.S. (1982b). Zinc stearate: its properties and uses. Pigment & Resin Technology 11: 9–14. Lower, E.S. (1990a). Calcium stearate – Part 1. Pigment & Resin Technology 19: 4–4. Lower, E.S. (1990b). Calcium stearate – Part 2. Pigment & Resin Technology 19: 4–6. Lower, E.S. (1990c). Calcium stearate – Part 3. Pigment & Resin Technology 19: 8–10. Lysogorskii, Y.V. and Aminova, R.M. (2015). Quantum chemical study of mechanisms of the reaction of cyclohexyl phenyl sulfide with water. Chemistry and Technology of Fuels and Oils 51: 113–116. Lysogorskiy, Y.V., Aminova, R.M., and Tayurskii, D.A. (2016). Initial steps in reactions of aquathermolysis of cyclohexyl phenyl sulfide by means of ab initio calculations. Computational & Theoretical Chemistry 1078: 138–145. Maity, S.K., Ancheyta, J., and Marroquín, G. (2010). Catalytic aquathermolysis used for viscosity reduction of heavy crude oils: a review. Energy & Fuels 24: 2809–2816. Mehrabi-Kalajahi, S., Varfolomeev, M.A., Yuan, C. et al. (2020). Oil-dispersed α-Fe2 O3 nanoparticles as a catalyst for improving heavy oil oxidation. Energy & Fuels 35: 10498–10511. Mironenko, O.O., Sosnin, G.A., Eletskii, P.M. et al. (2017). Catalytic steam cracking of heavy crude oil with molybdenum and nickel nanodispersed catalysts. Catalysis in Industry 9: 221–229. Morelos-Santos, O., Reyes de la Torre, A.I., Melo-Banda, J.A. et al. (2022). Novel direct method in one-step for catalytic heavy crude oil upgrading using iron oxide nanoparticles. Catalysis Today 392–393: 60–71. Nayebzadeh, M., Vahedpour, M., Shiroudi, A., and Rius-Bartra, J.M. (2020). Kinetics and oxidation mechanism of pyrene initiated by hydroxyl radical. A theoretical investigation. Chemical Physics 528: 110522. Nelson, P.N. and Taylor, R.A. (2014). Theories and experimental investigations of the structural and thermotropic mesomorphic phase behaviors of metal carboxylates. Applied Petrochemical Research 4: 253–285. Nordin, N., Hasan, S.Z., Zakaria, Z. et al. (2015). Effect of applied voltage on slow-release of cu(II) ions on the synthesis of copper(II) stearate complex by electrochemical technique. Journal of the Brazilian Chemical Society 26: 1115–1123. Perton, F., Cotin, G., Kiefer, C. et al. (2021). Iron stearate structures: an original tool for nanoparticles design. Inorganic Chemistry 60: 12445–12456. Ponduru, T.T., Qiu, C., Mao, J.X. et al. (2018). Copper(I)-based oxidation of polycyclic aromatic hydrocarbons and product elucidation using vacuum ultraviolet spectroscopy and theoretical spectral calculations. New Journal of Chemistry 42: 19442–19449. Prado, G.H.C., Rao, Y., and de Klerk, A. (2017). Nitrogen removal from oil: a review. Energy & Fuels 31 (1): 14–36. Raj, A., da Silva, G.R., and Chung, S.H. (2012). Reaction mechanism for the free-edge oxidation of soot by O2 . Combustion and Flame 159: 3423–3436.

References

Rodionov, N.O., Neklyudov, V.V., Yuan, C. et al. (2022). New insights of the initial processes of pyrene oxidation in the presence of cu-based catalysts: quantum-mechanical calculation and experimental evidence. Applied Catalysis A: General 643: 118775. Santos, R.G., Loh, W., Bannwart, A.C., and Trevisan, O.V. (2014). An overview of heavy oil properties and its recovery and transportation methods. Brazilian Journal of Chemical Engineering 31: 571–590. Shah, A., Fishwick, R., Wood, J. et al. (2010). Review of novel techniques for heavy oil and bitumen extraction and upgrading. Energy & Environmental Science 3: 700–714. Shannon, R.D. (1976). Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallographica. Section A 32: 751–767. Shokrlu, Y.H. and Babadagli, T. (2013). In situ upgrading of heavy oil/bitumen during steam injection by use of metal nanoparticles: a study of in situ catalysis and catalysis transportation. SPE Reservoir Evaluation and Engineering 333–344. Shokrlu, Y.H., Maham, Y., Tan, X. et al. (2013). Enhancement of the efficiency of in situ combustion technique for heavy-oil recovery by application of nickel ions. Fuel 105: 397–407. Siskin, M. and Katritzky, A.R. (2001). Reactivity of organic compounds in superheated water: general background. Chemical Reviews 101: 825–835. Siskin, M., Brons, G., Katritzky, A.R., and Balasubramanian, M. (1990a). Aqueous organic chemistry. 1. Aquathermolysis: comparison with thermolysis in the reactivity of aliphatic compounds. Energy & Fuels 4: 475–482. Siskin, M., Brons, G., Katritzky, A.R., and Murugan, R. (1990b). Aqueous organic chemistry. 2. Cross-Linked Cyclohexyl Phenyl Compounds. Energy & Fuels 4: 482–488. Siskin, M., Brons, G., Vaughn, S.N. et al. (1990c). Aqueous organic chemistry. 3. Aquathermolysis: reactivity of ethers and esters. Energy & Fuels 4: 488–492. Speight, J.G. (2016). Feedstock composition. In: Handbook of Petroleum Refining (ed. J.G. Speight), 101–142. Boca Raton, Florida, USA: CRC Press. Tumanyan, B.P., Petrukhina, N.N., Kayukova, G.P. et al. (2015). Aquathermolysis of crude oils and natural bitumen: chemistry, catalysts and prospects for industrial implementation. Russian Chemical Reviews 84: 1145–1175. Tverdov, I., Khafizov, N.R., Madzhidov, T.I. et al. (2020). Theoretical insights into the catalytic effect of transition-metal ions on the aquathermal degradation of sulfur-containing heavy oil: a DFT study of cyclohexyl phenyl sulfide cleavage. ACS Omega 5: 19589–19597. Vakhin, A. V., Mukhamatdinov, I. I., Aliev, F. A., et al. (2021). Industrial Application of Nickel Tallate Catalyst During Cyclic Steam Stimulation in Boca De Jaruco Reservoir. SPE Russian Petroleum Technology Conference, 2021, October 12–15, Paper Number: SPE-206419-MS. https://onepetro.org/ SPERPTC/proceedings-abstract/21RPTC/2-21RPTC/D021S007R002/470308. Van Hecke, G.R., Nakamoto, T., Clements, T.G., and Sorai, M. (2003). Adiabatic calorimetry of the metallomesogen purple cobalt stearate Co(O2 CC17 H35 )2 . Liquid Crystals 30: 831–837. Van Niekerk, J.N. and Schoening, F.R.L. (1953a). A new type of copper complex as found in the crystal structure of cupric acetate, Cu2 (CH3 COO)4 ⋅ 2H2 O. Acta Cryst 6: 227–232. Van Niekerk, J.N. and Schoening, F.R.L. (1953b). The crystal structures of nickel acetate, Ni(CH3 COO)2 ⋅4H2 O, and cobalt acetate, Co(CH3 COO)2 ⋅4H2 O. Acta Cryst 6: 609–612. Varfolomeev, M.A., Yuan, C., Bolotov, A.V. et al. (2021). Effect of copper stearate as catalysts on the performance of in-situ combustion process for heavy oil recovery and upgrading. Journal of Petroleum Science and Engineering 207: 1–11. (109125). Wen, S., Zhao, Y., Liu, Y., and Hu, S. (2007). A study on catalytic aquathermolysis of heavy crude oil during steam stimulation. In: International Symposium on Oilfield Chemistry February 28–March 2. Houston, Texas, USA: Paper Number: SPE-106180-MS. https://onepetro.org/SPEOCC/proceedingsabstract/07OCS/All-07OCS/SPE-106180-MS/143596.

433

434

8 Application of Quantum Chemical Calculations for Studying Thermochemistry, Kinetics, and Catalytic Mechanisms

Wu, C., Lei, G.-L., Yao, C.-J. et al. (2010). Mechanism for reducing the viscosity of extra-heavy oil by aquathermolysis with an amphiphilic catalyst. Journal of Fuel Chemistry and Technology 38: 684–690. Yamamoto, M., Taguchi, K., and Sasaki, K. (1991). Basic nitrogen compounds in bitumen and crude oils. Chemical Geology 93: 193–206. Yi, S., Babadagli, T., and Li, H.A. (2020). Catalytic-effect comparison between nickel and iron oxide nanoparticles during aquathermolysis-aided cyclic steam stimulation. SPE Reservoir Evaluation and Engineering 282–291. Yönder, Ö. and Hättig, C. (2019). Density functional theory study of CO formation through reactions of polycyclic aromatic hydrocarbons with atomic oxygen (O(3p)). Fuel 241: 851–860. Yuan, C., Varfolomeev, M.A., Emelianov, D.A. et al. (2018). Copper stearate as a catalyst for improving the oxidation performance of heavy oil in in-situ combustion process. Applied Catalysis A: General 564: 79–89. Yuan, C., Emelianov, D.A., Varfolomeev, M.A. et al. (2021). Mechanistic and kinetic insight into catalytic oxidation process of heavy oil in in-situ combustion process using copper(II) stearate as oil soluble catalyst. Fuel 284: 1–12. (118981). Zhang, T., Liang, Y., Li, B. et al. (2004). Identification of structures of nitrogen-containing compounds in crude oils in conjunction with chemometric resolution. Annali di Chimica 94: 783–794. Zhang, G., Lin, J., Guo, D.-W. et al. (2010). Infinite coordination polymers of one- and two-dimensional cobalt acetates. Zeitschrift für Anorganische und Allgemeine Chemie 636: 1401–1404. Zhang, T., Lang, X., Dong, A. et al. (2020). Difference of oxidation mechanism between light C3 –C4 alkane and alkene over mullite YMn2 O5 oxides’ catalyst. ACS Catalysis 10: 7269–7282. Zhao, Y. and Truhlar, D.G. (2008). The M06 suite of density functionals for Main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theoretical Chemistry Accounts 120: 215–241.

435

9 Behavior of Catalyst in Porous Media Timur R. Zakirov 1 , Rail I. Kadyrov 1 , Chengdong Yuan 2,3 , and Mikhail A. Varfolomeev 2 1

Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Department of Petroleum Engineering, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia 3 Center for Petroleum Science and Engineering, Skolkovo Institute of Science and Technology, Moscow Region, Russia 2

9.1

Introduction

The mass transfer process in porous media is an important phenomenon that takes place during the convective–diffusion transport of catalysts in the pore space. The interplay between convective and diffusion parameters, reaction rate constants, and the characteristic of a porous medium provides various patterns of mass transfer processes of catalysts (Ancheyta et al. 2002; Chen et al. 2022). The influence of porosity and particle size on the total (inside particles) adsorption dynamics has been investigated by Wang et al. (2017) and Zhou et al. (2015). The results showed that an increase in porosity promotes faster adsorption, and a decrease in particle size (at a fixed porosity) tends to increase the rate of total adsorption. It has been found that increasing the flow rate stimulates the adsorption dynamics. Chen et al. (2012) reported that the reduction in particle size negatively affects the total amount of adsorbed fraction. Zhou et al. (2016) showed that a direct relationship has been found between the particle size and the dynamics of surface adsorption, which contradicts the trend of total adsorption. This result is explained by the increase in the surface area of the particles during their grinding. Lei et al. (2021) found that the number of adsorbent particles negatively affects the surface adsorption rate. It has been also reported that the inlet adsorbate concentration does not affect the seepage velocity but leads to a decrease in the adsorption equilibrium time. Lv et al. (2020) reported the influence of the fractal characteristics of coal samples on oxygen adsorption. A direct relationship has been established among fractal dimension, porosity, and adsorption amount. Liu et al. (2019) studied the dynamic adsorption of ions in porous electrodes. It has been found that the inlet velocity does not affect the final ion adsorption amount but does change the time of the saturation state. Machado (2012) and Yu et al. (2019) investigated the influence of the adsorption rate constant, as well as the Damkohler and Peclet numbers, on the adsorption dynamics. The results showed that these reaction and convection–diffusion parameters are highly influencing factors that can significantly stimulate the rate of adsorption. The effect of the coefficients of inter- (outside the adsorbent) and intraparticle (inside the adsorbent) diffusion has been considered by Sullivan et al. (2005). It has been found that the flow at high Peclet numbers (low diffusion) results in low overall mass transfer due to the “preferential formation of larger saturated stagnant zones which effectively did not participate further in the reaction” and the total adsorption dynamics depend on the combination of inter- and intraparticle diffusion coefficients Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

436

9 Behavior of Catalyst in Porous Media

and Peclet and Reynolds numbers. The adsorption characteristics for different transport and adsorption properties are studied by Vanson et al. (2015). Zaafouri et al. (2021) investigated the important role of the relationship between the convection and diffusion mechanisms. This chapter presents a systematical investigation of the effect of pore space heterogeneity on the dynamics adsorption of catalyst dissolved in the water during a single-phase flow. The influence of heterogeneity on the adsorption process is studied in combination with porosity, Peclet, and Damkohler numbers. Also, this chapter presents a new methodology, which allows one to register the catalyst distribution in the pore space using 4D tomography.

9.2

Methods

9.2.1

Mathematical Model

In this chapter, mathematical modeling is used as a tool for studying dynamic adsorption processes in 2D and 3D spaces. The mathematical model consists of three blocks: governing equations for a single-phase fluid flow; model for interparticle transport of an adsorbate dissolved in a fluid; and equations describing surface reactions and adsorbate diffusion inside adsorbent particles. Fluid flow is described using the lattice Boltzmann equations. The variables such as time- and space-dependent pressure and velocity are calculated using the distribution functions f i (r, t) (Succi 2001), where the index i determines the direction of particles motion during the time step Δt. Potentially, D2Q9 and D3Q19 lattices can be considered (here and below, i = 1 ÷ 9 for the D2Q9 lattice and i = 1 ÷ 19 for the D3Q19 lattice). For the D2Q9 lattice, the directions of particles movement are described by the following basis: e1 = c⋅(0, 0); e2 = c⋅(1, 0); e3 = c⋅(0, 1); e4 = c⋅(−1, 0); e5 = c⋅(0, −1); e6 = c⋅(1, 1); e7 = c⋅(−1, 1); e8 = c⋅(−1, −1); e9 = c⋅(1, −1), where Δl is the grid step, and c = Δl/Δt is the lattice speed. For the D3Q19 case, the basis vectors are written as follows: e1 = c⋅(0, 0, 0); e2 = c⋅(1, 0, 0); e3 = c⋅(−1, 0, 0); e4 = c⋅(0, 1, 0); e5 = c⋅(0, −1, 0); e6 = c⋅(0, 0, 1); e7 = c⋅(0, 0, −1); e8 = c⋅(1, 1, 0); e9 = c⋅(−1, 1, 0), e10 = c⋅(1, −1, 0); e11 = c⋅(−1, −1, 0), e12 = c⋅(1, 0, 1); e13 = c⋅(−1, 0, 1); e14 = c⋅(1, 0, −1); e15 = c⋅(−1, 0, −1); e16 = c⋅(0, 1, 1); e17 = c⋅(0, −1, 1); e18 = c⋅(0, 1, −1); e19 = c⋅(0, −1, −1). The evolution of f i in time and space is described by Eq. (9.1) (Succi 2001): fi (r + ei Δt, t + Δt) = fi (r, t) + Ωi (r, t)

(9.1)

The first stage in solving Eq. (9.1) is a streaming of the particles, at which, during the time Δt, they move to the neighboring nodes in the directions ei possible for D2Q9 or D3Q19 lattices. Further, the collision of particles is considered, after which the distribution functions f i tend to an equilibrium state (operator Ωi ). Macroscopic density and velocity are calculated from the Eqs. (9.2) and (9.3): 𝜌(r, t) =

n ∑

fi (r, t)

(9.2)

1∑ e f (r, t) 𝜌 i=1 i i

(9.3)

i=1

n

u(r, t) =

In Eqs. (9.2) and (9.3) and below, n = 9 and 19 for D2Q9 and D3Q19 lattices, respectively. Fluid pressure and density are related as follows: P = 𝜌c2 /3. The relationship between kinematic viscosity 𝜂 (m2 /s) and the time step is established using the relaxation parameter 𝜏 f for fluid flow (Succi 2001): ( ) 2𝜏f − 1 Δl2 𝜂= (9.4) 6 Δt

9.2 Methods

The collision operator Ωi in Eq. (9.1) is described using the multirelaxation time (MRT) scheme (Pan et al. 2006), the advantage of which have been demonstrated by Pan et al. (2006) and Zakirov et al. (2018). The MRT operator is described as follows: ( eq ) Ωi = −M−1 S mi − mi (9.5) ∑9 Moments m = (𝜌, e, 𝜀, jx , qx , jy , qy , pxx , pxy )T are calculated as mi = k=1 Mik ⋅ fk . The form of eq matrix M, the equations for the equilibrium moments m , and the ninth-order diagonal collision matrix S are given by Pan et al. (2006) and Zakirov et al. (2018). At the initial time, the pore space is filled with a fluid, and a fluid with the same properties is injected through the inlet boundary with a known flow rate u0 . To ensure an unambiguous pressure field, the pressure on the outlet boundary is also fixed and known. On the external and internal impermeable boundaries, u = 0 (“bounce-back” conditions (Succi 2001)). The concept of the interparticle mass transfer involves two stages: interparticle mass transport and reaction on the outer surface of the particle. The adsorbate transport in the pore space occurs due to the fluid convection in which the adsorbate is dissolved and diffusion occurs, which is caused by the concentration gradient. These processes are described by the classical convective diffusion equation. In this chapter, the adsorbate concentration in the fluid is proposed to be sufficiently low that its influence on the flow characteristics is negligible (Kang et al. 2007). The adsorbate concentration, by analogy with the fluid flow, is determined by the concentration distribution functions gl , the evolution of which is written as follows (Lei et al. 2021; Zhou et al. 2015, 2016): gl (r + el Δt, t + Δt) = gl (r, t) −

] 1 [ eq g (r, t) − gl (r, t) 𝜏s l

(9.6) eq

where 𝜏 s is the relaxation parameter for the adsorbate component, and gl is the equilibrium distribution function. Equation (9.6), as in other research studies by Chen et al. (2012) and Machado (2012), contains a single-relaxation time model as a collision scheme, which is justified by the reduction in computational costs. According to Zakirov and Khramchenkov (2022), the dynamic adsorption curves weakly depend on 𝜏 s if its value is less than 1.5. As reported by Machado (2012) and Zhou et al. (2015), it is possible to reduce the directions of particles movement from 9 to 5 for a 2D space (D2Q5 lattice) and from 19 to 7 for a 3D space (D3Q7 lattice) without loss of accuracy of numerical results, which also saves computational time. For D2Q5 lattice, e1 = c⋅(0, 0); e2 = c⋅(1, 0); e3 = c⋅(0, 1); e4 = c⋅(−1, 0); e5 = c⋅(0, −1); for D3Q7 lattice, e1 = c⋅(0, 0, 0); e2 = c⋅(1, 0, 0); e3 = c⋅(−1, 0, 0); e4 = c⋅(0, 1, 0); e5 = c⋅(0, −1, 0); e6 = c⋅(0, 0, 1); e7 = c⋅(0, 0, −1). In Eq. (9.6) and below, l = 1 ÷ 5 for the D2Q5 lattice and l = 1 ÷ 7 for the D3Q7 lattice. The concentration C and the relaxation parameter 𝜏 s are calculated using the following formulas (Zhou et al. 2015): C(r, t) =

q ∑

gl (r, t)

l=1

𝜏s = 0.5 + 3 ⋅

Ds ⋅ Δt Δl2

(9.7) (9.8)

where Ds is the interparticle diffusion coefficient. In Eq. (9.6) and below, q = 5 and 7 for D2Q5 and D3Q7 lattices, respectively. The equilibrium distribution functions in Eq. (9.6) are as described by Zhou et al. (2015): ] [ (e ⋅ u) eq ̃ lC 1 + l 2 (9.9) gl (r, t) = w c

437

438

9 Behavior of Catalyst in Porous Media

̃ i are the weight coefficients. For D2Q5 lattice, w ̃ 1 = 1∕3, w ̃ 2−5 = 1∕6; for D2Q7 lattice: where w ̃ 2−7 = 1∕8. ̃ 1 = 1∕4, w w The boundary conditions at the particle surface have been proposed in Kang et al. (2007): q ∑ gl = Cu − Ds ∇C

(9.10)

i=1

Considering that u = 0 on the solid boundary, the Kang’s condition (Eq. (9.10)) can be written as follows (Zhou et al. 2015): 𝜕Cw 𝜕C ; g3 − g5 = −Ds w 𝜕x 𝜕y 𝜕Cw 𝜕Cw 𝜕C ; g4 − g5 = −Ds ; g6 − g7 = −Ds w D3Q7 lattice: g2 − g3 = −Ds 𝜕x 𝜕y 𝜕z

D2Q5 lattice: g2 − g4 = −Ds

(9.11) (9.12)

where Cw is the concentration on the particle surface. In this chapter, concentration gradients are estimated using a three-point, finite-difference scheme by Lei et al. (2021) and Zhou et al. (2015, 𝜕Cw 𝜕Cw and , the scheme is similar): 2016) (for 𝜕y 𝜕z i,j,k

i,j,k

3Cw − 4Ci+1,j,k + Ci+2,j,k 𝜕Cw = (9.13) 𝜕x 2Δl Porous channels of adsorbent particles exist on the outer surface (micropores) and inside them (nanopores). Therefore, the study of adsorption should involve modeling of the surface reaction, which is the stage of interparticle mass transfer, and the adsorption inside particles, called intraparticle diffusion, which depends on the surface reaction. In this chapter, the surface reaction between the adsorbate and the outer surface of particles is described by the Langmuir adsorption equation (Lei et al. 2021; Zhou et al. 2015, 2016): Ds

𝜕Cw 𝜕N surf = = [ka Cw (N sat − N surf ) − kd N surf ] 𝜕n 𝜕t

(9.14)

where ka and kd are the adsorption and desorption rate constants, respectively; N surf is the adsorbed phase amount on the particles surface; N sat is the maximum or saturation adsorbed amount; n is the normal vector to the solid surface. Based on Eq. (9.14), the new value of surface adsorption amount at time step “t + 1” is calculated as follows: surf

surf

Nt+1 = Nt

surf

+ Δt ⋅ [ka Cwt (N sat − Nt

surf

) − k d Nt

]

(9.15)

To find the surface adsorbate concentration (Cw ) on the right side of Eq. (9.15), Eq. (9.13) is substituted into Eq. (9.14), which describes the Langmuir adsorption kinetics. When describing the intraparticle diffusion, it is assumed that the diffusion coefficient (Dsp ) is uniformly distributed over the internal area of the particles. Such diffusion mechanics inside a particle is described by the second Fick’s law: 𝜕N total = Dsp ΔN total 𝜕t

(9.16)

where N total is the adsorbed phase amount inside the particles. When solving Eq. (9.16), N surf defined on the surface is used as boundary conditions. In this chapter, we study both surface-adsorbed amount and the adsorption amount inside the particles called total adsorption. These parameters are presented in the dimensionless form.

9.2 Methods surf

Surface-adsorbed amount Nt (i, j), calculated at the time t in computational node (i, j), (i) is averaged over all computational nodes belonging to the particles surface, and (ii) is normalized to the saturation adsorption value: ∑ surf i,j Nt (i, j) surf Nt = (9.17) W surf ⋅ N sat where W surf is the total number of nodes at the particles’ surface. The dimensionless total adsorbed amount is calculated as follows: ∑ total (i, j) i,j Nt total Nt = (9.18) W total ⋅ N sat In Eq. (9.18), (i, j) belongs to the nodes inside the particles, including nodes on the surface; W total is the total number of nodes of the adsorbent particles. At the initial time, there is no catalyst in the pore space. Being dissolved in the fluid, the adsorbate is injected together with the fluid through the inlet boundary (x = 0, 0 < y < Ly , C = C0 ). On the external top and bottom boundaries, 𝜕C/𝜕y = 0; on the outlet section, 𝜕C/𝜕x = 0. The described governing equations with boundary conditions can be used for any component dissolved in a fluid and having a low concentration. Specific types of adsorbate and adsorbent particles are characterized by inter- and intraparticle diffusion coefficients, reaction rate constants, and saturation adsorption amount. This chapter considers the adsorption of catalyst in the form of molecules dissolved in the injected water. The adsorbent is selected as Cu-BTC (Wang et al. 2019). For such a catalyst/water system, the physical parameters are given in Table 9.1. Table 9.1

Dimensionless and physical parameters for the adsorption calculation of catalyst.

Parameter

Physical symbol

Grid step

Δl

5⋅10−6 m

Time step

Δt

8.33⋅10−7 s

Kinematic viscosity

𝜂

1.5 ⋅10

Fluid density

𝜌

1000 kg/m3

Inlet concentration

C0

Physical value

−6

Lattice symbol

̂ Δl ̂ Δt 2

m /s

40.87 mol/m

3

Dimensionless value

1 1

𝜂̂

0.3332

𝜌̂ ̂ C0

1 1

Scaling relation

̂ = Δl∕(5 ⋅ 10−6 ) Δl ̂ = Δt∕(8.33 ⋅ 10−7 ) Δt 𝜂̂ = 𝜂 ⋅ Δt∕Δl2 𝜌̂ = 𝜌∕0.65 ( ) ̂ = C ∕ ΔN C 0

0

Δl3

[5⋅10−7 ÷3⋅10−5 ] m2 /s

û0 ̂s D

[0.0166 ÷ 1]

Dsp

[10−9 ÷10−7 ] m2 /s

̂sp D

[0.3 ÷ 33] 10−4

Adsorption rate constant

ka

2.5⋅103 m3 /(mol⋅s)

k̂a

0.085

k̂a = ka ⋅ Δt ⋅

Desorption rate constant

kd

5⋅103 s−1

k̂d

0.0042

k̂d = kd ⋅ Δt

sat ̂ N

13.21

̂ N sat = Nsat ∕

Flow rate

u0

5⋅10

Intraparticle diffusion coefficient

Ds

Interparticle diffusion coefficient

Saturation adsorption amount

N sat

−4

m/s

540 mol/m3

0.833⋅10−4

ûo = u ⋅ Δt∕Δl ̂s = Ds ⋅ Δt∕Δl2 D ̂sp = Dsp ⋅ Δt∕Δl2 D (

(

ΔN Δl3

ΔN Δl3

)

)

439

440

9 Behavior of Catalyst in Porous Media

(a) 350

(b) H=0

350

(c) H = 0.03

350

300

300

300

250

250

250

200

200

200

150

150

150

100

100

100

50

50

50

50

(d) 350

100 150 200 250 300 350

H = 0.104

(e) 350

50

100 150 200 250 300 350

H = 0.13

(f) 350

300

300

300

250

250

250

200

200

200

150

150

150

100

100

100

50

50

50

50

100 150 200 250 300 350

50

100 150 200 250 300 350

H = 0.055

50

100 150 200 250 300 350

H = 0.17

50

100 150 200 250 300 350

Figure 9.1 Digital porous media with various disorders: (a) H = 0 (ordered structure); (b) H = 0.03 (low heterogeneity); (c) and (d) H = 0.055 and 0.104 (medium heterogeneity); (e) and (f) H = 0.13 and 0.15 (high heterogeneity). Black indicates adsorbent particles, white is a pore space.

9.2.2

Artificial Digital Models of Porous Media

The mass transfer processes are studied using numerical simulations performed on artificially created digital images of porous media using the Monte Carlo movement method, which allows one to control the pore space heterogeneity, which is consistent with the objectives of this work. In this chapter, we investigate the porous media made up of grains of the same size that do not overlap. The solid phase, which is an adsorbent, is represented by particles with a shape close to round. Pore space heterogeneity is numerically described by the disorder √ parameter H, which N 2 ∑ (𝜑𝛼i −𝜑) characterizes the corrected standard deviation of local porosity: H = , where 𝜑 is the N−1 i=1

average porosity of the sample, 𝜑𝛼i is the local porosity calculated in the i-th unit cell, and N is the number of cells. To simplify the estimation of the disorder, the unit cells are in the form of uniform squares, but we must admit that the Voronoi diagram is more suitable for the accurate assessment of H. Figure 9.1 shows a set of porous media with 𝜑 = 0.64 and various heterogeneities.

9.2.3

Validation of Intraparticle Diffusion Model

As a benchmark test, we consider the dynamics of static adsorption in a round particle. An adsorbent particle is placed in a motionless field of a uniformly distributed adsorbate concentration that does not change with time. For this problem, the dimensionless dynamics of total adsorption is described by the analytical formula (Zhou et al. 2015): N total = 1 −

∞ ∑

( ) 4 exp −Dsp 𝛼n2 t 2 2 r ⋅ 𝛼 n n=1 s

(9.19)

9.2 Methods

Total adsorption amount, Ntotal

1,0

0,8

0,6

0,4

0,2 Analytical solution LBE results 0,0 0,0

0,2

0,4 0,6 Dimensionless time, Dsp∙t/r 2s

0,8

1,0

Figure 9.2 Time evolution of static adsorption in a round particle. Comparison of analytical solution (Eq. (9.19)) and LBEs results.

In calculations, Dsp = 1⋅10−8 m2 /s, r s = 15 l. u. or 75 μm (particle radius), 𝛼 n is the positive root of the Bessel function of the first type of order zero J 0 (r s ⋅𝛼 n ). Only the first three roots have a significant influence on the solution: r s ⋅ 𝛼 1 = 2.405, r s ⋅ 𝛼 2 = 5.52, and r s ⋅ 𝛼 3 = 8.654. The following boundary total | conditions are applied: N total ||r=r = N sat = 540 mol∕m3 and 𝜕N𝜕t | = 0. s |r=0 Time evolution of static adsorption obtained by the analytical solution and the LBEs simulations is shown in Figure 9.2. The results are presented versus the dimensionless time Dsp t∕rs2 . Images in Figure 9.2 show the distribution of an adsorbed amount at N total = 0.4, 0.7, and 0.8. A comparison of the curves reveals satisfactory agreement between the numerical and analytical results. The same insignificant lag of the LBEs results in relation to analytical solution has also been found in the study of Lei et al. (2021) and Zhou et al. (2015), and is explained by the imperfectly round shape of the adsorbent particle formed from rectangular units.

9.2.4 Observation of the Catalyst Distribution in the Pore Space Using 4D Microtomography The four-dimensional X-ray computed tomography is a temporary 3D image of dynamic processes, where “time” is considered the fourth dimension. In the course of the work, a “staged” approach was used, when the filtration processes were suspended for the duration of the scan. For work, a cylindrical sample of Indiana limestone, 5 mm in diameter and 5 mm long, was used. The sample was installed in a thin silicone hose on which tubes supplying liquid were put on both sides. The whole structure was attached to a metal platform-support. A radiolucent aluminum alloy tube was put on top of this structure, which was also attached to the support. Water was supplied to the support in the space between the two tubes from a pressing pump. Liquid was supplied from the pressure pump to the bottom of the inner tube with the sample. Crimping pressure was 2 MPa and temperature −20 ∘ C. The upper end of the inner tube was opened to the atmosphere. The sample was placed in the scanning zone of the tomograph in a core holder that was “transparent” for X-rays. The sample for this study was taken using a nanofocus tube at a voltage of 100 kV

441

442

9 Behavior of Catalyst in Porous Media

and a current of 100 mA. The rest of the survey parameters was adjusted by the operator separately depending on the density characteristics of the minerals that make up the sample. The process of obtaining a three-dimensional image because of X-ray computed tomography consists of three stages: shooting, reconstruction, and processing of models. The voxel model was reconstructed using the phoenix datos | x reconstruction. The resulting volumetric image is an array of voxels saved in the. vol and metadata file in.vgi. Processing and calculation of the characteristics of the three-dimensional model were carried out in the specialized program Avizo 7.1. The original Indiana limestone carbonate sample (135–220 mD) was scanned using micro-CT. The structure and ratio of total and effective porosity are determined. A diagram of the distribution of equivalent diameters for the effective volume of porosity is constructed. The analysis of pore distribution by equivalent diameters is based on the method of dividing pore chambers in the volume of effective porosity by necks (the narrowest points of connection of adjacent pore chambers), measuring the equivalent pore diameter, grouping the obtained pore diameters into separate diameter intervals, and constructing a diagram of the distribution of equivalent pore diameters by intervals, but depending on the share they occupy (in %) of the volume of effective porosity. The algorithm for dividing pores by throats is based on the method of constructing a skeleton (thin frame) inside the pore system, identifying sections on it associated with the central regions of pore chambers, zones of channels connecting adjacent chambers, and dead ends, as well as the subsequent determination of pore boundaries on the lines of zones channels and the final division of the skeleton into groups of lines belonging to the same pore. After that, based on the limited voxel growth algorithm, the effective porosity volume is geometrically divided into separate pore chambers. The√equivalent pore diameter is 6×V

calculated as the diameter of a sphere of the same volume Deq = 3 𝜋 p , where V p is the pore volume. Next, the sample was saturated with a model of formation water with the addition of KI contrast (200 g/l). The sample was rescanned to ensure that the pore space is well saturated. The catalyst solution was pumped into the sample at a rate of ≈0.01 ml/min. The filtering was stopped, and a scan was performed to estimate the saturation of the reservoir. Next, the catalyst solution was displaced with water contrasted with KI salts at a similar rate. At the same time, the displacement was not carried out completely, but partially. Since the catalyst is planned to be used in steam injection, the main goal was to achieve an approximate ratio between formation water and catalyst solution in the sample of 50–50%. The obtained three-dimensional models were combined with each other, and the saturation of the catalyst and water in pores of different equivalent diameters was estimated.

9.3

Results and Discussion

9.3.1

Effect of Heterogeneity Coupled with Peclet Number

This section considers the effect of the pore space heterogeneity on the evolution of the adsorbed phase amount Nttotal , calculated by the Eq. (9.18), under various convection–diffusion conditions described by the Peclet number: Pe = Ly ⋅ u0 /Ds . The study is carried out for a group of porous media having the same porosity 𝜑 = 0.582 and disorder ranging from 0 (completely ordered structure; for example, see Figure 9.1a) to 0.153 (highly heterogeneous media, for example, see Figure 9.1e,f). The size of the digital models is 800 × 400 lattice nodes; Lx = 4 mm and Ly = 2 mm. The flow rate ranges from 0.002 to 0.03 m/s, and the diffusion coefficient ranges from 6.5 to 13⋅10−6 m2 /s. Reaction rate constants and fluid properties are given in Table 9.2.

9.3 Results and Discussion

Table 9.2

Catalyst and water proportions for different steps.

Stage

Catalyst, %

Water, %

Initial saturation

94.3

5.7

Wipe out 10 VP

83.4

16.6

Displacement 30 VP

70.7

29.3

Displacement 50 VP

59.4

40.6

(a)

(b)

0,8

0,6

0,4

0,2 Pe = 0.3 0,0 0

50

100

(c)

200 150 Time (ms)

250

300

0,6

0,4

0,2 Pe = 0.6 0

(d)

0,6

0,4

0,2 Pe = 1.8

100

200 300 Time (ms)

400

500

H=0 H = 0.035 H = 0.092 H = 0.153

1,0

H=0 H = 0.035 H = 0.092 H = 0.153

0,8

0,8

0,0

350

Dimensionless adsorbed amount

Dimensionless adsorbed amount

1,0

H=0 H = 0.035 H = 0.092 H = 0.153

1,0

H=0 H = 0.035 H = 0.092 H = 0.153

Dimensionless adsorbed amount

Dimensionless adsorbed amount

1,0

0,8

0,6

0,4

0,2 Pe = 9 0,0

0,0 0

200

400 Time (ms)

600

800

0

200

400

600 800 Time (ms)

1000

1200

1400

Figure 9.3 Influence of the pore space heterogeneity on the dynamics of the adsorbed phase amount at different Peclet numbers: (a) Pe = 0.3; (b) Pe = 0.6; (c) Pe = 1.8; (d) Pe = 9.

The data presented in Figure 9.3 for the Peclet number in the range from 0.3 to 9 lead to several important findings. First, the pore space heterogeneity significantly affects the dynamics of the adsorption. It was found that an increase in heterogeneity contributes to a deceleration in the adsorption dynamics. The vertical dotted lines in Figure 9.3 indicate the time required to reach Nttotal = 0.98. This parameter is referred to as T sat . For example, at Pe = 0.3 (Figure 9.3a), an increase in the disorder from 0 to 0.153 leads to an increase in T sat from 225 to 350 ms. Second, the sensitivity of the adsorption dynamics to variations in the pore space heterogeneity is the highest at low Pe, when the diffusion process dominates. With an increase in Pe, the effect of heterogeneity on T sat decreases. At high Pe values (the convection mechanism prevails), heterogeneity practically does not affect the dynamics of adsorption. As seen in Figure 9.3a, at low Pe, the difference in T sat between ordered and highly heterogeneous samples reaches 60%, while at high Pe (Figure 9.3d), T sat is approximately the same for all four porous structures.

443

444

9 Behavior of Catalyst in Porous Media

(a)

(b)

Output boundary

H=0

Input boundary

0.14·Lx

Flow direction

(c)

H = 0.035

H = 0.092

0.19·Lx

(d)

H = 0.153

0.28·Lx

0.25·Lx

1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0

Figure 9.4 The distribution of adsorbate concentration at low Pe = 0.3 (diffusion process prevails), detected at t = 180 ms in porous structures with the same 𝜑 = 0.582 and different disorder parameters: (a) H = 0; (b) H = 0.035; (c) H = 0.092; (d) H = 0.153. Vertical yellow dotted lines indicate the position of the concentration front.

(a) Input boundary

Output boundary

(b) H=0

H = 0.035

0.15·Lx

0.15·Lx

Flow direction

(c)

H = 0.092

(d)

0.15·Lx

H = 0.153

0.15·Lx

1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0

Figure 9.5 The distribution of the adsorbate concentration at high Pe = 9 (convection mechanism prevails), detected at t = 1020 ms in the porous structures with the same 𝜑 = 0.582, and different disorder parameters: (a) H = 0; (b) H = 0.035; (c) H = 0.092; (d) H = 0.153. Vertical yellow dotted lines indicate the position of the concentration front.

Figures 9.4 and 9.5 show the adsorbate concentration distributions for low Pe = 0.3 and high Pe = 9, respectively. The concentration fields show that the increasing disorder contributes to the uneven distribution of the adsorbate in the convective–diffusion process. Vertical yellow dotted lines in Figures 9.4 and 9.5 indicate the position of the concentration front. As seen in Figure 9.3, with the prevalence of the diffusion process, an increase in heterogeneity leads to a decrease in the velocity of the concentration front.

9.3.2

Effect of Heterogeneity Coupled with Damkohler Numbers

In this section, we study the influence of the Damkohler number on the dynamics of the adsorption amount. The Damkohler number is estimated by the following relationship: Da = ka CL2y ∕Ds .

9.3 Results and Discussion

(b)

(a) H=0 H = 0.057 H = 0.108 H = 0.145

0,8

H=0 H = 0.057 H = 0.108 H = 0.145

1,0 Dimensionless adsorbed amount

Dimensionless adsorbed amount

1,0

0,6

0,4

0,2

0,8

0,6

0,4

0,2

Da = 250

Da = 750

0,0

0,0 0

100

50

150 200 Time (ms)

(c)

300

350

0,6

0,4

0,2

50

100 Time (ms)

150

200

H=0 H = 0.057 H = 0.108 H = 0.145

1,0 Dimensionless adsorbed amount

0,8

0

(d)

H=0 H = 0.057 H = 0.108 H = 0.145

1,0 Dimensionless adsorbed amount

250

0,8

0,6

0,4

0,2 Da = 6400

Da = 3000 0,0

0,0 0

20

40

60

80 100 Time (ms)

120

140

160

180

0

20

40

60

80 100 Time (ms)

120

140

160

Figure 9.6 Influence of the pore space heterogeneity on the dynamics of the adsorbed phase amount at different Damkohler numbers: (a) Da = 250; (b) Da = 750; (c) Da = 3000; (d) Da = 6400.

The convection–diffusion parameters are u0 = 0.005 m/s and Ds = 50⋅10−6 m2 /s (Pe = 0.2). The Damkohler number is modified by changing the adsorption rate constant from 0.1 to 2.5⋅103 m3 /(mol s). Thus, the Damkohler number varies from 250 to 6400. These values are 10 times higher than in Yu et al. (2019) due to larger sample sizes (2 mm versus 0.2 mm in Yu et al. (2019)). The desorption rate constant is equal to 0.1 s−1 . The numerical simulations are performed for a group of porous media with a size of 800 × 400 lattice nodes, having the same porosity 𝜑 = 0.615, and a disorder parameter ranging from 0 to 0.145. The curves of the adsorption dynamics detected at various Da are presented in Figure 9.6. According to shown data, it is noted that an increase in heterogeneity contributes to a slowdown in the adsorption process. The adsorption dynamics curves indicate an increase in the effect of heterogeneity with an increase in the intensity of mass transfer. The strongest impact of the disorder on the adsorption dynamics was found for the highest Da (Figure 9.6d). T sat for ordered and highly heterogeneous structures varies from 155 to 105 ms (about 50%). With a decrease in Da (Figure 9.6a,b), the sensitivity of the adsorption dynamics to variations in heterogeneity significantly reduces: the difference in T sat reaches only 20 ÷ 23%.

9.3.3

Effect of Heterogeneity Coupled with Porosity

In this section, we study how porosity affects the sensitivity of adsorption dynamics to pore space heterogeneity. The influence of heterogeneity in combination with porosity is investigated for four porosity values equal to 0.52, 0.61, 0.71, and 0.79. For each porosity, four digital models were generated with a disorder parameter ranging from 0 to 0.12 ÷ 0.156. To reveal only the effect of porosity,

445

9 Behavior of Catalyst in Porous Media

(a)

(b) H=0 H = 0.046 H = 0.081 H = 0.121

0,8

1,0 Dimensionless adsorbed amount

Dimensionless adsorbed amount

1,0

0,6

0,4

0,2

φ = 0.52 Pe = 0.6

H=0 H = 0.042 H = 0.104 H = 0.156

0,8

0,6

0,4

0,2

φ = 0.61 Pe = 0.6

0,0

0,0 0

100

200

(c)

300 Time (ms)

400

500

1,0

0,6

0,4

0,2

100

(d)

Dimensionless adsorbed amount

0,8

0

600

H=0 H = 0.036 H = 0.084 H = 0.127

1,0 Dimensionless adsorbed amount

446

φ = 0.71 Pe = 0.6

200 300 Time (ms)

400

500

H=0 H = 0.035 H = 0.092 H = 0.153

0,8

0,6

0,4

0,2

φ = 0.79 Pe = 0.6

0,0

0,0 0

50

100

150

200 250 Time (ms)

300

350

400

0

50

100 150 Time (ms)

200

250

Figure 9.7 Influence of the pore space heterogeneity at Pe = 0.6 on the dynamics of the adsorbed phase amount in samples with different porosities: (a) 𝜑 = 0.52; (b) 𝜑 = 0.61; (c) 𝜑 = 0.71; (d) 𝜑 = 0.79.

numerical simulations for each flow region were performed with the same Pe = 0.6 (u0 = 0.005 m/s, Ds = 16.5⋅10−6 m2 /s). The results are presented in Figure 9.7. The data in Figure 9.7 show that the sensitivity of the adsorption dynamics to variations in heterogeneity strongly depends on porosity. As shown in Figure 9.7, the effect of heterogeneity is most pronounced at the smallest value 𝜑 = 0.52 (Figure 9.7a). T sat changes from 450 to 620 ms in order of increasing disorder. With an increase in porosity (Figure 9.7b, c), the differences in the adsorption dynamics become smaller and almost level at the maximum value 𝜑 = 0.79 (Figure 9.7d): T sat = 255 ms for H = 0.153 versus T sat = 220 ms for H = 0. This finding can be implicitly associated with the results, according to which, with an increase in porosity, the sensitivity of flow properties to changes in disorder also decreases.

9.3.4

Catalyst Distribution in the Pore Space 4D X-Ray CT

The original sample was scanned (Figure 9.8). The total sample porosity was 6.78% (Figure 9.9a), and the effective porosity was 5.81% (Figure 9.9b). The distribution of the fraction of the volume of effective pores from the equivalent diameter is shown in Figure 9.10. They show that the dominant pores in the sample have an equivalent diameter of 0.1–0.4 mm, which accounts for more than 85% of the total effective porosity. Next, the sample was saturated with formation water, and the catalyst solution was filtered. After the sample was completely saturated with the catalyst solution, the catalyst solution was displaced with formation water at a rate of 0.06 ml/min (1 PV/min). Orthoslices of the sample after complete

9.3 Results and Discussion

1 mm

(a)

(b)

1 mm

(c) Figure 9.8

3D rendering (a) and orthogonal slices (b–d) of an Indiana limestone sample.

(a) Figure 9.9

1 mm

(d)

(b)

3D visualization of the total porosity (a) and the effective porosity of the sample (b).

saturation with the catalyst solution (A) and after its partial displacement with formation water through 10 (B), 30 (C), and 50 (D) pore volumes are shown in Figure 9.11. Three-dimensional visualization of the volumes of catalyst and water for these displacement stages are presented in Figure 9.12. The percentage of catalyst and water in the volume of effective porosity for various stages are presented in Table 9.2.

447

9 Behavior of Catalyst in Porous Media

35.00 Volume fraction of effective porosity (%)

448

30.00 25.00 20.00 15.00 10.00 5.00 0.00 0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Equivalent pore diameter (mm) Figure 9.10

Distribution of effective porosity fraction over equivalent pore diameters.

(a)

(b)

(c)

(d)

Figure 9.11 Orthoslices of the sample after stages: (a) after saturation with the catalyst, (b–d) after displacement with formation water after 10 PV, 30 PV and 50 PV, respectively (the catalyst is lighter, the water is darker).

9.3 Results and Discussion

(a)

(b)

(c)

(d)

Figure 9.12 Distribution of the catalyst (green) and formation water (blue) for different stages of the experiment: (a) after saturation with the catalyst, (b–d) after displacement with formation water after 10, 30, and 50 PV, respectively.

Fraction of effective porosity (%)

35 30

Pores

Catalyst

25 20 15 10 5 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Equivalent diameter (mm) Figure 9.13

Distribution of catalyst fraction after saturation by equivalent pore diameters.

The distribution of catalyst fractions by equivalent effective porosity diameters is shown in Figures 9.13–9.16. It demonstrates that there are two main trends in the displacement of the catalyst solution by the formation of water. First, more efficient displacement comes from larger pores. For example, a catalyst solution from pores with an equivalent diameter of 0.35–0.5 mm turned out to be almost completely displaced. Second, active displacement occurs from the pores having the maximum share in the effective porosity from 0.15 to 0.35 mm.

449

9 Behavior of Catalyst in Porous Media

Fraction of effective porosity (%)

35 30

Pores

Catalyst

25 20 15 10 5 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Equivalent diameter (mm) Figure 9.14

Distribution of catalyst fraction after displacement of 10 PV by equivalent pore diameters.

35 Fraction of effective porosity (%)

450

30

Pores

Catalyst

25 20 15 10 5 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Equivalent diameter (mm) Figure 9.15

Distribution of catalyst fraction after displacement of 20 PV by equivalent pore diameters.

Thus, a similar effect on the catalyst solution in the rock can be expected when steam is pumped through the sample. First, the displacement will go through the largest pore channels and pores of the most common sizes. The distribution of the catalyst generally follows the pore equivalent diameter distribution curve, which means that it will persist in all pores with a slight bias toward the smallest pores.

9.4

Conclusion

This chapter presents a systematic numerical study of the influence of pore space heterogeneity on the dynamics of the adsorbed phase amount. The effect of heterogeneity was investigated under various Damkohler numbers, convection–diffusion conditions described by the Peclet number, as well as in its combination with porosity. The best findings are summarized as follows:

References

Fraction of effective porosity (%)

35 30

Pores

Catalyst

25 20 15 10 5 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Equivalent diameter (mm) Figure 9.16

Distribution of the catalyst fraction after displacement of 30 PV by equivalent pore diameters.

(1) The dynamics of the adsorbed phase amount are significantly influenced by pore space heterogeneity. An increase in disorder promotes a decrease in the velocity of the concentration front and a slowdown in the adsorption dynamics. (2) Depending on the Peclet number and porosity, the response of the adsorption dynamics to a change in heterogeneity is different: (a) The sensitivity of the adsorption dynamics to variations in the disorder parameter is greatest when the diffusion mechanism is dominant (low Pe) and significantly decreases as the Peclet number increases. When the convection mechanism prevails (high Pe), the adsorption dynamics is practically unaffected by heterogeneity. (b) The effect of heterogeneity is most pronounced at low porosity (𝜑 ≈ 0.5) and decreases significantly with its growth. At high porosity (𝜑 ≈ 0.8), the influence of heterogeneity was not revealed. (3) The Damkohler number influences the sensitivity of adsorption dynamics to changes in pore space heterogeneity. The effect of heterogeneity increases with the increasing intensity of mass transfer. (4) Methodology for estimating the distribution of a catalyst in a porous medium using 4D X-ray CT is presented.

References Ancheyta, J., Angeles, M.J., Macías, M.J. et al. (2002). Changes in apparent reaction order and activation energy in the hydrodesulfurization of real feedstocks. Energy Fuels 16: 189–193. Chen, L., Kang, Q., He, Y.-L., and Tao, W.-Q. (2012). Pore-scale simulation of coupled multiple physicochemical thermal processes in micro reactor for hydrogen production using lattice Boltzmann method. International Journal of Hydrogen Energy 37: 13943–13957. Chen, L., He, A., Zhao, J. et al. (2022). Pore-scale modeling of complex transport phenomena in porous media. Progress in Energy and Combustion Science 88: 100968.

451

452

9 Behavior of Catalyst in Porous Media

Kang, Q., Lichtner, P.C., and Zhang, D. (2007). An improved lattice Boltzmann model for multicomponent reactive transport in porous media at the pore scale. Water Resources Research 43: W12S14. Lei, J., Ding, R., and Zhang, J. (2021). Study on seepage and adsorption characteristics of porous media containing adsorbent based on lattice Boltzmann. AIP Advances 11: 045126. Liu, R., Yao, S., Li, Y., and Cheng, J. (2019). Pore-scale study of dynamic ion adsorption process in porous electrodes of capacitive deionization using lattice Boltzmann method. International Journal of Heat and Mass Transfer 135: 769–781. Lv, X., Liang, X., Xu, P., and Chen, L. (2020). A numerical study on oxygen adsorption in porous media of coal rock based on fractal geometry. Royal Society Open Science 7: 191337. Machado, R. (2012). Numerical simulations of surface reaction in porous media with lattice Boltzmann. Chemical Engineering Science 69: 628–643. Pan, C., Luo, L.S., and Miller, C.T. (2006). An evaluation of lattice Boltzmann schemes for porous medium flow simulation. Computers and Fluids 35: 898–909. Succi, S. (2001). The Lattice Boltzmann Equation for Fluid Dynamics and Beyond, 290. UK: Oxford University Press. Sullivan, S.P., Sani, F.M., Johns, M.L., and Gladden, L.F. (2005). Simulation of packed bed reactors using lattice Boltzmann methods. Chemical Engineering Science 60: 3405–3418. Vanson, J.-M., Coudert, F.-X., Rotenberg, B. et al. (2015). Unexpected coupling between flow and adsorption in porous media. Soft Matter 11: 6125–6133. Wang, H., Qu, Z.G., and Zhou, L. (2017). A combined GCMC and LBM simulation method for CH4 capture in Cu-BTC particle adsorption bed. International Communications in Heat and Mass Transfer 88: 48–53. Wang, H., Bai, J.Q., Qu, Z.G., and Chang, M. (2019). Numerical study of heat transfer in process of gas adsorption in a Cu-benzene-1, 3, 5-tricarboxylic acid particle adsorption bed. International Journal of Hydrogen Energy 44 (23): 11989–12002. Yu, X., Regenauer-Lieb, K., and Tian, F.-B. (2019). A hybrid immersed boundary-lattice Boltzmann/finite difference method for coupled dynamics of fluid flow, advection, diffusion and adsorption in fractured and porous media. Computers and Geosciences 128: 70–78. Zaafouri, Z., Batôt, G., Nieto-Draghi, C. et al. (2021). Lattice Boltzmann method for adsorption under stationary and transient conditions: interplay between transport and adsorption kinetics in porous media. Physical Review E 104: 015314. Zakirov, T.R., Galeev, A.A., Statsenko, E.O., and Khaidarova, L.I. (2018). Calculation of filtration characteristics of porous media by their digitized images. Journal of Engineering Physics and Thermophysics 91 (4): 1069–1078. Zakirov, T.R. and Khramchenkov, M.G. (2022). Effect of pore space heterogeneity on the adsorption dynamics in porous media at various convection-diffusion and reaction conditions: A lattice Boltzmann study. Journal of Petroleum Science and Engineering 212: 110300. Zhou, L., Qu, Z.G., Chen, L., and Tao, W.-Q. (2015). Lattice Boltzmann simulation of gas–solid adsorption processes at pore scale level. Journal of Computational Physics 300: 800–813. Zhou, L., Qu, Z.G., Ding, T., and Miao, J.Y. (2016). Lattice Boltzmann simulation of the gas-solid adsorption process in reconstructed random porous media. Physical Review E 93: 043101.

453

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process Allan Rojas 1 , Denis Shevchenko 1,2 , Vladislav Sudakov 1 , Sergey Usmanov 1 , and Michael Kwofie 1 1 2

Institute of Geology and Petroleum Technologies, Kazan Federal University, Kremlyovskaya str. 18, Kazan 420008, Russia Department of Higher Mathematics, Kazan Innovative University named after V. G. Timiryasov, Moskovskaya 42, Kazan, Russia

10.1 The Reaction Scheme A reaction model within the context of an oil system is a model which best represents the changes in oil composition at a specific reaction temperature, pressure, time, and concentration conditions. The reaction models are valuable tools that allow the prediction of changes in oil composition when combined with analytical kinetic methods to later be integrated into a numerical simulation model. The accuracy of the predictions of oil composition depends on the thoroughness and sophistication of the reaction model itself. Regarding aquathermolysis, most of the published numerical simulation models point out toward the overall changes in the liquid phase composition. These changes are mostly represented by variations in SARA composition and the gas formation (i.e. light hydrocarbons, CO2 , H2 S, and H2 ) (Hyne 1986; Kapadia et al. 2012; Ayache et al. 2015; Huang et al. 2017). As the aquathermoylsis window is 200–300 ∘ C (Hyne 1986), products in the solid phase are not included in these models due to the fact that they occur at higher temperature ranges, during thermal cracking reactions. Consequently, after satisfying the condition of establishing the reaction model, it is suitable to proceed with the definition of the equation of state (EoS) parameters in a process called oil characterization (see Section 10.2.1). In numerical simulation, the oil characterization defines the oil fractions, the distribution of its composition, and their EoS parameters. The oil fractions participate under the nomination of components. When dealing with reaction systems, the components or grouping of components of an unreacted oil are usually described in one of the, but not restricted to, following general ways: (i) by the content of its distillation fractions at the respective boiling point where volatiles C1 —C4 (0–30 ∘ C), naphtha C5 —C10 (30–180 ∘ C), kerosene C10 —C16 (180–260 ∘ C), gas oils C16 —C60 (260–350 ∘ C), lubricants >C60 (350–575 ∘ C), fuel oil >C70 (>490 ∘ C), asphalt >C80 (>580 ∘ C) or (ii) by the content of saturates, aromatics, resins, asphaltenes (SARA) fractions. The correct characterization of the components of the unreacted oil and fitting of PVT experimental data to the EoS leads to more reliable predictions of oil phase behavior at any pressure and temperature. While performing material balance operations of a reaction system as for aquathermolysis processes in numerical simulation, two main conditions must be compiled: (i) the properties of arranged number and grouping of components of a fluid model such as molecular weight (MW), specific gravity (SG), and boiling point (T b ), must honor and match those properties previously defined in the reaction model (Figure 10.1). Commonly, the average MW of a nonpure component Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

454

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Reactants

Products

H2 H2S S A R A Coke

S MW = w A MW = x R MW = y A MW = v

MW = t MW = u MW = v1 MW = w1 MW = x1 MW = y1 MW = z

Components

H2 H2S S A R A Coke

MW = t MW = u MW = v MW = w MW = x MW = y MW = z

Aquathermolysis

Reaction model Figure 10.1

Fluid model

Scheme of conditions for a creation of a fluid model to simulate a reaction system.

in a reacted oil differs from the unreacted one. However, obtaining these data may become time-consuming with dependence on the number of components and reactions in a certain reaction system. For this reason, in numerical simulation, it is usual practice to assume the same properties of the unreacted and reacted oil components; (ii) the fluid model must include the product components of the previously defined reaction model (Figure 10.1). Some components, such as light hydrocarbons, H2 , H2 S, and CO2 , may be absent in an unreacted oil but may be formed during chemical reactions, thus their integration into the components list is of high priority. After the successful definition of the reaction model and replication of the phase behavior of the unreacted oil, two main advantages can be leveraged with numerical simulation: history matching of experimental runs (i.e. kinetic cells, autoclave reactors, laboratory-scale displacement facilities, etc.) and forecasting of a reaction system using a reservoir 3D field-scale numerical model. Forecasts can allow to estimate the economics of a pilot EOR project when supported by low-uncertainty data.

10.2 Modeling the Phase Behavior of Oil This section discusses the key facts during oil characterization for numerical simulation purposes. This includes the most used correlations for the estimation of critical parameters of the C7+ fraction and prediction of the dead oil viscosity with temperature dependence. Moreover, the main matching EoS parameters are discussed in detail along with a systematic workflow for its achievement.

10.2.1

Oil Characterization

The oil characterization process is needed when working with any EoS and consists of the representation of a petroleum reservoir fluid with the correct number of pure components, properties of pseudocomponents, and composition. The pure hydrocarbon component is a well-defined component that can be nonhydrocarbon (i.e. CO, CO2 , H2 S, etc.) or hydrocarbon, often called single carbon number (SCN) (i.e. C1 , C2 , C3 , etc.). They are usually available from an extensive data base integrated in each simulator. A pseudocomponent is a component which is not a pure component. A pseudocomponent is the mixture of two or more SCN. A well-known pseudocomponent is the C7+ fraction. The properties of the C7+ fraction such as T b , MW, and SG vary from oil to oil. For this reason, the properties of the C7+ fraction are not incorporated into the internal library of numerical

10.2 Modeling the Phase Behavior of Oil

simulators. As additional properties other than T b , MW, and SG are required to fulfill an EoS, they are estimated via in-built correlations (see Section 10.2) within the simulator. The C7+ fraction is the largest pseudocomponent which demands its extension and splitting into several fractions using split functions (i.e. exponential or gamma distribution) (Whitson et al. 1990). The main challenge of these techniques is to identify the right extension and the molar distribution of the split fractions. The C7+ fraction is usually not only extended up to C21+ but also extended up to C135+ (Gutiérrez et al. 2018) and C200+ (Pedersen et al. 2004). As the number of extended SCN can be quite large and impractical, it is usually grouped into a certain range of SCN (e.g. C2 —C6 , C7 —C19 , C20 —C80 or any) or also called multicarbon number (MCN). Some numerical fluid models involved in reaction systems are defined by the use of pseudocomponents based on SARA (Gutiérrez et al. 2018); maltenes and asphaltenes (Rojas et al. 2021); naphta, distillates, VGO, and VR (Nguyen et al. 2017). One key feature of these types of models is that their composition is mostly made-up of the C7+ fraction. The relative importance of the correct characterization of the C7+ fraction is on properly honoring the T b , MW, and SG of a certain pseudocomponent or MCN. The correct oil characterization and definition of the EoS parameters may enable the fitting of the PVT experimental data and predict the phase behavior of oil (see Section 10.2.4) Usually, the characterization process involves the following three general steps (Hong 2007; Varzandeh et al. 2017): ●

●

●

Estimating the relation between carbon number or molecular weight and molar composition distribution; Estimating parameter values of the EoS (i.e. critical properties such as critical temperature, T c and critical pressure, Pc , and acentric factor, 𝜔) and the binary interaction coefficients (K ij ) for each pair of pseudocomponents; Lumping of oil fractions into MCN.

10.2.2

Correlations for Property Estimation of Hydrocarbons

10.2.2.1 Critical Parameters 10.2.2.1.1 Twu’s Method

Twu (1984) derived correlations that accurately predict the critical and physical parameters of pure hydrocarbons only as a function of the boiling point by using the family of n-alkanes in the range of C1 H4 up to C100 H202 as the reference system. The correlations exclude values for methane’s critical pressure. Based on a perturbation–expansion technique, a method was derived to predict the critical and physical parameters of pseudocomponents (Eqs. (10.1)–(10.14)). These can be estimated by the following steps: Step 1 Choose the boiling point of a n-Alkane that represents the boiling point of the pseudocomponent. Step 2 Use data of step 1 and solve X cnA by using 𝛽 coefficients presented in Tables 10.1–10.4. Step 3 Solve f x , and 𝛥𝛾 x to obtain X cps . Table 10.1

Coefficients for estimation of critical temperature of pseudocomponents by Twu’s method. 𝜷0

𝜷1

𝜷2

𝜷3

𝜷4

T cnA

0.533272

0.191017E-3

0.779681E-7

0.284376E-10

0.959468E-28

ft

0.362456

0.0398285

0.948125

—

—

455

456

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Table 10.2

Coefficients for estimation of critical pressure of pseudocomponents by Twu’s method. 𝜷0

𝜷1

𝜷2

PcnA

3.83354

1.19629

fp

2.53262

46.1955

34.8888 0.00127885

𝜷3

𝜷4

𝜷5

36.1954

104.193

—

11.4277

252.140

0.00230535

Table 10.3 Coefficients for estimation of critical volume of pseudocomponents by Twu’s method. 𝜷0

𝜷1

𝜷2

𝜷3

V cnA

0.419869

0.505839

1.5643

9481.70

fv

0.466590

0.182421

3.01721

—

Table 10.4 Coefficients for estimation of specific gravity of pseudocomponents by Twu’s method.

𝛾 nA

𝜷0

𝜷1

𝜷2

𝜷3

0.843593

0.128624

3.36159

13 749

The Critical Temperature The critical temperature is the maximum temperature at which a compound can exist in liquid phase: [( )]2 1 + 2ft Tcps = TcnA (10.1) 1 − 2 ft Tb (10.2) TcnA = ( ) 𝛽0 + 𝛽1 Tb + 𝛽2 Tb2 − 𝛽3 Tb3 + 𝛽4 Tb13 ( [ ) ] 𝛽0 𝛽2 (10.3) ft = Δ𝛾t − 1∕2 + 𝛽1 − 1∕2 Δ𝛾t Tb Tb

Δ𝛾t = exp[5(𝛾nA − 𝛾ps )] − 1

(10.4)

The Critical Pressure The critical pressure is the minimum pressure at which a compound can be

liquified at the critical temperature: ]2 [ PcnA Tcps VcnA (1 + 2fp ) Pcps = TcnA Vcps (1 − 2fp ) PcnA = (𝛽0 + 𝛽1 𝛼 1∕2 + 𝛽2 𝛼 + 𝛽3 𝛼 2 + 𝛽4 𝛼 4 )2 T 𝛼 =1− b TcnA ⎡⎛ ⎞ ⎛ ⎞ ⎤ 𝛽 𝛽 fp = Δ𝛾p ⎢⎜𝛽0 − 11 − 𝛽2 Tb ⎟ + ⎜−𝛽3 + 41 + 𝛽5 Tb ⎟ Δ𝛾p ⎥ ⎢⎜ ⎟ ⎜ ⎟ ⎥ Tb2 Tb2 ⎣⎝ ⎦ ⎠ ⎝ ⎠ Δ𝛾p = exp[0.5(𝛾nA − 𝛾ps )] − 1

(10.5) (10.6) (10.7) (10.8) (10.9)

10.2 Modeling the Phase Behavior of Oil

The Critical Volume The critical volume is the volume of a mass unit of a gas or vapor when found

in its critical state:

[

Vcps = VcnA

(1 + 2fv ) (1 − 2fv )

]2 (10.10)

VcnA = [1 − (𝛽0 − 𝛽1 𝛼 − 𝛽2 𝛼 3 − 𝛽3 𝛼 14 )]−8 ⎡ 𝛽 ⎛ 𝛽 0 + ⎜−𝛽1 + 21 fv = Δ𝛾v ⎢ 1∕2 ⎢T ⎜ Tb2 ⎣ b ⎝ [ ( 2 )] 2 Δ𝛾v = exp 4 𝛾nA − 𝛾ps −1

(10.11)

⎞ ⎤ ⎟ Δ𝛾 ⎥ ⎟ v⎥ ⎦ ⎠

(10.12) (10.13)

The Specific Gravity The specific gravity is the density of oil relative to that of water:

𝛾nA = 𝛽0 − 𝛽1 𝛼 − 𝛽2 𝛼 3 − 𝛽3 𝛼 12 10.2.2.1.2

(10.14)

Kesler–Lee’s Method

Kesler and Lee (1976) developed correlations for the estimation of critical parameters of hydrocarbons by using the boiling point and specific gravity as input data. These can be obtained by using Eqs. (10.15) and (10.16) and the following steps: Step 1 Obtain the specific gravity and the boiling point of the pseudocomponent. Step 2 Use data of step 1 to solve X cps and by using 𝛽 coefficients presented in Tables 10.5 and 10.6. The Critical Temperature

Tcps = 𝛽0 + 𝛽1 𝛾ps + [𝛽2 + 𝛽3 𝛾ps ]Tb +

[𝛽4 − 𝛽5 𝛾ps ]105

(10.15)

Tb

The Critical Pressure

[ [ ] ] 𝛽3 𝛽6 𝛽7 𝛽1 𝛽4 −3 ln(Pcps ) = 𝛽0 − − 𝛽2 + + + 10 Tb + 𝛽5 + 10−7 Tb2 𝛾ps 𝛾ps 𝛾ps 2 𝛾ps 𝛾ps 2 ] [ 𝛽9 − 𝛽8 + 2 10−10 Tb3 𝛾ps

(10.16)

10.2.2.2 Dead Oil Viscosity with Temperature Dependence

The oil viscosity 𝜇 is one of the most important parameters in oil recovery and enhanced oil recovery processes. The oil viscosity has dependence on temperature, and in the cases of data unavailability, it can be predicted by empirical correlations (Eqs. (10.17)–(10.24): 141.5 − 131.5 (10.17) API = 𝛾 𝜌 γ= 0 (10.18) 𝜌w Table 10.5 Coefficients for estimation of critical temperature of pseudocomponents by Kesler–Lee’s method.

T cps

𝜷0

𝜷1

𝜷2

𝜷3

𝜷4

𝜷5

341.7

811.1

0.4244

0.1174

0.4669

3.26238

457

458

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Table 10.6 Coefficients for estimation of critical pressure of pseudocomponents by Kesler–Lee’s method.

Pcps

10.2.2.2.1

𝜷0

𝜷1

𝜷2

𝜷3

𝜷4

8.3634

0.0566

0.24244

2.2898

0.11857

𝜷5

𝜷6

𝜷7

𝜷8

𝜷9

1.4685

3.648

0.47227

0.42019

1.6977

Beal’s Method (Beal 1946)

( 𝜇o =

1.8(10)7 0.32 + API 4.53

a = 10 10.2.2.2.2

)a 360 Tp − 260

10(4.9563−0.00488Tp ) (T ) )2.709 API + 30p − 14.29

(10.20)

(10.21)

Hossain’s Method (Hossain et al. 2005)

𝜇o = 10(−0.71523API+22.13766) Tp (0.269024API−8.268047) 10.2.2.2.4

(10.19)

Al-Khafaji’s Method (Al-Khafaji et al. 1987)

𝜇o = (

10.2.2.2.3

( ) 0.43+ 8.33 API

)(

(10.22)

Bergman’s Method (Bergman and Sutton 2009)

This method requires at least two measurements at two different temperatures. These two temperatures are at 100 and 210 ∘ F. Extrapolation of viscosity at temperatures lower than 100 ∘ F are not well corresponding. However, at temperatures higher than 210 ∘ F, the estimations are satisfactory, regardless of the oil type: 𝜇o = exp(exp{ln[ln(𝜇100 + 1)] + B[ln(Tp + 310) − ln(410)]}) − 1

(10.23)

ln[ln(𝜇210 + 1)] − ln[ln(𝜇100 + 1)] (10.24) B= ln(520) − ln(410) The rheological parameters of a real oil called oil A are provided in Table 10.7. By substituting the values of Table 10.7 into the correlations provided above (Eqs. (10.17)–(10.24)), a plot of predicted versus experimental data is illustrated in Figure 10.2. Figure 10.2 shows that the Bergman’s correlation has the best fitting of experimental data for the dead oil viscosity with temperature dependence. It has a good fitting above 100.4 ∘ F/38 ∘ C and tends to offset at lower temperatures. Other correlations showed low fitting to experimental data for the presented oil sample.

10.2.3

Special Data Requirement

10.2.3.1 Oil Viscosity with Temperature Dependence

The viscosity is defined as the resistance of a fluid to flow. Usually, the viscosity is given in mPa*sec or centipoise (cP). The magnitude of 𝜇 in situ is a function of the reservoir’s pressure (P) and

10.2 Modeling the Phase Behavior of Oil

Table 10.7

Rheological properties of oil A.

Property

Oil density (kg/m3 ) ∘ API Temp. (∘ F)

Value

967 14.82 Viscosity (cP)

77

11 070

86

9785

89.6

8883

93.2

8174

96.8

7422

100.4

6464

104

5790

113

4248

122

2948

131

2142

140

1528

158

858.5

176

511.6

194

310

212

206.2

230

148.4

248

105.5

302

30.63

356

15.96

temperature (T) conditions and gas in solution (Rs ). At standard conditions, the oil viscosity magnitude increases with the oil density. According to P, the oil viscosity can be classified in two groups: ●

●

Pressure > 1 atm ⚬ Saturated oil viscosity (Rs @ P=T b and T = T res ) ⚬ Unsaturated oil viscosity (Rs @ P > T b and T = T res ) Pressure = 1 atm ⚬ Dead oil viscosity (no Rs )

Conventionally 𝜇 o is obtained by Pressure-Volume-Temperature (PVT) analysis such as differential Liberation, Constant Volume Depletion (CVD), Constant Composition Expansion (CCE) (Pedersen and Christensen 2006) at reservoir P and T to preserve in situ conditions, specially, for the case of live oils, where there is Rs . However, it is common practice to measure the dead oil viscosity at increased temperatures by using rheological devices such as the bob and cup and the cone and plate viscometers. The oil viscosity-temperature curve has the tendency to decrease in a quadratic manner as the oil temperature increases. The relevance of its accurate experimental measurement within a certain range of temperatures relies on that the reduction of 𝜇 is the main recovery mechanism of heavy oils. Also, when there is lack of experimental analysis results, some available published correlations may be used (see Section 10.2.2.2).

459

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

12 000 10 000 Predicted viscosity (cP)

460

8000 6000

4000 2000 0

0

2000

8000 4000 6000 Experimental viscosity (cP)

Bergman Figure 10.2

10.2.4

Hossain

Beal

10 000

12 000

A1-Khafaji

Predicted versus experimental viscosity of oil A.

The Cubic EoS and Phase Behavior

10.2.4.1 The Peng and Robinson EoS

Peng and Robinson (1976) proposed a two-constant EoS. It is a well-accepted expression (Eq. (10.25)) that uses the critical parameters as critical pressure and temperature as well as the acentric factor to calculate the parameters “a” and “b” (Eqs. (10.26–10.29) and Table 10.8) which will ultimately supply the EoS to reproduce the phase behavior and volumetric properties of a fluid mixture. The two-constant PR EoS predicts the vapor-liquid equilibrium (VLE) values of hydrocarbon mixtures with satisfactory accuracy, but the volumetric predictions have an off-set from experimental data. An EoS has two main tasks: (1) Match the phase behavior and molar phase composition (2) Match phase density a RT − v − b v(v + b) + b(v − b) R2 Tc2 a = 𝛺a 𝛼 Pc RT b = 𝛺b c Pc [ ( √ )]2 T 𝛼 = 1+m 1− Tc

(10.25)

P=

(10.26) (10.27) (10.28)

m = m + m1 𝜔 − m2 𝜔2 ; 𝜔 < 0.49

(10.29)

m = 0.3796 + 1.485𝜔 − 0.1644𝜔 + 0.0166𝜔 ; 𝜔 > 0.49 2

3

(10.30)

10.2 Modeling the Phase Behavior of Oil

Table 10.8 Constants for parameters 𝛺 and m of PR EoS. Constant

Value

𝛺a

0.45724

𝛺b

0.0778

m (0.49)

−0.1644

m3 (>0.49)

0.0166

The mixtures of the parameters “a” and “b” are determined by ∑∑ 1∕ 1∕ xi xj (1 − 𝛿ij )ai 2 aj 2 a= i

b=

(10.31)

j

∑ xi bi

(10.32)

i

ai and aj are the PR EoS constants of components “i” and “j,” respectively (Peng and Robinson 1977). 10.2.4.2 Binary Interaction Coefficients

As the saturation pressure is a key parameter that describes the energy of an oil system, it is used for reliable predictions of oil phase behavior calculations. The energy of a system consists of the sum of the individual energy of each component in the system and is given by their chemical energy. In an oil system, the molecules that conform the system can interact differently due to their distinctive chemical energy. The effect of this variation in chemical energy has an impact on the saturation pressure. The binary interaction coefficient between a component i and j was introduced as a fitting parameter of the saturation pressure experimental data, whether in the bubble point or retrograde dew point region of oil mixtures by implementing it into the PR EoS (Eq. (10.33)). Then, the calculated binary coefficient 𝛿 ij is introduced into “a” of the EoS as in Eq. (10.31). This coefficient has proved to produce accurate results during fitting of experimental fluid behavior data (Peng and Robinson 1976, 1977; Katz and Firoozabadi 1978; Whitson 1983): n √ ⎡ 2 v 1∕3 ∗ v 1∕3 ⎤ ci ci ⎥ ⎢ (10.33) 𝛿ij = 1 − ⎢ ⎥ 1∕3 1∕3 ⎥ ⎢ vci + vcj ⎦ ⎣ The binary interaction coefficients are given for the Hydrocarbon-Hydrocarbon (HC—HC) and the Hydrocarbon-Nonhydrocarbon (HC_NHC) systems. Usually, the HC—HC binary interaction coefficients for well-defined components as the C1 —C6 are set to zero. Other HC—HC binary interaction coefficients as for the C7+ fraction and all its split pseudocomponents, it is recommended to couple with C1 (Figure 10.3). These coefficients must be found by trial and error technique which

461

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

10 000 9000 8000 Pressure (KPa)

462

7000 6000 5000 4000 3000 2000 1000 0

0

0,2

0,4 0,6 CO2 mole fraction

0,8

Liquid @71.17 ˚C

Vapor @71.17 ˚C

Model - Int. coeff. = 0.125

Model - Int. coeff. = 0

1

Figure 10.3 Dependence of methane binary interaction coefficient on SCN of the C+ fraction. Source: Adapted from Whitson (1983).

commonly is performed by mathematical regression operations. The binary interaction coefficients from C2 to C6 with the C7+ fraction are set to zero (Pedersen et al. 2004). On the other hand, the binary interaction coefficients between methane and C7+ fraction increase with the SCN as depicted in Figure 10.4. As nonhydrocarbon components are typical gases found in a reservoir (i.e. N2 , CO2 , H2 S), the HC—NHC binary interaction components are taken into account. The binary interaction coefficients of these gases with light hydrocarbons have shown their importance while matching experimental phase behavior using the PR EoS as illustrated in Figure 10.4 (Peng 1986). Since these gases are well-defined components, the reported values of binary interaction coefficients can be used in any phase behavior simulation Table 10.9. 10.2.4.3 Volume Translation

The original two-constant PR EoS overpredicts liquid densities at lower temperature and underpredicts at high temperatures in comparison to the experimental data (Peng and Robinson 1976). Liquid densities are predicted with error in the gas phase z factors up to 5% and liquid density up to 12% (Jhaveri and Youngren 1988). The volume translation technique was initially proposed for the Soave–Redlich–Kwong (SRK) EoS to improve the liquid density predictions. A third parameter was derived and introduced into the two-constant EoS (Péneloux et al. 1982). The third parameter “Ci ” or also known as volume shift is represented in Eq. (10.34). This parameter is used to adjust the original EoS calculated volumes to a new and more corresponding prediction with the actual experimental data. The method works well for pure components except in the near critical point. The EoS accurately predicts the equilibrium constants with acceptable volumetric predictions for the vapor and liquid phases when using the volume translation technique. The advantage of using the volume translation technique relies on the fact that it does not alter the equilibrium conditions (saturation pressure or temperature or composition) but has an effect on the fluid density, compressibility factors, and molar volumes which are not irrelevant terms in reservoir engineering

10.2 Modeling the Phase Behavior of Oil

Volume shift (dimensionless)

0.1

Xylene Toluene

0.05

Benzene

Octane

0 Methylcyclopropane

Nonane Decane

–0.05 Methylcyclobutane

–0.1 Methylcyclopentane

–0.15 –0.2

20

0

40

60

80 100 MW (g/mol)

120

140

160

Figure 10.4 Effect of binary coefficients in phase behavior matching of iC5 -CO2 system. Source: Experimental Adapted from Besserer and Robinson (1975). Table 10.9

Binary interaction coefficients HC-NHC components. C1

C2

C3

C4

C5

nC6

nC7

N2

0.036

0.05

0.05

0.09

0.1

0.1

0.1

CO2

0.1

0.13

0.135

0.13

0.125

0.125

0.1

H2 S

0.085

0.084

0.075

0.06

0.065

0.06

0.06

Source: Adapted from Peng (1986).

(i.e. compressibility, GOR, and mass density). Therefore, the third parameter is important and should be taken with care: V = V EoS − c ∑ xi ci

(10.34)

N

VL = VLEoS −

(10.35)

i=1 N

Vv = VvEoS −

∑ yi ci

(10.36)

i=1

While performing tasks of numerical PVT simulation, the actual values of volume shift are those given by s (10.37), which searches for c. The values of s are negative for low molecular weight of well-defined components and become positive at higher molecular weighs (Figure 10.5). It has been reported while matching the experimental results on phase behavior of pure or well-defined hydrocarbon components (i.e. alkanes) (Jhaveri and Youngren 1988). For this reason, the volume shift values of pure components can be found in built-in libraries of numerical simulators. Opposite to the pure oil components, the C7+ fraction is a type of undefined mixture of oil and its composition and physicochemical properties (i.e. density, average molecular weight, etc.) can vary in a wide range of values. In this sense, the volume shift coefficient of the C7+ fraction is not available in built-in libraries of simulators and can be obtained numerically for each oil after mathematical regression operations. In order to approximate the volume shift coefficient of the C7+ fraction, the

463

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Methane binary interaction coefficient (Dimensionless)

464

0.07 0.06 0.05 0.04 0.03 0.02 0.01 0

10

0

20

30

40

50

SCN Figure 10.5

Estimated volume shifts versus MW for selected hydrocarbons.

Table 10.10

Volume shift correlation coefficients for C7+ hydrocarbons.

Component type

Correlation coefficient

Deviation (%)

d

e

n-alkanes

2.258

0.1823

0.19

n-alkylcyclohexanes

3.004

0.2324

0.28

n-alkylbenzenes

2.516

0.2008

0.24

Source: Jhaveri and Youngren (1988)/Society of Petroleum Engineers.

Eq. (10.38) can be used for n-alkanes, n-alkylcyclohexanes, and n-alkylbenzes and its corresponding coefficients in Table 10.10 (Jhaveri and Youngren 1988): c (10.37) s= b d s=1− e (10.38) M During any of the thermal process as in aquathermolysis, which takes place at high temperatures, thermal expansion of fluid occurs. The thermal expansion of stable oils is underpredicted with the original PR EoS. According to the ASTM 1250-80, the density of a stable oil can be determined at higher temperatures. In this manner, a temperature-dependent volume-shift term was introduced as shown in Eq. (10.39) to allow for better volumetric predictions at high temperatures (Pedersen et al. 2004): si = s + s1i (T − Tref )

(10.39)

10.2.4.4 Tuning an EoS

When referring to PVT numerical simulation, the term “tuning” is related to the adjustment of certain parameters in order to accurately match experimental data to an EoS for reliable predictions of fluid phase behavior. The task of tuning commonly involves the adjustment of the following EoS parameters:

10.2 Modeling the Phase Behavior of Oil

(1) (2) (3) (4)

The critical parameters Binary interaction coefficients Volume shift Molecular weight and composition distribution

Since parameters of well-defined components is redundantly well known, these parameters are tuned for the undefined pseudocomponents only, as those of the C+ fraction-type. The parameters are computed with a mathematical regression technique that uses an objective function with its general expression as presented in Eq. (10.40). The regressed parameters search for the optimal value by previous user-given thresholds: [ ]2 ∑ (xi,calc − xi,exp ) R= (10.40) wi xi,exp j=1 Some EoS tuning techniques have been published as the one by Coats (1985) called the pseudoization method that discards the mixing rules effects on the properties of pseudocomponents and the stepwise procedure by Whitson and Brule (2000). However, the general procedure to tune into an EoS can be summarized as follows: Step 1 Match saturation pressure experimental data by regressing T c and Pc for each MCN group and the 𝛿 ij between methane and each of the MCN groups. Step 2 Match oil density, GOR, compressibility experimental data by regressing s for each MCN group. Step 3 Match oil viscosity experimental data by regressing MW. Step 4 Lump SCN components and MCN pseudocomponents after successful EoS tuning until the new lumping scheme misrepresents results obtained in step 3. The practical use of the tuning workflow above can be illustrated by predicting the oil phase behavior of oil B, using experimental data provided in Table 10.11 and Figure 10.6. To match the experimental saturation pressure (379.212 KPa at 62.77 ∘ C), it is necessary to tune the binary interaction coefficients as in Eq. (10.33) by making use of regression tools. The results after regression that match the experimental data are shown in Table 10.12. Next, it is necessary to match oil density, ROV, and liquid%, which requires the tuning of the volume shift. As can be seen in Figure 10.6, a successful matching of PVT data can be achieved at all pressure regimes by using the parameters provided in Table 10.13. Table 10.11

EoS parameters of oil B with 11∘ API gravity. T c (K)

𝝎

72.8

304.2

0.225

44.01

0.069

N2

33.5

126.2

0.04

28.013

0.046

C1 —C7

46.36

214.4

0.0195

19.489

1.325

Saturates (C8 —C34 )

Components

P c (atm)

CO2

MW (g/mol)

Mole (%)

23.49

614.7

0.8757

294.976

62.396

Aromatics (C35 —C44 )

9.42

959.3

1.0051

531.786

16.11

Resins (C45 —C70 )

8.6

1051.9

1.0731

742.844

16.136

Asphaltenes (C71+ )

6.96

1180.2

1.2480

1164.782

3.379

Source: Data taken from Gutiérrez et al. (2018).

465

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

120 1.9 80

1.5

ROV - Sim.

60

Liq. Vol. - Exp.

1.3

40

Liq. Vol. - Sim.

1.1

20 0 20 000

0.9 5000

0

10 000 Pressure (KPa)

1000 900 800 700 600 500 400 300 200 100 0

15 000

990 985 980 Oil visc. - Exp.

975

Oil visc. - Sim. Oil dens. - Exp.

970

Oil density (Kg*m–3)

ROV

ROV - Exp.

Liquid vol. (%)

100 1.7

Oil viscosity (cP)

466

Oil dens. - Sim.

0

5000

10 000 Pressure (KPa)

15 000

965 20 000

Figure 10.6 Comparison of experimental versus simulation results of PVT for oil B, by using SARA components-based fluid model. Source: Adapted from Gutiérrez et al. (2018)/Society of Petroleum Engineers.

Finally, the phase behavior of oil B can be predicted by using the defined EoS parameters. The results of phase oil behavior are presented in Figure 10.7. 10.2.4.5 Lumping Sensitivity

The lumping has two sensitive aspects: (i) the computational cost. The higher the number of pseudocomponents, the higher the computational cost of calculations. For example in a fluid model the computational time can be incresed by two times when increasing the number of components from 10 to 20 (Michelsen et al. 2013). This is not desirable, especially, when dealing with field-scale reservoir numerical models that include high-grid block resolution and high number of operating wells; (ii) the influence of the main recovery mechanisms. In order to generate a consistent PVT model, a systematic lumping must be performed with the least number of pseudocomponents that properly capture the essential thermal physics. During thermal-enhanced oil recovery (EOR), the thermal physics of the main recovery mechanisms such as evaporation, thermal expansion, and viscosity reduction depend on the oil type. For example recovery of a light oil with 32.1∘ API gravity is highly sensitive to evaporation effects when using five components compared to that when using three components (Figure 10.8). On the other hand, recovery of an intermediate oil with 22.0∘ API gravity is roughly insensitive to evaporation effects when using five components compared to that when using three components. Whether in light oil or in intermediate oil, the governing thermal

10.2 Modeling the Phase Behavior of Oil

Table 10.12 model.

EoS matching binary interaction coefficients for oil B with SARA components-based fluid

CO2

CO2 N2

N2

C1 —C7

C8 —C33

C35 —C43

C44 —C68

C71+

zero

0

0.1056

0.12032575

0.15

0.15

0.15

0

zero

0.0298

0.11152164

0.12

0.12

0.12

C1 —C7

0.1056

0.0298

zero

0.08443653

0.1165954

0.134558

0.1426546

C8 —C33

0.120325

0.111521

0.0844365

zero

0.0

0.0

0.0 0.0

C35 —C43

0.15

0.12

0.1165954

0.0

zero

0.0

C44 —C68

0.15

0.12

0.134558

0.0

0.0

zero

0.0

C71+

0.15

0.12

0.1426546

0.0

0.0

0.0

zero

Table 10.13 EoS matching volume shift coefficients for oil B basing on SARA components-based fluid model. Components

S

S(t)

CO2

−0.0817

0

N2

−0.01927

0

C1 —C7

−0.1465

0

Saturates (C8 —C34 )

0.055

−0.0056

Aromatics (C35 —C44 )

0.145

−0.0056

Resins (C45 —C70 )

0.185

−0.0056

Asphaltenes (C71+ )

0.250

−0.0056

6000 Phase behavior

Critical point

Pressure (KPa)

5000 4000 3000 2000 1000 0

0

200

400 600 Temperature (˚C)

800

Figure 10.7 Predicted phase behavior of oil B by using SARA components-based fluid model. Source: Adapted from Gutiérrez et al. (2018)/Society of Petroleum Engineers.

recovery mechanism is oil viscosity reduction due to temperature dependence (Figure 10.2) regardless of the number of components in the model (Zhang et al. 2014). In fact, the petroleum fluid while modeling steam injection under an aquathermolysis reaction system has been reduced to a single pseudocomponent (Hyne 1986; Ibatullin et al. 2011) or multi pseudocomponents (Kapadia et al. 2012, 2013; Ayache et al. 2015; Huang et al. 2017).

467

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process MCN #3

Tb (˚C)

MCN #2

7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39

MCN #1

Figure 10.8 Average boiling point dependence on the lumping scheme and number of MCN (a) three pseudocomponents and (b) five pseudocomponents. Source: Modified from Zhang et al. (2014).

SCN (a) MCN #5 MCN #4

Tb (˚C)

MCN #3 MCN #2 MCN #1

7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39

468

SCN (b)

For instances it is recommended to merge well-defined components of NHC with HC (N2 + C1 and CO2 + C2 ). Also, iso and normal butane and pentane can be merged into one pseudocomponent (iC4 + nC4 , iC5 + nC5 ). In addition, the C1 —C6 SCNs can be represented with two pseudocomponents (i.e. N2 + C1 + CO2 + C2 and C3 + iC4 + nC4 + iC5 + nC5 + C6 ). It is also true that the lumping schemes may depend on the existing chemical reactions to be reproduced by the established reaction model.

10.3 Numerical Validation of Experimental Data The aim of lab-scale numerical validation is to prove the accuracy of a set of experimentally gathered property values, which ultimately enables matching experimental results of oil upgrading by using an EoS. This section discusses the validation of lab-scale kinetics in both static (batch reactor) and dynamic conditions (core displacement) with numerical simulation. It introduces the numerical validation of a reaction system of both noncatalytic and catalytic aquathermolysis. The last refers to the numerical simulation of oil upgrading by catalyst injection into the porous media. The validation of kinetic models of lab-scale investigations through numerical simulation is a prerequisite before simulation of oil upgrading in upscaled models (see Section 10.4).

10.3.1

Numerical Validation in Static Conditions

In a static condition, in both noncatalytic and catalytic aquathermolysis mainly the phenomena of oil mass conversion are investigated in lab-scale facilities. This mass conversion is run at constant

10.3 Numerical Validation of Experimental Data

temperature. When crude oil is exposed to aquathermolysis, the oil composition is modified by reducing the heavy oil fractions and increasing the light ones which include the formation of light hydrocarbon gases and nonhydrocarbon gases as well. In general, in order to carry on a successful numerical validation of oil upgrading when no flow exists, we must obey the following three general conditions: ● ● ●

Reproduce the experimental thermodynamic conditions of P and T Reach mass balance between initial and final mass that includes the sum of the reaction products The final mass distribution of the simulated components must equal those observed experimentally at a certain temperature and time.

Both temperature and pressure are primary variables of any thermal process and key matching parameters during numerical simulation. Since in practice, experimental error is likely, the Antoine correlation Eq. (10.41) predicts the vapor pressure of a well-defined component at any given subcritical temperature. The coefficients of some well-defined components are provided in Table 10.14. Furthermore, during the aquathermolysis gasification of oil occurs. Despite the fact that the overall oil gasification percentage is low in magnitude (Hyne 1986), it adds an excess pressure or gas partial pressure to the system. The Dalton’s law (Eq. (10.42)) states that the total pressure of nonreacting gases in a system is the sum of the partial pressure of each gas: ( ) B log10 (Psat ) = A − (10.41) T+C (10.42)

PT = P1 x1 + P2 x2 + P3 x3 … … + Pn xn

The general form of the mass balance equation can be represented as in Eq. (10.43). Most of the aquathermolysis kinetic models for numerical simulation assume that the water molecule is not dissociated and that water is completely insoluble in the oil phase. Thus, the aquathermolysis reaction products result from the oil phase: ∑ ∑ Mf (10.43) Mi = The accuracy of the prediction of changes in oil composition in numerical simulation is governed by the reaction rates when oil is thermally induced (Figure 10.9). The Arrhenius equation (Eqs. (10.44) and (10.45)) is a temperature-dependent relation that calculates the kinetics of the

Table 10.14 Correlation coefficients of Antoine correlation for prediction of vapor pressure of some well-defined compounds commonly found in Aquathermolysis process. Component

Temperature (K)

A

B (K)

C(K)

P

References

H2 O

379–573

3.55959

643.748

−198.043

Bar

Liu and Lindsay (1970)

CH4

118–190

5.7687

−6.469

Kpa

C2 H6

198–305

5.95405

663.72

−16.469

Kpa

Martienssen et al. (1999)

C3 H8

248–369

5.92828

803.997

−26.108

Kpa

C4 H10

298–425

5.93266

935.773

−34.361

KPa

C5 H12

328–470

5.97786

1064.84

−41.138

Kpa

C6 H14

358–508

6.00139

1170.88

−48.833

KPa

H2 S

228–363

6.22882

−21.761

KPa

3.95744

806.933

469

1.3

0.009

1.2

0.008 0.007

1.1

0.006

1

0.005 0.9 0.004 0.8

0.003

0.7

0.002

0.6

Mf, Hyd. Sulf./Mi, oil (Dimensionless)

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Mf, oil/Mi, oil (Dimensionless)

470

0.001

0.5

0

50 Sat. Exp. Sat. Sim. Hyd. Sulf. Exp.

100 Time (hr) Arom. Exp. Arom. Sim.

150 Res. Exp. Res. Sim.

200

0

Asp. Exp. Asp. Sim.

Hyd. Sulf. Sim.

Figure 10.9 Comparison experimental versus simulation during kinetics validation under static conditions. Source: Modified from Ayache et al. (2015).

reaction rates in each grid block of the numerical model in most of the commercial simulators that incorporate an EoS. The kinetic constants A and Ea of the Arrhenius-type equation are obtained using analytical methods of experimental results (see Chapter 7) (Ancheyta 2017). Ea is the activation energy for a reaction to take place, and A is the pre-exponential factor or sometimes also called frequency factor, that describes how often and effectively molecules collide in dependence on their orientation. Usually, A becomes an adjustable parameter that fits the experimental changes in oil composition: ) ( −Eai Ki = Ai x exp (10.44) RT Or Eai (10.45) RT Usually, a 0D grid-type is used for numerical simulation of reactors. However, it is not a rule. The reason for choosing this grid type is because it best represents the absence of flow in the x, y, or z vectors as it occurs in a batch reactor where this commonly consists of a few hundred cm3 of volume. The 0D grid-type model is illustrated in Figure 10.10. ln ki = ln Ai −

10.3.2

Validation of Lab-Scale Kinetic Models in Dynamic Conditions

Dynamic lab-scale tests are meant to estimate the oil recovery factor at reservoir conditions when EOR methods are implemented. In the dynamic conditions unlike the static conditions, simultaneous processes take place influenced by the flow of suspended catalyst particles in the porous media during high temperature in in situ oil upgrading (e.g. mass conversion, mass transport, retention). Each of these processes contribute in a certain degree to the overall EOR. To capture the kinetics of

10.3 Numerical Validation of Experimental Data

Figure 10.10

3D view of a 0D numerical model that simulates a reactor experiment. Producer

Injector

Figure 10.11

3D view of a 1D numerical model to simulate a displacement experiment.

these processes, individual tests are required. This avoids misinterpretations due to process overlapping and ultimately assures data accuracy. Last but not least, lab-scale tests are used to calibrate numerical simulation models that could serve for further upscaling tasks (see Section 10.4). The numerical simulation model reproducing the dynamic condition of oil upgrading is accomplished with a 1D grid-type model (Figure 10.11). This is because the catalytic aquathermolysis oil upgrading process is usually simulated as a flooding experiment in a displacement tube with an oil saturated core sample (Zamani et al. 2010; Hashemi and Pereira 2011; El-Amin et al. 2012; Shokrlu and Babadagli 2013; Hosseini and Javadpour 2018), which implies that the fluid flow is limited to one vector, whether the x or y direction. The accurate numerical validation of lab-scale dynamic conditions is achieved by matching the kinetics of the processes associated to the main recovery mechanisms participating in the catalytic oil upgrading in in situ conditions (e.g. oil viscosity reduction due to thermal induction and oil upgrading, wettability alteration due to catalyst particle retention). The general workflow to calibrate a model for numerical validation of an oil upgrading catalytic aquathermolysis dynamic test follows the next steps: Step 1 Validation of catalytic oil upgrading kinetics in static conditions (see Section 10.3.1). Step 2 Determination of catalyst transport and adsorption kinetics in porous media (see Chapter 9). Step 3 Validation of the overall oil recovery and oil upgrading in dynamic conditions. Other steps may integrate into the numerical model and the solubility equilibrium constants of aquathermolysis gas products (Kapadia et al. 2012, 2013; Ayache et al. 2015). The numerical validation of oil upgrading in dynamic conditions can be illustrated by the above general workflow. The validation of step 1 uses a calibrated model by Zapata et al. (2019). The input data regarding fluid and kinetic models are presented in Table 10.15. The dynamic experimental conditions here refer to a core flooding experiment. The parameters of the core assemblage are provided in Table 10.16.

471

472

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Table 10.15

Fluid and kinetic calibrated models corresponding to a heavy oil field.

Comp.

𝝎

Gases

0.0913

Lights Sat. Arom. Res. Asphalt.

P c (atm)

E a (J/gmol)

A (min−1 )

—

—

—

−0.0478

—

—

—

0.0021

—

—

—

204.49

0.0893

Sat.

1.45E05

9.4E06

366.24

0.1404

Arom.

1.90E05

3.97E07

Sat.

1.46E05

1.19E11

Res.

1.36E05

3.52E07

Arom.

1.67E05

1.13E10

Sat.

1.92E05

3.34E11

Gas

2.61E05

7.96E16

T c (K)

MW

Vol. shift

53.63

222.5

27.48

−0.0792

0.16109

40.27

377.9

49.08

0.29488

31.35

525.6

90.42

0.67658

17.04

723.2

1.06821

10.06

859.4

1.38001

7.49

1160.4

1087

0.2212

Prod.

Sat. = Saturates, Arom. = Aromatics, Res. = Resins, Asphalt. = Asphaltenes. Source: Data from Bueno and Mejía (2021); Zapata et al. (2019).

Table 10.16

Core model features.

Parameter

Fluid and reaction models

Value

(Zapata et al., 2019; Bueno and Mejía 2021)

Core length, cm

60.96

Distance y and z, cm

3.3312

Porosity, %

26.2

Permeability (i, j, k), D

1

Initial oil saturation, %

78.4

Initial core temperature, ∘ C

80

Initial core pressure, KPa

3447

Injection rate (water equivalent), (cm3 /min) Injection temperature, ∘ C

210

0.3

While catalyst particles flow within the porous media, they undergo adsorption or sometimes also called retention. The adsorption kinetic rates of catalyst particles increase at low-reservoir quality characteristics and decrease at high-reservoir quality characteristics (i.e. low porosity and permeability) (Shokrlu and Babadagli 2013). The increase in the adsorption kinetics leads to an increase in the concentration of catalyst particles inside the porous media. The adsorption phenomenon is simulated by the Langmuir isotherm, and its general form is presented in Eqs. (10.46) and (10.47). In Table 10.17 some parameter values are shown for the adsorption of catalyst particles in porous media: bP (1 + bP) k b= a kd 𝜃=

(10.46) (10.47)

Experimental results with catalytic upgrading generally range between 2 and 15% of additional recovered oil (Hamedi-Shokrlu and Babadagli 2014; Li et al. 2019; Vakhin et al. 2022). The recovery

10.3 Numerical Validation of Experimental Data

Table 10.17

Catalyst particle adsorption parameters.

Parameter

Value

Catalyst injection (wt%)

0.1a)

Molar fraction

0.00012

Adsorbed moles (gmole/cm3 )

2.6E–6

Max adsorption (gmole/cm3 )

2.6E–6

Accessible Resistance factor @ Q = 2 cm3 /min, K = 6D, poro = 0.3

3.88a)

Accessible pore volume

1

a) Source: Adapted from Shokrlu and Babadagli (2013).

Table 10.18 Relative permeability sets for calibrated models with and without catalyst retention during oil displacement experiments. Corey parameters

Without adsorption

With adsorption

Irreducible oil mobility

0.17

0.13

Irreducible water mobility

0.22

0.16

Maximum oil permeability

1

0.98

Maximum water permeability

0.0026

0.0041

Oil exponent

1.28

1.00

Water exponent

1.11

1.62

Recovery factor (%)

Source: Data from Bueno and Mejía (2021).

90 85 80 75 70 65 60 55 50 45 40

Steam Steam + cat

0

3 1 2 Pore volume (Dimensionless)

4

Figure 10.12 Numerical simulation showing the effect of steam + catalyst injection during displacement and oil upgrading in lab-scale. Source: Adapted from Bueno and Mejia (2021)/Elsevier.

mechanism responsible for the additional recovery other than viscosity improvement due to oil upgrading is attributed to a change in the rock wettability as a result of particle retention (Ju and Fan 2009; Sun et al. 2017). Wetting of a solid surface is one of the most influencing and effective mechanisms among the enhanced oil recovery mechanisms. Hydrophilic catalyst particles modify rock wettability to a more water-wet system (i.e. reducing the residual oil saturation) (Li et al. 2019). Since at a certain catalyst concentration onto the reservoir rock the relative permeabilities are altered, then, they are called suitable matching parameters. This approach leads to the matching of oil production and recovery factor of experimental tests (Figure 10.12). In Table 10.18 is presented

473

474

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

an example of a set of relative permeability under steam injection and after steam + catalyst injection.

10.4 Upscaling Laboratory-Scale to Field-Scale This section discusses the upscaling procedure from lab-scale results to field-scale of in situ catalytic oil upgrading and EOR process. Also, it summarizes some approaches and the relevant parameters to upscale laboratory results of steam processes. Additionally, it is provided as an example of an upscaling case with the injection of a hydrophilic catalyst to show the EOR matching of lab-scale results. Furthermore, key details of the grid selection criteria during numerical simulation of steam processes are provided.

10.4.1

Some Approaches for Upscaling Steam Processes

There are two critical points within the theory of dimensionally scaled models: (1) The possibility of transferring experimental results from lab-scale to the real processes in big oil fields. (2) Validation of numerical simulation models according to experimental results. For the direct transfer of results to a real reservoir, it is of high priority to observe equality of all dimensionless parameters for both laboratory and field models that significantly affect the solution of the results which represent a practical interest. As a rule, all these parameters are in the order of unity (small values can usually be neglected). As will be shown below, it is unrealistic to fully satisfy the equality of all similarity criteria for a catalyst propagation model during upscaling processes. Thus, it is impossible to draw conclusions about the parameters of the catalyst influence on the development of the entire field directly from the experimental results. In this case, the results of the experimental investigations on EOR are used to match numerical models using a set of scaled parameters. After successful numerical validation of experimental results, these parameters are changed to real ones that correspond to a real oil reservoir to carry on numerical simulation. For the process of catalyst propagation in a porous medium, heat and mass transfer equations are used, taking into account the following factors: (1) (2) (3) (4) (5) (6) (7)

Transfer by the general filtration flow Gravitation Diffusion Dispersion Adsorption Mass conversion of oil factions Thermal activation of the catalyst

The approaches for upscaling laboratory data of steam processes are classified in relation to the pressure of the system: low-pressure (Stegemeier et al. 1980) and high-pressure (Pujol and Boberg 1972; Kimber and Farouq Ali 1989; Huygen 2007). The low-pressure models may require different fluids at both scales. On the other hand, high-pressure models commonly use the same fluids at both scales. The upscaling approaches for high-pressure models are presented in Figure 10.13 (Kimber and Farouq Ali 1989).

10.4 Upscaling Laboratory-Scale to Field-Scale

Scaling of high-pressure reservoir fluids Preserved conditions: – ΔTmax – Porosity, saturation

Preserved conditions: – ΔTmax, ΔPmax – Porosity, permeability, saturation

Modified conditions: – Reservoir height and width – Injection rate – Time – Absolute permeability – Pressure drop, ΔPmax

Modified conditions: – Reservoir height and width – Injection rate – Time

Only steam injection processes

Steam or steam-additive injection processes

Hot liquid injection-type processes

Approach #1 Aim: processes with important gravity effects

Approach #2 Aim: processes with poor gravity effects as in thin formations

Approach #3 Aim: processes with gravity effects in thick formations

Approach #4 Aim: processes with transverse dispersion in very thin formations

Scales: – Diffusion – Gravity forces – Viscosity forces

Scales: – Diffusion – PVT steam & additive – Viscosity forces – Heat transfer – Irreducible saturation and relative permeability

Scales: – Gravity forces – Viscosity forces

Scales: – Transverse dispersion – Viscosity forces

Relaxes: – Steam saturation P/T – Velocity distribution within steam zone – Dispersion – Capillary pressure – Irreducible saturation and relative permeability

Relaxes: – Dispersion – Vertical heat transfer – Capillary pressure

Relaxes: – Capillary pressure – Gravity forces

Relaxes: – Dispersion – Gravity forces

Figure 10.13 Scaling approaches for high-pressure reservoir systems. Source: Modified from Kimber and Farouq Ali (1989).

Let us generalize the approach proposed by Kimber et al. (1988) and Kimber and Farouq Ali (1989) for a system of catalyst heat and mass transfer equations with the equations of state and process kinetics, as well as with the corresponding boundary and initial conditions. For the scaling option with the preservation of internal processes (PVT properties, the form of the equations of phase permeabilities), one can come to the following dimensionless parameters that determine the solution: Ratios of characteristic geometric dimensions HR ∕LR ; WR ∕LR

(10.48)

Porosity (10.49)

mR Dimensionless pressures Pwell i PR Pcapij PR

(10.50) (∗)

(10.51)

Here and below in Eq. (10.57) the symbol (*) denotes dimensionless parameters, which are most often small and their influence can be neglected.

475

476

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Dimensionless well injection rates wwell i wwR

(10.52)

Dimensionless densities 𝜌o R 𝜌vR , 𝜌wR 𝜌wR

(10.53)

Dimensionless mass transfer parameters kwR 𝜇oR kvR 𝜇oR , koR 𝜇wR koR 𝜇vR DLiR DTiR , DLcR DLcR

ki = kkri

(10.54) (10.55)

Dimensionless heat transfer parameters ho R hgR 𝜌rR urR 𝜌sR usR 𝜆i , , , , hwR hwR 𝜌wR hwR 𝜌wR hwR 𝜆rR

(10.56)

Ratio of catalyst transport by diffusion to filtration 𝜇oR DLcR (∗) koR PR D D= M F

(10.57) (10.58)

Ratio of catalyst transport by dispersion to filtration 𝜇oR 𝛼LcR |uLoR | (∗∗) koR PR

(10.59)

This parameter is small for the case of constant flows but can be significant for cyclic processes. Ratio of heat transfer by conduction to convection. 𝜆rR TR 𝜇oR (∗∗∗) koR PR 𝜌oR hoR

(10.60)

This relationship is small for light oil but may be significant in the propagation of heat in heavy oil fields. Ratio of transport by transverse gravity to transport by longitudinal filtration Δ𝜌woR gLR PR

(10.61)

Ratio of the characteristic injection rate to the filtration transport wwR 𝜇oR 𝜌wR koR PR HR

(10.62)

Ratio of oil fraction conversion by catalyst to oil transfer by filtration KijR L2R 𝜇oR koR PR

(10.63)

Ratio of catalyst adsorption to catalyst transport by filtration 𝛽R L2R 𝜇oR koR PR

(10.64)

10.4 Upscaling Laboratory-Scale to Field-Scale

Table 10.19 Some scaling factors for steam or steam + additive process using approach #3. Parameter

Scaling factor

PVT

1

Porosity

1

Permeability

1

Oil saturation

1

Pressure-drop

1

Steam quality

1

Thermal properties

1

Distance x

au

Distance y Heigh z

au a)a u

Injection rate

au

Reaction kinetics

au −2

Adsorption

au −2

a) It provides better fitting when using (1E–02 au 2 ).

With this definition of the dimensionless parameters, the characteristic time of the process is defined as follows: tR =

L2R (koR ∕𝜇oR )PR

Characteristic mass flow of water in the well 𝜌 k P H wwR = wR oR R R 𝜇oR

(10.65)

(10.66)

Thus, to scale the results of an experiment of catalytic aquathermolysis where additives are injected into the porous media, it is necessary to use the following upscaling factors in Table 10.19.

10.4.2

Upscaling Laboratory Data

The approach #3 scales PVT properties, porosity, permeability, pressure drop, thermal properties, and steam quality by simulating the same conditions as in lab-scale tests (Table 10.20). In this approach, vertical and horizontal heat transfer are properly upscaled in both magnitude and time. On the other hand, the geometrical properties must be scaled (Figure 10.14). The geometry of the model is defined by two parameters, the length (x, y) and the thickness (z). The length is scaled by a factor “au ” (Eq. (10.67)), which is the relation of the distance between the wells in the lab-scale and the field-scale model. The definition of the scaling factor “au ” allows to scale other parameters, according to the influence of the desired scaling group, as provided in Table 10.19: au =

Lf Lm

(10.67)

The validated kinetic and oil recovery models under static and dynamic conditions, respectively (see Section 10.3), are requirements before the implementation of upscaling techniques. Moreover,

477

478

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Table 10.20

Some scaling factors for steam or steam + additive process using approach #3.

Parameter

PVT Porosity, % Permeability, D Oil saturation, %

Lab-scale

Upscaled 1

Upscaled 2

—

—

—

26.2

26.2

26.2

1

1

1

78.4

78.4

78.4

Distance x, m

0.6096

100

170

Distance y, m

0.033

100

170

Heigh z, m

0.033

8.96

25.91

Injection rate, (cm3 /min; m3 /d) −1

−1

Reaction kinetics, min ; day

1/a2

11.62

33.60

3.52E + 07

1.88E + 06

6.51E + 05

1.13E + 10

6.03E + 08

2.09E + 08

3.34E + 11

1.79E + 10

6.19E + 09

7.96E + 16

4.26E + 15

1.47E + 15

3.97E + 07

2.12E + 06

7.34E + 05

1.19E + 11

6.39E + 09

2.21E + 09

9.40E + 06

5.03E + 05

1.74E + 05

Injector Producer

Figure 10.14

3D view of a 3D numerical model to simulate upscaling of laboratory data.

calibrated lab-scale models of oil upgrading have shown to foster additional oil recovery when forecasting with upscaled numerical models (Nguyen et al. 2017; Bueno and Mejía 2021; Vakhin et al. 2022). Additionally, field EOR implementations of oil upgrading technologies with hybrid steam processes (i.e. steam + Hydrogen donor + catalyst) demonstrated that field-scale numerical models can fairly predict oil recovery rates after stimulation (Vakhin et al. 2022). The upscaling techniques of numerical models are valuable tools for estimation of the economics of EOR pilot projects. Here is considered an upscaling case with approach #3, where porosity and permeability are as in the lab-scale model (Table 10.20). The lab-scale input parameters are converted into field-scale by using the scaling factors discussed in Section 10.4.1 as provided in Table 10.20.

Recovery factor (%)

10.4 Upscaling Laboratory-Scale to Field-Scale

100 90 80 70 60 50 40 30 20 10 0

Lab-scale Upscaled 1 Upscaled 2

0

0.4 0.8 Pore volume (Dimensionless)

1.2

Figure 10.15 Comparison of lab-scale and upscaled models recovery factor versus injected water equivalent pore volume with approach #3.

Figure 10.15 illustrates convergence on the oil recovery factor observed in lab-scale results and upscaled models discussed in this section. Also, the two upscaled cases overlap each other, meaning good agreement with the approach. One important fact to keep in mind during upscaling process is that among the constraints of field-scale models are the real geological parameters of the reservoir. The pore network of oil reservoirs has the tendency to behave not homogeneously, and instead, heterogeneously. The heterogeneity of a reservoir is always under certain range of uncertainty, specially, for parameters such as porosity and permeability. The porosity and permeability have a linear relationship in sandstone reservoirs (Yousef et al. 2021). However, no relationship is found in carbonate reservoirs. The permeability in carbonate rocks is generally low and is greatly dependent on the interconnectivity of the pore network, which is often low due to secondary processes of pore space replacement (Ibrahem et al. 2022). Usually, these two parameters are updated in geological models while performing tasks of oil production history matching. The uncertainty produced by these parameters must be carefully taken into account during upscaling and prediction of oil recovery with the application of any EOR technology.

10.4.3

Criteria for the Selection of the Optimal Grid Type and Size

In the simulation of the catalyst co-injection with steam, it is of high priority to take into account the following two facts: ●

●

The accuracy of determining volumetric flows rates at the given pressure or vice versa, the bottomhole pressure and pressure in the vicinity of the well-known flows; The accuracy of determining the sweep efficiency of the reservoir by the catalyst co-injection and heating.

The first is important due to the strong relationship between the properties of steam and its vapor pressure. The second is important due to the numerical dispersion that arises when calculating the equation for the transport of catalyst particles. During the catalyst co-injection with steam displacement, it is mandatory to take into account the following key factors that may affect the choice of the optimal grid for numerical simulation: ● ●

Radial flow for vertical wells and elliptical flow for horizontal wells arise in the well vicinity; A common filtration field in the reservoir is superimposed on the local flow.

479

480

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

The solution of the tasks set can be achieved with the use of local grid refinement. Several approaches are available for the solution of such tasks where the grid size and type may play a meaningful role, specially, when dealing with the mass transport phenomena. Among the available approaches we should notice the dynamic gridding (Biterge and Ertekin 1992; Ding and Lemonnier 1993; Liu et al. 2003) in which the grid is compacted in the vicinity of the main flow of the process or the main field gradients (Figure 10.16). The advantage of this method is the best match of the grid area to the flow pattern. Such grids would best suit the catalyst distribution area. However, in this case, there are significant requirements for re-interpolation of values between grids, selection of optimal calculation parameters for each area, taking into account their relationship. After the operations of steam injection within the numerical model are completed, the flow regime changes significantly. This requires re-gridding for further analysis. This raises the need to re-interpolate the distribution of the catalyst, leading to a significant loss of accuracy. Also, the loss of accuracy occurs during the initial transfer of the calculated water saturation and oil saturation from the available history-matched field models to dynamic grids. Despite the large number of articles that solve problems on this topic, this approach is practically not implemented in commercial simulators, with the exception of certain CMG tools. A similar problem arises for all variants of grids that require rebuilding during calculations. Finite element grids (Figures 10.17 and 10.18) are a variant of a stationary computational grids that accurately take into account the flows in the given areas (Chavent and Jaffre 1986). On such grids, it is possible to accurately describe the bottom hole itself without using the source model. In this case, only the problem of transferring the simulation model previously built in CornerPoint to a new mesh remains. For small finite element grids, unfortunately, high computation time is required. This factor is often critical in reservoir simulation. Also, they are not implemented in

Figure 10.16

Some types of dynamic grids.

Figure 10.17

Example of finite element meshes in the interwell area for a pair of SAGD wells.

10.4 Upscaling Laboratory-Scale to Field-Scale

Figure 10.18 Examples of finite element meshes in the vicinity of the well.

Figure 10.19

Types of hybrid grids (cornerPoint with radial grid-types).

commercial simulators. In this regard, at the moment this approach can be well used for scientific study of the effect of a catalyst on a single well or a small group of wells but is practically unrealistic on the scale of fields. To analyze catalyst injection, it is possible to use a combined CornerPoint grid with radial or radial-elliptical gridding in the vicinity of the well – the so-called “hybrid grid refinement” (Morita et al. 1990; Norris and Piper 1990; Goktas and Ertekin 1999). Two possible options for vertical wells are shown in the Figure 10.19. At this moment, a number of modern commercial simulators allow the use of local radial gridding. This is certainly an important advantage of the practical application of this approach. These kind of grids are of meaningful utility for numerical simulation with vertical wells. However, for horizontal sections in a reservoir with small thicknesses of layers, radial refinement of the existing grid is practically impossible, or a preliminary combination of layers is required. Also, the radial refinement of the current grid is difficult for inclined wells and wells of complex shape. A fairly effective and simple method for constructing grid refinement is the method of stationary local multiple refinement of CornerPoint cells (Quandalle and Besset 1985; Wasserman 1987), most often along the wellbore. Such grids are implemented in most modern reservoir simulators.

481

482

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

The use of this approach allows not only to solve the problems of better localization of the catalyst distribution by reducing the numerical dispersion and, in general, increasing the accuracy of numerical schemes, but also to make it possible to better describe the complex geometry of well trajectory, which also increases the accuracy of calculations. However, there is a lower limit on the size of computational cells. The size of the effective radius of the block (Eq. (10.68)), corresponding to the Peaceman formula (Peaceman 1993) should remain noticeably larger than the well radius, i.e.: rb = 0,2•Δx > rw

⇒

Δx > 5rw

(10.68)

From this ratio for the characteristic radii of wells, it can be determined that the minimum longitudinal cell size should not be less than 1 m. To account for the features of the bottomhole zone, it is desirable that this size be even slightly larger. Local mesh refinement should be used only in the area where it is needed. As a result, the efficiency of computational procedures will be optimal. There are two ways to determine the required area: ●

●

Semianalytical method of integral estimates. In this case, the field of filtration flow is calculated on the initial grids when the catalyst and steam are injected. After that, according to the given volume of the dissolved catalyst and the volume of steam, it is possible to estimate the region of the catalyst distribution and the region of the thermal activation front. The maximum total area is assumed to be necessary for mesh refinement. Fully computational estimate. In this case, a wide mesh refinement is carried out around the well with sufficiently large grid spacings, and for this option, the area of distribution is calculated. The resulting area is covered with a more detailed mesh. If necessary, the procedure can be used iteratively.

Nomenclature T cps T cnA Tb 𝛾 nA 𝛾 ps 𝛾 Pcps PcnA V cps V cnA 𝛽 𝜌0 𝜌w 𝜇 𝜇 210 𝜇 100 Tp R v

the critical temperature of the pseudocomponent (R) the critical temperature of the normal alkane (R) the normal boiling point (R) the specific gravity of the n-alkane (dimensionless) the specific gravity of the pseudocomponent (dimensionless) the specific gravity of oil (dimensionless) the critical pressure of the pseudocomponent (psia) the critical pressure of the normal alkane (psia) the critical volume of the pseudocomponent (ft3 /lb/mol) the critical volume of the normal alkane (ft3 /lb/mol) coefficients of the critical parameters (dimensionless) oil density at standard conditions (kg/m3 ) water density at standard conditions (kg/m3 ) dynamic viscosity (cP) the measured viscosity at 210 ∘ F (cP) the viscosity measured at 100 ∘ F (cP) temperature of the predicted oil viscosity (∘ F) ideal gas constant 8.3145 (J/mol/K) molar volume (l/mol)

Nomenclature

a b 𝛿 ij vci n V V EoS c ci VL Vv VLEoS VvEoS M d e s si s1i T ref T res wi xi, calc xi,exp Psat T A B C PT Pi xi Mi Mf Ai Eai Ki 𝜃 ka kd P L W H m 𝜌 Δ𝜌

first constant of the P&R EoS second constant of the P&R EoS interaction coefficient between component i and j (dimensionless) critical volume of component i and j (l/mol) regression coefficient of the binary interaction correlation (dimensionless). Corrected molar volume (l/mol) EoS calculated molar volume (l/mol) third parameter of the P&R EoS or volume shift (dimensionless) third parameter of the P&R EoS or volume shift of component “i” (dimensionless) corrected liquid molar volume (l/mol) corrected vapor molar volume (l/mol) EoS calculated liquid molar volume (l/mol) EoS calculated vapor molar volume (l/mol) molecular weight (g/mol) coefficient of the volume shift correlation 10.38 (dimensionless) coefficient of the volume shift correlation 10.38 (dimensionless) volume shift coefficient (dimensionless) volume shift that takes into account temperature dependence (dimensionless) temperature-dependent volume shift coefficient (1/K) reference temperature (K) the reservoir temperature (K) weight factor (fraction) calculation value experimental value vapor pressure of gas (bar or KPa) temperature (K) coefficient of the Antoine correlation (dimensionless) coefficient of the Antoine correlation (K) coefficient of the Antoine correlation (K) total pressure of the system (bar) partial pressure of component i (bar) molar composition of component i (fraction) input mass (g) final mass (g) frequency factor of reaction i (sec−1 ) activation energy of reaction i (J/mol) reaction rate of reaction i (sec−1 ) fractional surface covered by the adsorbed molecule (dimensionless) respective rate constant for adsorption (sec−1 ) Respective rate constant for desorption (sec−1 ) pressure (bar) length (m) width (m) thickness (m) porosity (fraction) density (kg/m3 ) differential density of i and j phase (kg/m3 )

483

484

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

k kri K ij Pcap ij Pwell i w D DM F 𝛼 𝜆 h u |u| 𝛽 g au Lf Lm rb rw Δx

permeability (Darcy) relative permeability of i phase (Dimensionless) kinetic coefficient of i fraction conversion into j (sec−1 ) capillary pressure between phases i and j (bar) pressure in well i (bar) mass flow in the well (Kg) diffusion coefficient of fraction i (m2 /day) molecular diffusion coefficient (m2 /day) tortuosity (dimensionless) dispersivity (m) heat conductivity coefficient (J/m/day/C) enthalpy (J/kg) internal energy (J/kg) magnitude of interstitial velocity (m/day) kinetic coefficient (constant of adsorption rate)/(sec) acceleration of gravity (m/sec2 ) upscaling factor (dimensionless) distance between wells in the field-scale model (m) distance between wells in the lab-scale model (m) block radius (m) well radius (m) cell size (m)

Indexes R o w g c L T r s i

reference value oil water gas catalyst longitudinal transverse reservoir rock surround formations some of phase or well number or oil fraction.

References Al-Khafaji, A.H., Abdul-Majeed, G.H., and Hassoon, S.F. (1987). Viscosity correlation for dead, live and undersaturated crude oils. Journal of Petroleum Research 6 (2): 1–16. Ancheyta, J. (2017). Chemical Reaction Kinetics. Hoboken, NJ: Wiley. Ayache, S.V., Lamoureux-Var, V., Michel, P., and Preux, C. (2015). Reservoir simulation of H2 S production during a SAGD process using a new sulfur-based-compositional kinetic model. SPE Journal 22 (1): 80–93. Beal, C. (1946). The viscosity of air, water, natural gas, crude oil and its associated gases at oil field temperatures and pressures. Transactions of AIME 165 (01): 94–115.

References

Bergman, D.F. and Sutton, R.P. (2009). A consistent and accurate dead-oil-viscosity method. SPE Reservoir Evaluation and Engineering 12 (6): 815–840. Besserer, G.J. and Robinson, D.B. (1975). Equilibrium-phase properties of isopentane-carbon dioxide system. Journal of Chemical and Engineering Data 20 (1): 93–96. Biterge, M.B. and Ertekin, T. (1992). Development and testing of a static/dynamic local grid-refinement technique. Journal of Petroleum Technology 44 (4): 487–495. Bueno, N. and Mejía, J.M. (2021). Heavy oil in-situ upgrading evaluation by a laboratory-calibrated EoS-based reservoir simulator. Journal of Petroleum Science and Engineering 196: 107455. Chavent, G. and Jaffre, J. (1986). Mathematical Models and Finite Elements for Reservoir Simulation-Single Phase, Multiphase and Multicomponent Flows Through Porous Media. Amsterdam, The Netherlands: Elsevier Science Publishers B.V. Coats, K.H. (1985). Simulation of gas condensate reservoir performance. Journal of Petroleum Technology 37 (11): 1870–1886. Ding, Y. and Lemonnier, P.A. (1993). Development of dynamic local grid refinement in reservoir simulation. Paper presented at the Proceedings of the SPE Symposium on Reservoir Simulation, New Orleans, Louisiana (28 February 1993). 501–510. El-Amin, M.F., Salama, A., and Sun, S. (2012). Modeling and simulation of nanoparticles transport in a two-phase flow in porous media. Society of Petroleum Engineers-SPE International Oilfield Nanotechnology Conference 2012, Noordwijk, The Netherlands (12–14 June 2012). 59–67. Goktas, B. and Ertekin, T. (1999). Implementation of a local grid refinement technique in modeling slanted, undulating horizontal and multi-lateral wells. Proceedings-SPE Annual Technical Conference and Exhibition, Houston, Texas (3–6 October 1999), 2 SIGM. SPE-56624-MS. Gutiérrez, D., Moore, R.G., Mehta, S.A. et al. (2018). Phase behavior modelling of oils in terms of SARA fractions. Society of Petroleum Engineers-SPE Canada Heavy Oil Technical Conference, Calgary, Alberta, Canada, (13–14 March 2018). SPE-189727-MS. Hamedi-Shokrlu, Y. and Babadagli, T. (2014). Kinetics of the in-situ upgrading of heavy oil by nickel nanoparticle catalysts and its effect on cyclic-steam-stimulation recovery factor. SPE Reservoir Evaluation and Engineering 17 (3): 355–364. SPE-170250-PA. Hashemi, R. and Pereira, P. (2011). Experimental study of simultaneous athabasca bitumen recovery and upgrading using ultradispersed catalysts injection. Society of Petroleum Engineers-Canadian Unconventional Resources Conference 2011, Calgary, Alberta, Canada (15–17 November 2011), CURC 2011 3: 1727–1739. SPE-149257-MS. Hong, K.C. (2007). Lumped-component characterization of crude oils for compositional simulation. Paper Presented at the SPE Enhanced Oil Recovery Symposium, Tulsa, Oklahoma, (4–7 April 1982). SPE-10691-MS. Hossain, M.S., Sarica, C., and Zhang, H.Q. (2005). Assessment and development of heavy-oil viscosity correlations. SPE/PS-CIM/CHOA International Thermal Operations and Heavy Oil Symposium Proceedings 2005, Calgary, Alberta, Canada (1–3 November 2005). B321–B329. Hosseini, A. and Javadpour, F. (2018). Determination of nanoparticle macrotransport coefficients from pore scale processes. Transport in Porous Media 25 (2): 377–394. Huang, S., Huang, Q., Liu, H. et al. (2017). A modified model for aquathermolysis and its application in numerical simulation. Fuel 207: 568–578. Huygen, H. (2007). Laboratory steamfloods in half of a five-spot. Paper Presented at the SPE Annual Fall Technical Conference and Exhibition, New Orleans, Louisiana (3–6 October 1976). SPE-6171-MS. Hyne, J.B. (1986). Aquathermolysis: a synopsis of work on the chemical reaction between water (steam) and heavy oil sands during simulated steam stimulation. Synopsis report No.: 50, AOSTRA Contracts No.: 11, 103, 103B/C.

485

486

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

Ibatullin, T.R., Yang, T., Petersen, E.B. et al. (2011). Simulation of hydrogen sulfide and carbon dioxide production during thermal recovery of bitumen. Society of Petroleum Engineers-SPE Reservoir Characterisation and Simulation Conference and Exhibition 2011, Abu Dhabi, UAE, (9–11 October 2011). RCSC 2011: 276-286. SPE-147934-MS. Ibrahem, Y., Morozov, V.P., Sudakov, V. et al. (2022). Sedimentary diagenesis and pore characteristics for the reservoir evaluation of Domanik formations (Semiluksk and Mendymsk) in the central part of Volga-Ural petroleum province. Petroleum Research 7 (1): 32–46. Jhaveri, B.S. and Youngren, G.K. (1988). Three-parameter modification of the Peng-Robinson equation of state to improve volumetric predictions. SPE Reservoir Engineering (Society of Petroleum Engineers) 3 (3): 1033–1040. SPE-13118-PA. Ju, B. and Fan, T. (2009). Experimental study and mathematical model of nanoparticle transport in porous media. Powder Technology 192 (2): 195–202. Kapadia, P.R., Wang, J., Kallos, M.S., and Gates, I.D. (2012). New thermal-reactive reservoir engineering model predicts hydrogen sulfide generation in steam assisted gravity drainage. Journal of Petroleum Science and Engineering 94–95: 100–111. Kapadia, P.R., Kallos, M.S., and Gates, I.D. (2013). A new reaction model for aquathermolysis of Athabasca bitumen. Canadian Journal of Chemical Engineering 91 (3): 475–482. Katz, D.L. and Firoozabadi, A. (1978). Predicting phase behavior of condensate/crude-oil systems using methane interaction coefficients. Journal of Petroleum Technology 30 (11): 1649–1655. SPE-6721-PA. Kesler, M.G. and Lee, B.I. (1976). Improve prediction of enthalpy of fractions. Hydrocarbon Processing 55 (3). Kimber, K.D. and Farouq Ali, S.M. (1989). New options for scaling steam injection experiments. Paper Presented at the SPE California Regional Meeting, Bakersfield, California (5–7 April 1989). SPE-18751-MS. Kimber, K.D., Farouq Ali, S.M., and Puttagunta, V.R. (1988). New scaling criteria and their relative merits for steam recovery experiments. Journal of Canadian Petroleum Technology 27 (4 Jul.-Aug. 1988): 86–94. PETSOC-88-04-07. Li, S., Torsæter, O., Lau, H.C. et al. (2019). The impact of nanoparticle adsorption on transport and wettability alteration in water-wet Berea sandstone: an experimental study. Frontiers in Physics 7 (May): 74. Liu, C.T. and Lindsay, W.T. (1970). Vapor pressure of D2O from 106 to 300∘ C. Journal of Chemical and Engineering Data 15 (4): 510–513. Liu, W., Mezzatesta, A.G., and Hildebrand, M.A. (2003). Intergrid operations-dynamic gridding techniques in nonlinear domain decomposition methods. Paper Presented at the SPE Reservoir Simulation Symposium, Houston, Texas (3–5 February 2003). SPE-79714-MS. Martienssen, W., Hall, K.R., Dykyj, J. et al. (1999). Vapor Pressure and Antoine Constants for Hydroncarbons and Sulfur, Selenium, Tellurium and Halogen Containing Organic Compounds. Berlin Heidelberg: Springer-Verlag. Michelsen, M.L., Yan, W., and Stenby, E.H. (2013). A comparative study of reduced-variables-based flash and conventional flash. SPE Journal 18 (5): 952–959. Morita, N., Singh, S.P., Chen, H.S., and Whitfill, D.L. (1990). Three-dimensional well model pre-processors for reservoir simulation with horizontal and curved inclined wells. Proceedings-SPE Annual Technical Conference and Exhibition Sigma, New Orleans, Louisiana, (23–26 September 1990). 49–64. SPE-20718-MS. Nguyen, N., Chen, Z., Pereira Almao, P. et al. (2017). Reservoir simulation and production optimization of bitumen/heavy oil via nanocatalytic in situ upgrading. Industrial and Engineering Chemistry Research 56 (48): 14214–14230.

References

Norris, S.O. and Piper, L.D. (1990). Modeling fluid flow around horizontal wellbores. Proceedings-SPE Annual Technical Conference and Exhibition Sigma, New Orleans, Louisiana, (23–26 September 1990). 65–75. SPE-20719-MS. Peaceman, D.W. (1993). Representation of a horizontal well in numerical reservoir simulation. SPE Advanced Technology Series 1 (1): 7–16. SPE-21217-PA. Pedersen, K.S. and Christensen, P.L. (2006). Phase Behavior of Petroleum Reservoir Fluids, 1–408. https://doi.org/10.1201/9781420018257. Pedersen, K.S., Milter, J., and Sørensen, H. (2004). Cubic equations of state applied to HT/HP and highly aromatic fluids. SPE Journal 9 (2): 186–192. SPE-88364-PA. Péneloux, A., Rauzy, E., and Fréze, R. (1982). A consistent correction for Redlich-Kwong-Soave volumes. Fluid Phase Equilibria 8 (1): 7–23. Peng, D.-Y. (1986). An empirical method for calculating vapor-liquid critical points of multicomponent mixtures. The Canadian Journal of Chemical Engineering 64 (5): 827–830. Peng, D.Y. and Robinson, D.B. (1976). A new two-constant equation of state. Industrial and Engineering Chemistry Fundamentals 15 (1): 59–64. Peng, D.Y. and Robinson, D.B. (1977). A rigorous method for predicting the critical properties of multicomponent systems from an equation of state. AICHE Journal 23 (2): 137–144. Pujol, L. and Boberg, T.C. (1972). Scaling accuracy of laboratory steam flooding models. Society of Petroleum Engineers-SPE California Regional Meeting, CRM 1972. Bakersfield, California (8–10 November 1972). SPE-4191-MS. Quandalle, P. and Besset, P. (1985). Reduction of grid effects due to local sub-gridding in simulations using a composite grid. Paper Presented at the SPE Reservoir Simulation Symposium, Dallas, Texas (10–13 February 1985). 295-305. SPE-13527-MS. Rojas, A.A., Yuan, C., Emelianov, D.A. et al. (2021). A 3-step reaction model for numerical simulation of in-situ combustion. Society of Petroleum Engineers-SPE Russian Petroleum Technology Conference 2021, Virtual, (12–15 October 2021), RPTC. SPE-206430-MS. Shokrlu, Y.H. and Babadagli, T. (2013). In-situ upgrading of heavy oil/bitumen during steam injection by use of metal nanoparticles: a study on in-situ catalysis and catalyst transportation. SPE Reservoir Evaluation and Engineering 16 (3): 333–344. SPE-146661-PA. Stegemeier, G.L., Laumbach, D.D., and Volek, C.W. (1980). Representing steam processes with vacuum models. Society of Petroleum Engineers of AIME, Journal 20 (3): 151–174. SPE-6787-PA. Sun, X., Zhang, Y., Chen, G., and Gai, Z. (2017). Application of nanoparticles in enhanced oil recovery: a critical review of recent progress. Energies 10 (3): 345. Twu, C.H. (1984). An internally consistent correlation for predicting the critical properties and molecular weights of petroleum and coal-tar liquids. Fluid Phase Equilibria 16 (2): 137–150. Vakhin, A., Sitnov, S., Mukhamatdinov, I. (2022). Improvement of CSS method for extra-heavy oil recovery in shallow reservoirs by simultaneous injection of in-situ upgrading catalysts and solvent: laboratory study, simulation and field application. Paper Presented at the SPE Conference at Oman Petroleum & Energy Show, Muscat, Oman (21–23 March 2022). SPE-200082-MS. Varzandeh, F., Stenby, E.H., and Yan, W. (2017). General approach to characterizing reservoir fluids for EoS models using a large PVT database. Fluid Phase Equilibria 433: 97–111. Wasserman, M.L. (1987). Local grid refinement for three-dimensional simulators. Society of Petroleum Engineers of AIME, San Antonio, Texas (1–4 February1987). 231–241. SPE-16013-MS. Whitson, C.H. (1983). Characterizing hydrocarbon plus fractions. Society of Petroleum Engineers Journal 23 (4): 683–694. SPE-12233-PA. Whitson, C.H., Anderson, T.F., and Soreide, I. (1990). Application of the gamma distribution model to molecular weight and boiling point data for petroleum fractions. Chemical Engineering Communications 96 (1): 259–278.

487

488

10 Numerical Simulation of Catalytic In Situ Oil Upgrading Process

C. H. Whitson and M. R. Brule (2000). Phase Behavior. SPE monograph series vol. 20. Yousef, I., Morozov, V., Sudakov, V., and Idrisov, I. (2021). Cementation characteristics and their effect on quality of the upper Triassic, the lower cretaceous, and the upper cretaceous sandstone reservoirs, Euphrates Graben, Syria. Journal of Earth Science 32 (6): 1545–1562. Zamani, A., Maini, B., and Pereira-Almao, P. (2010). Experimental study on transport of ultra-dispersed catalyst particles in porous media. Energy and Fuels 24: 4980–4988. Zapata, N.B., Morales Mora, O.A., and Mejía Cárdenas, J.M. (2019). Practical kinetic coupling to multi-component and multi-phase flow transport during in-situ heavy oil upgrading processes using an equation of state-based numerical reservoir simulation. In: Society of Petroleum Engineers - SPE Reservoir Characterisation and Simulation Conference and Exhibition 2019, RCSC 2019. Zhang, L., Pieterson, R., Dindoruk, B. et al. (2014). A new and practical oil-characterization method for thermal projects: application to belridge diatomite steamflood. SPE Reservoir Evaluation and Engineering 17 (1): 26–36.

489

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil Khusain Kadiev 1 , Anton L. Maximov 1 , and Jorge Ancheyta 2 1

A.V. Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, 29 Leninsky prospect, 119991 Moscow, Russia Mexicano del Petróleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, Mexico City, 07730, Mexico

2 Instituto

11.1 Introduction Despite the intensive development of renewable energy sources, nonrenewable fossil resources of raw materials for petrochemicals and fuels play a leading role in the global energy balance. Oil will remain the main feedstock to produce motor fuels for air, river, and road transport for now and, apparently, in the near future. In recent years, the intensive development of oil fields has been accompanied by an increase in the weight of oil, its viscosity, the content of resins, asphaltenes, heteroorganic compounds, and metals. Growing reserves of heavy, extra-heavy, and bituminous oils are becoming increasingly important in the oil industry. Distinctive features of such raw materials are a low H : C ratio as compared with light and medium oils, a low yield of light fractions, high viscosity and increased content of asphaltic-resin components (ARC), sulfur, nitrogen, and oxygen compounds, metals (vanadium, nickel, etc.), as well as high coke value. Composition and properties of heavy, extra heavy oils, and bitumen complicate their production and processing. The use of traditional industrial processes based on carbon redistribution (coking, catalytic cracking, and viscosity breaking) for the processing of heavy oil feedstock (HOF) is not effective, since the yield of carbon-enriched by-products (coke, asphaltite, etc.) is higher than for the processing of light and medium oils, and the yield of distillate fractions does not exceed 60%. Hydrogenation technologies are a promising direction for the upgrading of heavy petroleum feedstock and can significantly increase the yield of distillate hydrocarbon fractions (Ancheyta and Speight 2007). During HOF hydrocracking on bifunctional catalysts, including hydrogenating (Mo, Ni, Co, W) and cracking (Al2 O3 , SiO2 , aluminosilicates, etc.,) components, asphaltenes, resins, and heavy metal compounds undergo thermal degradation, which is accompanied by coke and metal deposit formation, and, as a consequence, be the deactivation of supported catalysts (Ahmed et al. 2013; Purón et al. 2013). Improvement in HOF hydroprocessing performance was achieved in hydroconversion processes with a suspension phase of a dispersed catalyst (slurry processes) (Sahu et al. 2015; Biswas et al. 2000; Rana et al. 2007; Zhang et al. 2007; Bellussi et al. 2013a). As evidenced by the results of numerous studies, the greatest activity in the hydrogenation processes of HOF processing is exhibited by dispersed catalysts with particle sizes of less than 1 μm. The features of catalysis in the presence

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

490

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

of nanosized catalysts make it possible to consider it as a separate new direction of heterogeneous and nanoheterogeneous catalysis (Biswas et al. 2000; Khadzhiev 2011, 2016; Kadiev et al. 2016). The use of nanosized catalysts made it possible to achieve the feedstock conversion of 90–95% and the yield of distillate fractions above 80% (Khadzhiev et al. 2014; Khadzhiev 2016; Kang et al. 2019; Kapustin et al. 2021). Processes for HOF hydroconversion in slurry phase are being intensively developed, tested, and implemented on an industrial scale, for example, ENI’s EST process, HTI’s HCAT process, Chevron Lumus Global’s LC-slurry process, etc. In Russia, the refinery of JSC TAIF NK in Nizhnekamsk introduced the Veba Combi Cracking (VCC) and suspension hydrocracking process licensed by KBR (Kapustin et al. 2021). The plant will process 3.7 million tons of vacuum residue per year. Hydroconversion is carried out at 400–450 ∘ C and a hydrogen pressure of 20 MPa, in the presence of 1–5% of a special noncatalytic dispersed additive based on products of incomplete coal combustion. Innovative Russian technologies include hydroconversion of heavy residues in the presence of in situ synthesized nanodispersed catalysts (Russian Slurry Hydro conversion process), developed at the A.V. Topchiev Institute of Petroleum Synthesis and currently being mastered at the JSC TANECO refinery on the pilot plant scale with a capacity of 50 thousand tons per year (Kapustin et al. 2021). Giving the prospects of HOF hydrogenation processing in the presence of dispersed catalysts suspensions, this review presents information on the HOF hydrocracking, with the main synthesis patterns and the catalytic properties of dispersed catalysts considered. Particular attention in the review is given to nanodispersed catalysts, which are the most active in the hydroconversion reactions of HOF.

11.2 Features of the Composition and Properties of Heavy Oil Feedstock Heavy oils and bitumen are characterized by a high content of resins, asphaltenes, heteroatomic compounds, metals, and a low H/C ratio (Meyer et al. 2007). Residues after the HOF atmospheric-vacuum distillation with the yield of 40–60 wt% on the feed are considered as raw material for the further processing by various thermal catalytic methods. The methods used are based on thermolysis of high-molecular-weight components of vacuum residues (VR) with the generation of products with a lower molecular weight, which are the gas and distillate fractions. Molecules of organic compounds which make up VO include various structural elements with different dissociation energies. The weakest bonds in the molecules of organic compounds are the bonds between the aromatic ring and aliphatic side substituents, as well as C—S, C—N bonds (McMillen and Golden 1982; Yu-Ran 2002). The content of heteroatoms and aromatic structural elements increases in the following order: saturated hydrocarbons < aromatic hydrocarbons < resins and asphaltenes. The thermal stability of the said compounds decreases in the same sequence (Ancheyta et al. 2009). Thermal cracking of asphaltenes is accompanied by the formation of solid-phase carbenes and carboides that are one of the main problems in the development of catalytic processes for deep HOF processing. Asphaltene molecules represent a complex three-dimensional formation with a high content of heteroatomic structural elements and polycyclic aromatic fragments with short aliphatic substituents (Mullins et al. 2007; Speight 2014). According to 1 H and 13 C NMR data, the fraction of aromatic carbon in asphaltene molecules reaches 58.6–60.8%, and the aliphatic hydrogen fraction is 9–12% (Silva et al. 2004). The last is predominantly included in the alkyl substituents of aromatic

11.2 Features of the Composition and Properties of Heavy Oil Feedstock

rings. The presence of structural elements with low bond dissociation energy and of polycyclic aromatic nuclei with alkyl substituents in the asphaltene molecules explains the ease of coke formation during the HOF thermal destruction (Alshareef 2020). Thermal degradation of resins is also accompanied by the cleavage from light fragments, such as methyl and ethyl radicals. However, the residual radical fragments of resin molecules polymerize to form not coke, but asphaltenes (Dmitriev and Golovko 2010). In general, thermal degradation of HOF can be described by the scheme (Rahimi et al. 1997; Joshi et al. 2008; Tankov et al. 2017): saturated HC → aromatic HC (+LH) → resins (+LH) → asphaltenes (+LH) → coke (carbenes, carboids) + LH

(11.1)

where LH are light hydrocarbons with a lower molecular weight than the initial component. High polarity of the molecules is an additional factor contributing to coke formation during the thermal asphaltene destruction. A strong interaction is observed between asphaltene molecules, including hydrogen bonding, van der Waals interaction, π–π stacking, acid–base interaction, etc. (Mullins et al. 2007; Ancheyta et al. 2009; Sedghi and Goual 2010; Gray et al. 2011; Chen et al. 2015). As a result, associates with different molecular weights are formed. The association degree for asphaltene molecules depends on the properties of the medium, asphaltene concentration, and temperature. The greater is the difference between the solubility parameters of asphaltenes and the medium, the greater is the degree of association. HOF components have different solubility values: 15.9–16.5 for saturated HC, 20.2–20.8 for aromatic HC, and 19.6–20.5 and 19.6–21 MPa0.5 for resins (Akbarzadeh et al. 2004). Therefore, the group composition affects the state and aggregative instability of asphaltenes in raw materials (Ancheyta et al. 2009; Sedghi and Goual 2010; Chen et al. 2015; Tankov et al. 2017). When the difference between the solubility parameters of asphaltenes and the medium is more than 1.5–2.5 MPa0.5 , the associates of asphaltene molecules lose their stability, resulting in the formation of asphaltene precipitate (Akbarzadeh et al. 2005). As the properties of the medium, asphaltene concentration, pressure, or temperature change, asphaltene associates can become larger, that makes asphaltene dispersions unstable and promotes coke formation in thermal processes (Chen et al. 2015). To characterize the stability of asphaltenes under HOF processing conditions, the colloidal instability index (CI) is used (Stratiev et al. 2020) CI = (Csat + Sasph )∕(Sar + Sresin )

(11.2)

where Csat , Sasph , Sar , and Sresin are the content of saturated hydrocarbons (paraffins + naphthenes), asphaltenes, aromatic hydrocarbons, and resins, respectively. At the initial stage of thermal cracking, an increase in the asphaltene concentration in the liquid phase occurs without the formation of the solid coke phase (Ancheyta et al. 2009). After the critical concentration is reached, the formation of solid asphaltenes begins, accompanied by the intensive coke formation (Ancheyta et al. 2009; Rogel et al. 2013). At this stage of thermal cracking, the group composition of the liquid phase changes, the content of resins decreases, and the content of light hydrocarbons increases, resulting together in the reduced asphaltene colloidal instability and increased coke formation (Pevneva et al. 2020). Under the hydrogen atmosphere, in the presence of active hydrogenation catalysts, radical fragments of asphaltene thermal degradation products are stabilized due to hydrogen addition, leading to a significant decrease in coke yield (Ancheyta et al. 2009; Panariti et al. 2000; Nguyena et al. 2018).

491

492

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

11.3 Main Directions of Processing of Oil Residues, Heavy Oils, and Bitumens Further processing of heavy oils and bitumens at existing refineries focused on the processing of light and medium oils is connected with difficulties due to the deterioration of their characteristics: low H/C ratio and high content of heavy components such as resins, asphaltenes, and heteroorganic compounds. Processing schemes for heavy oils and bitumens include, as a rule, atmospheric and vacuum distillation units. Vacuum distillation residues are recycled using secondary processes to produce additional highly profitable products. Secondary processes for vacuum residues are based either on the partial carbon removal (deasphalting, thermal, or thermal catalytic cracking), or on the hydrogen addition to increase H/C ratio in the residues (hydrocracking, hydroconversion processes) (Ancheyta and Speight 2007).

11.3.1

HOF Upgrading with Carbon Part Removal

11.3.1.1 Deasphalting

Asphaltenes are the most polar HOF components (Ancheyta et al. 2009). Therefore, they are almost insoluble in C3 —C7 alkanes with the solubility parameter below 15 MPa0.5 . Mixing the oil with hydrocarbons C3 —C7 leads to asphaltene precipitation. The process of crude oil deasphalting is based on this principle. As a result of the asphaltene removal from feedstock, the content of heteroatoms decreases in the deasphalted oil, the coking capacity decreases, and the H/C ratio increases. After the asphaltene removal, the possibilities for further processing of deasphalted oil are significantly expanded (Figure 11.1) (Magomedova et al. 2019). 11.3.1.2 Thermal Cracking

Thermal cracking is used most often for the processing of HOF (including oil distillation residues): delayed coking, fluid coking (thermocontact coking in a fluidized bed in a continuous mode); Production of lubricating materials

Selective extraction

Production of motor fuels and electrode coke Deasphalted oil

Catalytic cracking Hydrocracking Delayed coking

Crude oil

AVD distillation SDA Vacuum residue Asphalt

Base oils

Motor fuel components Electrode coke

Compounding/oxidation

Bitumen binders

Coking/thermal polycondensation

Petroleum coke/ pitch

Gasification/combustion

Energy/syngas

Slurry-phase hydroconversion

Motor fuel components

Figure 11.1 Possible ways for processing the products of solvent deasphalting of vacuum residues into two fractions: deasphalted oil and asphalt.

11.3 Main Directions of Processing of Oil Residues, Heavy Oils, and Bitumens

flexicoking (coking with subsequent gasification of coke in a continuous mode), and visbreaking (Castaneda et al. 2014). Thermal cracking processes are attractive for oil refineries as the safest and most well developed, as well as from an economic point of view; however, the yield of distillate fractions does not exceed 60% (Stratiev et al. 2014). When processing VR in the long term, heavy oils and natural bitumens, characterized by a low H/C ratio and high coking capacity, the yield of distillate fractions will be low. 11.3.1.3 Catalytic Cracking

The high coking capacity and the presence of significant metal amounts limit the possibilities of HOF processing in the catalytic cracking. Thus, in the fluid catalytic cracking (FCC) process, the maximum allowable coke value is below 5%, with the restrictions on the metal content in the feedstock. For this reason, vacuum gas oils, coking gas oils, deasphalting residues from atmospheric and the vacuum oil distillation are used as feedstock in FCC units, as a rule (Oloruntoba et al. 2022). To improve the performance of thermal cracking, new technical solutions are sought. In particular, Fluid Oil Limited has developed a technology FluidOil VHTL (Viscositor Heavy-To-Light) (VHTL 2019) for thermal cracking of heavy oil residues. The process is based on a reactor with a very short residence time of the raw material (about two seconds). For the rapid feedstock heating, a contact heat transfer system was used in which hot sand subjects are HOF components to the thermal splitting. The raw material enters the reactor by spray-up in the water vapor stream (Figure 11.2). A large sand-to-oil ratio and efficient stirring are used to ensure high turbulence in the injection zone. Along with coking, the VHTL process involves partial steam reforming to form hydrogen and partial hydrogenation of the resulting products. According to the developers, the VHTL process is more reliable and easier to operate than FCC.

11.3.2

Hydrogenation Processes for HOF Upgrading

Hydrogenation processes for HOF distillation residues make it possible to increase the yield of distillate fractions up to 80%, but the share of such processes in the total volume of HOF processing does not exceed 30% due to higher capital and operating costs (Castaneda et al. 2014). Hydrogenation processes using reactors with a stationary, moving, or fluidized bed of bifunctional granular catalysts have received the greatest distribution in the vacuum residues processing. Catalysts include hydrogenating components (Mo, Ni, Co, W sulfides) and support (alumina, silica, titania, aluminosilicates, zeolites, etc.). In addition to the hydrogenating function, catalysts are active in cracking reactions. Pre-fractionation distillate to SCO

C4- Gas Reactor product vapor Reactor liquid product to SCO

Bottoms to reactor

Feed pre-fractionation Atm. and Vac. distillation

Figure 11.2

Reactor and reheater circulates hot sand

FluidOil VHTL scheme.

Product quench and recovery

493

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

AT HOF hydrocracking in reactors with a fixed catalyst bed, the feedstock, as a rule, is preliminarily subjected to deasphalting to prevent the catalyst deactivation by coke and metal compound deposits as a result of asphaltene thermal destruction. In processes with moving and fluidized catalyst bed, the reactor design makes it possible to permanently withdraw a part of deactivated catalyst from the process, replacing it with the fresh one. Slurry processes use dispersed catalysts, highly active in hydrogenation reactions. The process is organized in such a way that the catalyst suspension enters the reactor together with the feedstock and is removed from the reactor together with the hydroconversion products. The main parameters of hydrogenation processes for HOF upgrading are presented in Table 11.1.

11.3.3

Efficiency Analysis HOF Processing

Thermal cracking processes are the most widely used for processing oil residues, as they are characterized by low capital and operating costs (Figure 11.3). The main disadvantages of thermal and thermal catalytic cracking processes are the formation of significant amounts of by-products: coke, asphalt, flue gases, and a low yield of low-quality distillate fractions that require special approaches for further upgrading. Hydroprocesses make it possible to increase the conversion of raw materials up to 90–95% and the yield of distillate fractions up to 80%. The most efficient are the processes of hydroconversion in the catalyst suspension phase (Table 11.1). The use of highly dispersed nanosized catalysts in HOF hydrotreatment in the suspension phase promotes the rapid absorption of hydrogen and deactivates intermediate free radical fragments, thereby suppressing coke formation, increasing the overall conversion to 95% and enhancing the yield of distillate fractions. Processes can be used for feedstocks with a high content of asphaltenes, metal compounds, and high coking capacity. The obvious advantages of HOF hydroconversion in the suspension phase of dispersed catalysts contribute to the intensive development of technology and the industrial implementation of such processes. The first industrial facility using slurry technology for the oil residue hydroconversion 9.0

8.0

Hydrocracking (90%)

Hydrocracking (60%)

VB/CSD

Flexi coking

6.0

Fluid coking

7.0

Delayed coking

Operating cost ($/bb1)

494

Processes Figure 11.3

Comparison of operating costs for various HOF processing methods.

Table 11.1

Comparison of various heavy oil feedstock hydrotreatment processes (Taken from different sources). Fixed catalyst bed

Moving and fluidized catalyst bed

Industrial and developed processes

OCR, UFR, Up-flow reactor Hycon, Bunker type reactor Hyvall

H-Oil, T-Star, LC-Fining, LC-MAX

EST

LC-SLURRY

Veba combi-cracking

HCAT

Uniflex

Precursor

Supported Mo, Ni, Co, W, 3–6 mm

Supported Mo, Ni, Co, W, 0.5–3 mm

Mo Hexanoate

Ultradispersed catalyst ISOSLURRYTM

Fe2 O3 + additives, 1–50 μm

Mo Octoate

MicroCatTM catalyst MoS2

Reactor type

Slurry reactor with catalyst suspension phase

MoS2 , 95

>90

94

95

Catalyst rate per feed, %

—

1–2

0.2

—

2–5

—

—

Licensor

Chevron Lumus Global (CLG), Shell (Bunker flow) Axen (Swing reactor), Shell IFP (Axen)

Axen (HRI/IFP),Chevron Lumus Global (CLG)ABB Lummus

ENI

Chevron Lumus Global (CLG)

KBR

Headwaters Technology Innovations Group

UOP

Table 11.1

(Continued) Fixed catalyst bed

Moving and fluidized catalyst bed

Advantages

Low loss of catalyst with products, high mechanical strength of the catalyst, low catalyst consumption

Disadvantages

Feedstock with low metal and asphaltenes content. Uneven temperature profile across the catalyst bed. Blocking of active acid sites by coke and metals, diffusion difficulties in mass transfer inside the granule, high risk of coking

Improved mass transfer, partial catalyst regeneration and recirculation, uniform temperature profile throughout the reactor volume, feedstock contains asphaltenes and metals The need to constantly replace the spent deactivated catalyst with a fresh one, high linear velocities in the reactor to keep the catalyst in suspension, high catalyst consumption

Reactor type

Slurry reactor with catalyst suspension phase

Maximum surface accessibility, absence of pores, high concentration of active sites on the surface, high specific activity, low catalyst consumption, low concentration in the reaction zone, operation at low pressures in the reactor, operation with feedstock of extremely high asphaltene, and metal content

The need for careful dispersion of the catalyst in the feedstock Possible takeaway with light products Difficulty in extraction from hydroconversion products

11.3 Main Directions of Processing of Oil Residues, Heavy Oils, and Bitumens Gas

Reaction products

Fractionation system

H2

Slurry reactor

H2 recovery

Refined products

Catalyst precursor Catalyst and residue recycle Purge

Feed

EST simplified process scheme

Figure 11.4

Schematic plot of the EST process.

is a complex at the ENI refinery in Sannazaro de Burgondi, commissioned in 2013. The ENI Slurry Technology (EST) cracking process uses an oil-soluble Mo-containing catalyst precursor (Bellussi et al. 2013a). The process is carried out at 430 ∘ S, hydrogen pressure of 13–16 MPa, in the presence of 1000–3000 ppm of Mo (Figure 11.4). The EST plant has a design capacity on 3600 tons of tar per day and allows processing vacuum residues into Euro V standard diesel fuel, naphtha, aviation fuel, liquefied gas, and other products. The HCAT/HC process is licensed by headwaters technology innovation (HTI). The technology is based on the in situ formation of a molecularly dispersed catalyst from organometallic precursors, such as molybdenum octoate, molybdenum 2-ethylhexanoate (Kunnas and Smith 2011). HCAT catalyst is proposed to be used to improve the performance of hydrocracking reactors with a fluidized bed of a bifunctional-supported catalyst. In 2020, JSC LUKOIL Neftekhim Burgas put into operation a technology using the H-CAT nanocatalyst, licensed by an American company (HTI), in alliance with Axens, and the licensor of the H-Oil hydrocracking unit. The complex is the largest catalytic hydrocracking unit in Europe with a throughput capacity of 2.5 million tons of crude oil per year. This is the fourth plant in Europe and the sixth one in the world using the innovative HCAT nanocatalyst system. It is noted that the introduction of the HCAT catalyst precursor in the feed increased the conversion from 6% to 73%. In Russia, significant deposits of heavy oils and bitumen are concentrated on the territory of Tatarstan. For this reason, the oil companies of Tatarstan are showing great interest in the development of new HOF processing processes. Currently, a complex for deep processing of heavy oil residues using the Veba Combi Cracker technology is in commercial operation. Hydroconversion of tar is carried out at 400–450 ∘ C and a hydrogen pressure of 22 MPa, in the presence of 2–5% noncatalytic highly dispersed additive. The process diagram is shown in Figure 11.5. The process is carried out in two stages (Dohler et al. 1987; Kapustin et al. 2021). First, the thermal cracking of the tar is carried out in the presence of a finely dispersed adsorbent flowing at an elevated pressure of about 20.0 MPa in a once-through reactor. The feed enters the reactor bottom and exits at the reactor top. The gases and reaction products enter into the hydrocracking reactor, where the process is carried out in the presence of hydrogen at the same pressure and the temperature of 430–450 ∘ C. The second stage is hydrocracking of the vacuum residue to obtain naphtha, kerosene, jet fuel, diesel fuel, and hydrocracking residue (fraction 520 ∘ C+). The residue is taken from the bottom of the once-through reactor and consists of a concentrate containing asphaltenes, an additive, and metals (vanadium, nickel, etc.).

497

498

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

IX I II

X 4

9

7

V 3

1

6

VI

2

VII XI

5

III

IV

8

VIII

Figure 11.5 Process scheme for vacuum oil thermal hydrocracking in a suspended adsorbent bed using the Veba Combi Cracker technology. 1 is heat exchanger; 2 is furnace; 3 is reactor; 4 is hot separator; 5 is vacuum column; 6 is second-stage reactor (hydrocracking); 7 is a separator; 8 is atmospheric column; 9 is gas preparation unit; I is raw materials; II is adsorbent catalyst; III is hydrogen; IV is residue (>520 ∘ C); V is gases S 1–S 4; VI is naphtha; VII is diesel fuel; VIII is vacuum gas oil; IX is hydrogen sulfide-containing gas; X is hot hydrogen gas; XI is water vapor.

HBG

Hydrogen

Amine treatment and concentration HBG

Modifier

Feed

NH3, H2S, H2O

Heating reaction block HC

Separation block

Atmospheric fractionation

Catalyst precursor preparation block

Gasoline

Diesel fuel

Recycle

APM

Fuel gas

Vacuum fractionation

Light gasoil

Vac.gasoil to processing

To catalyst precursor regeneration

Figure 11.6

Block diagram of TATNEFT hydroconversion unit.

The refinery of AO TANECO (TATNEFT) and TIPS RAS has developed the technology of tar hydroconversion in the presence of a suspension of MoS2 nanoparticles synthesized in situ from inversed emulsion of a water-soluble precursor, ammonium paramolybdate (APM). The process is carried out at a temperature of 430 ∘ C, a hydrogen pressure of 7 MPa, and in the presence of 0.05% Mo. The block diagram of a pilot plant (demo plant) with a capacity of 50 thousand tons per year is shown in Figure 11.6 (Kadiev 2015).

11.4 Catalysts for HOF Hydroprocessing

11.4 Catalysts for HOF Hydroprocessing 11.4.1.1 Morphology of Catalysts

As seen from Table 11.1, a catalyst is a factor that determines the hardware technological scheme and process performance. Two classes of catalysts are used in HOF hydroprocessing. In batch reactors with a fixed, moving, and fluidized bed, tableted bifunctional catalysts are used, including a hydrogenating component (Mo, Ni, Co, and W sulfides) and a carrier (aluminum, silicon, titanium oxides, aluminosilicates, zeolites, etc.) containing acid sites and having activity in cracking reactions (Ancheyta and Speight 2007; Rana et al. 2007; Kapustin et al. 2021). These catalysts have a developed system of macro- and micropores, a large specific surface area. Granules with a size of 3–6 mm are used in reactors with a fixed catalyst bed, and those of 0.5–3 mm are used in reactors with a moving and fluidized bed. In terms of the structure and process mechanism, these catalysts are similar to those used in the hydrotreatment and hydrocracking of distillate fractions (Ancheyta and Speight 2007; Peng et al. 2018). They are highly efficient in hydrocracking of petroleum distillate fractions, but in the HOF processing, they are rapidly deactivated due to deposits of coke and metal compounds that resulted from asphaltene cracking and decomposition of metal complexes (V, Ni, Fe), (Rana et al. 2007; Ahmed et al. 2013; Purón et al. 2013). For this reason, the feedstock processed in fixed bed reactors is preliminarily subjected to deasphalting. In reactors with a moving or fluidized bed of catalyst, part of the catalyst is constantly withdrawn from the process and replaced with the fresh one. Due to deactivation, bifunctional catalysts do not provide heavy feedstock conversion above 80%. The main purpose of dispersed catalysts is to inhibit the reactions of polycondensation and coke formation. To achieve this goal, catalysts used are active in hydrogenation reactions (Ancheyta and Speight 2007; Rana et al. 2007; Bellussi et al. 2013a; Khadzhiev 2016). The particle size of catalysts significantly affects their catalytic activity. By mechanical grinding, it is possible to obtain powders with particle sizes of 1–50 μm. Such catalysts have been used in the VCC and CanMet HCR processes (Ancheyta and Speight 2007; Rana et al. 2007; Kapustin et al. 2021). The catalysts represented powders containing iron compounds. Due to the low catalytic activity, the consumption of catalysts was 2–5%. Suspensions of ultrafine (or nanosized) catalysts with particle sizes less than 1 μm made it possible to achieve significant progress in the HOF hydroconversion technology (Table 11.1). A relatively small number of chemical elements form compounds that exhibit catalytic activity in HOF hydroconversion. These are elements of groups 6, 7, and 8 of Mendeleev periodic system: Mo, W, Re, Fe, Co, Ni. In this row, molybdenum sulfide is the most efficient in HOF hydroconversion (Rana et al. 2007; Zhang et al. 2007; Khadzhiev 2016; Kang et al. 2019). Suspensions of nanosized particles are synthesized under hydroconversion conditions from oil- or water-soluble precursors previously introduced into the raw material (Bellussi et al. 2013a; Zhang et al. 2007; Khadzhiev 2016; Kang et al. 2019). High polarity of precursor molecules leads to the formation of joint associates with asphaltene molecules, thus increasing the efficiency of hydroconversion (Kunnas and Smith 2011; Kadieva et al. 2011; Kadiev et al. 2018a). The mechanism of catalysis by molybdenum disulfide is due to the peculiarities of its layered crystal structure (Chianelli et al. 2006; Kadiev et al. 2018a; Peng et al. 2018). In the MoS2 crystal, molybdenum atom is coordinated with six sulfur atoms. At the edges of the MoS2 plates, there are coordinately unsaturated Mo atoms, which are active in the hydrogenation reactions. The maximum concentration of catalytically active centers is observed on the brims located at the edges of the upper and lower crystal plates. Under hydroconversion conditions, coordinately unsaturated centers with vacancies of sulfur atoms are formed at the edges of MoS2 slabs. Hydrogen and hydrogen sulfide are sorbed on the active centers with the formation of Mo—H and S—H bonds. These bonds are unstable and can rapidly transfer hydrogen to asphaltenes and radical fragments

499

500

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

of molecules adsorbed on the catalyst particle surface (Chianelli et al. 2006; Ancheyta et al. 2009; Kim and Lee 2019). The catalytic activity of MoS2 in hydrogenation reactions increases with the increasing dispersity (the ratio of the number of Mo atoms on the plate edge to the total number of Mo atoms in crystal) (Thian 2008; Bruix et al. 2015; Kim and Lee 2019). For catalyst particles of the same shape, catalyst dispersion increases with the size decreased. Catalyst efficiency in the inhibition of coke formation depends on the concentration of catalytically active centers per unit volume of the reaction medium (Chianelli et al. 2006; Kim and Lee 2019). A significant effect can be achieved by changing the morphology of the catalyst particles, reducing the size, and increasing the particle dispersion (Chianelli et al. 2006; Thian 2008; Bruix et al. 2015; Kim and Lee 2019). Another method to increase the concentration of active sites in the reaction volume is the increase of catalyst content in the reaction zone (Fixari et al. 1994; Panariti et al. 2000; Zhang et al. 2007; Bellussi et al. 2013a; Go et al. 2018). Thus, at hydroconversion of heavy oil in the presence of suspension of MoS2 nanoparticles synthesized from a water-soluble precursor, an increase in the catalyst concentration from 0% to 0.05% leads to a decrease in the coke yield from 7% to 0.1% (Khadzhiev et al. 2018). 11.4.1.2 Hydroconversion Conditions in Slurry Processes

Catalysis in slurry processes is heterogeneous and proceeds in the liquid phase. With the pressure increase, there increases the concentration of hydrogen in the liquid phase and the degree of hydrogen saturation of the catalytically active MoS2 edges. Increasing the hydrogen pressure up to 7–16 MPa significantly reduces the yield of compaction products (Zhang et al. 2007; Bellussi et al. 2013a; Bruix et al. 2015). Since hydroconversion processes are based on the thermal cracking of feedstock, they are characterized by similar patterns. The product yield and raw material conversion increase with the increasing temperature and reaction time (Panariti et al. 2000; Go et al. 2018; Kadiev et al. 2018a; Khadzhiev et al. 2018; Lim et al. 2018). Optimum values of temperature and reaction time are determined from the conditions of minimal coke formation. Summarizing the known data, we can formulate the following conditions for reducing the formation of coke during HOF hydroconversion in the presence of highly dispersed Mo-containing catalysts: ● ● ●

Increase in the catalyst particle dispersion Increase in the catalyst concentration in the range from 100 to 2000 ppm Mo Increase in hydrogen pressure up to 7–16 MPa

11.5 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion Development of efficient processes for the deep heavy feedstock processing involves the creation of highly dispersed next-generation catalysts. In this regard, the main methods of synthesis and properties of dispersed nanosized catalysts used in slurry processes of HOF hydroconversion should be considered. There are many methods for the synthesis of nanosized catalysts (Kang et al. 2019; Somwanshi et al. 2020). Most of these methods have not found wide practical application in HOF hydroprocessing. As it follows from the data given in the literature and patents, it is expedient to use the dispersions of catalyst particles synthesized directly in the feedstock from oil- or water-soluble precursors, stabilized in a hydrocarbon medium without the use of traditional solid carriers, in HOF hydroconversion processes.

11.5 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion

Before considering examples of hydroconversion using the dispersed catalysts, it is worthwhile to make a comment about the criteria used to evaluate the effectiveness of catalysts. For this purpose, indicators of the feedstock conversion depth and the yield of distillate fractions with minimal coke formation are often used. However, in slurry HOF processing, these indicators are determined by the intensity of thermal cracking processes (Del Bianco et al. 1994). In a number of studies with experiments without catalyst, these indicators were higher than in the experiments with catalyst. More important for evaluating the activity of nanosized catalysts are data on changes in the H/C atomic ratio, group composition, asphaltene conversion degree, and the yield of compaction products (coke) (Nguyena et al. 2016).

11.5.1

Catalysts Derived from Oil-Soluble Precursors

Naphthenates, octanoates, 2-ethylhexanoate, carbonyls, acetylacetonates, and metal carbonyls are most often used as oil-soluble precursors of highly dispersed catalysts in HOF hydroconversion (Fixari et al. 1994; Bruix et al. 2015; Go et al. 2018; Khadzhiev et al. 2018; Kang et al. 2019). The use of such precursors makes it possible to distribute the catalyst in the feed at the molecular level, thus providing the production of catalyst particles with small sizes and high catalytic activity. Various studies have experimentally compared the activities of dispersed catalysts formed from oil-soluble precursors of different metals. Hydroconversion of vacuum residue from oil distillation in the presence of catalysts synthesized from Mo, Ru, Co, Ni, and Fe naphthenates was studied by Fixari et al. (1994), Thian (2008) Bruix et al. (2015) Go et al. (2018), and Kim and Lee (2019). The experiments were carried out in an autoclave, the initial hydrogen pressure was 7.5 MPa, the temperature was 440 ∘ C, the duration of the experiment was two hours, and the concentration of catalysts was 230–1400 ppm (in terms of metal). The catalyst was sulfidized with sulfur from the feedstock. It was found out that the coke yield in the order: Mo, Ru, Co, Ni, Fe, increased from 2.2% to 12.8%. In the same sequence, the hydrodesulfurization degree decreased. With molybdenum, concentration increased from 0 to 1400 ppm, the coke yield decreased from 15% to 2%. The results of studying the properties of catalyst suspensions synthesized in situ from iron-containing oil-soluble precursors that are given by Kadiev et al. (2019c). The hydroconversion of the vacuum residue was carried out in an autoclave with a constant flow of hydrogen at 435 ∘ C, a hydrogen pressure of 7 MPa, for two hours. Iron acetylacetonate Fe(CH3 COCH=C(CH3 )O)3 (IAA), ferrocene Fe(C5 H5 )2 (FC), and iron oleate (C17 H33 COO)3 Fe (IO) were used as precursors. The latter were introduced to reach the concentration of 0.2 wt% iron in tar. An additive of elemental sulfur was used as a sulfidizing agent with the concentration of 1 wt% per tar. In the comparative experiments, APM in the form of an aqueous solution in tar emulsion was used as a precursor. As compared with MoS2 , the in situ synthesized iron-containing catalysts had cracking properties to a greater extent and hydrogenating properties to a lesser extent, as evidenced by the yields of gaseous products and coke, being higher in the experiments with IO and FC than in the experiment without a catalyst. It was concluded that when processing HOF with low coking values, for example, after preliminary deasphalting, the use of iron-containing catalysts may be reasonable. A comparison of the catalytic activity of suspensions of WS2 and MoS2 nanoparticles in the tar hydroconversion was carried out in (Jeong and Lee 2019). Tungsten and molybdenum carbonyls were used as precursors. Hydroconversion was carried out in an autoclave at temperature of 419 ∘ C, hydrogen pressure of 9.5 MPa, in the presence of 0.1 wt% W or Mo. According to the TEM data, WS2 consisted of monolayer plates with an average size of 13.9 nm. When using W(CO)6 with the DMDS addition, an increase in the activity of the tungsten catalyst and a decrease in the average plate size

501

502

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

to 10.6 nm were observed. Suspensions of WS2 and MoS2 showed similar catalytic activity in the tar hydroconversion. Hydroconversion of atmospheric residue (AR) was carried out using molybdenum naphthenate, nickel octoate, and vanadium acetylacetonate acac precursors (Nguyen et al. 2015). The reaction was carried out in an autoclave at 450 ∘ C, residence time of one hour, and catalyst concentration of 300–900 ppm with the addition of DMDS as sulfidizing agent. As experiments showed, individual sulfides formed from Ni and V precursors did not control thermal cracking processes, resulting in the coke yield exceeding 4%, while in the experiments with Mo precursor this value did not exceed 0.2%. In the AR hydroconversion with a mixture of Ni and Mo precursors (Mo 300 ppm, Ni 80 ppm), a promoting effect of nickel was observed: conversion, desulfurization degree, H/C ratio in the hydrogenation product were higher than in experiments with individual precursors. Vanadium acetylacetonate did not exhibit the catalytic properties of the promoter. It was established that single-layer curved plates of MoS2 on carbon were formed from MoNaph with an average length of about 6 nm and dispersion of 20%. One plate included 300 atoms an average. V3 S4 were agglomerates of spherical particles with a size of 200 nm. Ni catalyst (600 ppm) consisted of bulk Ni3 S2 particles of 5–30 nm in size. For mixed Ni—Mo catalyst, MoS2 slabs were deposited on the surface of Ni3 S2 crystallites, providing a synergistic effect. The influence of the cobalt promoter on the activity of the Mo precursor synthesized from the oil-soluble Mo precursor was studied in Jeon et al. (2011). The hydroconversion of Athabasca oil sands bitumen was carried out in an autoclave at initial pressure of H2 of 7 MPa and temperature of 420 ∘ C. Cobalt naphthenate (Co naph), molybdenum octoate (Mo octo), and a mixed oleate synthesized with a ratio of Co/Mo = 1 (CoMo) were used as precursors with the assumed content of 1000 ppm for metals. It was established that the hydrogen absorption, 500 ∘ S + fraction conversion, and distillate fraction yields increase in the order: “without a catalyst” < Co naph < Mo octo < Co naph + Mo octo < CoMo. The coke yield decreased in the same sequence. The experiments confirmed the effectiveness of the promoter when using oil-soluble molybdenum-containing precursors. Manek and Haydary (2017) studied the effect of adding an oil-soluble molybdenum precursor (0.021% Mo per feedstock) on the hydroconversion of a vacuum residue in an industrial reactor with a fluidized bed of supported Ni–Mo/Al2 O3 catalyst granules. The process was carried out in three successive reactors at 413–419 ∘ C and 18 MPa. It was established that the addition of a dispersed catalyst reduced the yield of coke residue from 0.28 wt% down to 0.24 wt%. León et al. (2017) carried out an experimental comparison of the processes of thermal cracking and catalytic hydrocracking of the vacuum residue of Colombian oil and of group composition components. The experiments were performed in an autoclave at 430 ∘ C and 13.5 MPa for 25 min in a nitrogen or hydrogen atmosphere. Molybdenum naphthenate, in concentration 1000 ppm Mo, was used as a catalyst precursor. Additionally, catalytic hydrocracking was carried out after a part of asphaltenes had been removed from the tar by sorption with activated carbon at 380 ∘ C, 12.2 MPa H2 for 60 minutes. It was established that, at thermal cracking of maltenes, a small yield of gas and coke was observed (about 5% in total), while at catalytic hydrocracking, the yield of these products decreased to 2.6%. At thermal cracking of asphaltenes, 35% of coke was formed. In the presence of catalyst, this indicator decreased to 27%. Preliminary sorption removal of the part of asphaltenes from the vacuum residue with activated carbon reduced the yield of coke during hydroconversion. It was found out that the yield of coke during hydrocracking was significantly lower than that calculated from the data given above. This effect was attributed to an increase in the solubility of asphaltenes in the presence of resins contained in the feed.

11.5 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion

1 nm

10.00 nm

(a)

(b)

Figure 11.7 Transmission electron micrographs (HRTEM) of THF-insoluble vacuum residue hydroconversion product containing MoS2 plates synthesized from an oil-soluble precursor (molybdenum hexanoate): (a) dispersed catalyst; (b) a single layer of MoS2 with a superimposed atomic model.

The activity of oil-soluble catalysts in hydrogenation and removal of heteroatoms was studied by Kim and Curtis (1990). There was established by the high catalytic activity of MoS2 suspensions formed in a liquid hydrocarbon medium from molybdenum naphthenate (MoNaph) and molybdenum octoate (MoOct) with additions of elemental sulfur in the hydrogenation reactions of organic compounds presenting in the HOFs: naphthalene, decalin, benzothiophene, ortho-cresol, benzofuran, quinoline, and indole. Experiments were carried out in an autoclave at 380 ∘ C, hydrogen pressure of 18 MPa, and a catalyst content of 3000 ppm Mo. The sizes of the catalyst particles in the suspension were 50–250 nm. The catalytic activity of the dispersed catalyst was several times higher than that of the granular NiMo/Al2 O3 catalyst. It was noted, that at catalyst concentration reduced to 1500 ppm, the reaction rate decreased several times. The EST hydroconversion process developed by ENI uses a dispersed catalyst formed from Mo(IV) 2-ethylhexanoate (Bellussi et al. 2013b; León et al. 2017). The process is carried out at 430 ∘ C, hydrogen pressure of 13–16 MPa, and 1000–3000 ppm of molybdenum. The active phase of the catalyst consists of monolayer MoS2 particles of 5–20 nm in size (Figure 11.7). The structure of the catalyst is maintained even after the repeated recirculation of vacuum residue. The given examples give evidence to the high catalytic activity of MoS2 suspensions synthesized from oil-soluble precursors in the reactions of activated hydrogen addition to radical fragments of thermal destruction of asphaltenes and in the inhibition of coke formation processes. When using oil-soluble precursors, monolayer crystals are formed with the minimum size of 5–20 nm and high degree of dispersity. Review (Nguyena et al. 2016) analyzes the advantages and disadvantages of using the oil-soluble molybdenum-containing precursors. It is noted that suspensions of MoS2 particles obtained from such precursors exhibit high activity in the inhibition of coke formation, however the high cost of their synthesis and the difficulty in extracting the catalyst from hydroconversion products prevent their use in the suspension hydrocracking process.

11.5.2

Catalysts Synthesized from Water-Soluble Precursors

The formation of nanosized particle suspension from an inverse emulsion of a catalyst precursor in a hydrocarbon medium occurs under the reaction conditions because of sequential processes: evaporation of emulsion water from emulsion droplets, reduction of precursor particles with hydrogen,

503

504

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

and their sulfidizing with sulfur-containing feedstock components or an added sulfidizing reagent (Khadzhiev et al. 2014). The patterns of formation and stability of inverse emulsions of precursor aqueous solutions in hydrocarbon medium containing synthetic and native petroleum surfactant components were studied in (Kadiev et al. 2019b). The catalyst precursors were water-soluble salts (NH4 )6 Mo7 O24 , (NH4 )2 WO4 , Fe(NO3 )3 , and Ni(NO3 )2 . Oleic acid, SPAN 80, and a mixture of SPAN 80 + TWEEN 80 were used as surfactants. We also studied emulsions with native petroleum surfactants: fractions of asphaltenes, resins and oils (a mixture of hydrocarbon components). Toluene, heptane, and vacuum residue of West Siberian oil distillation (tar) were used as dispersion media. A model mixture of toluene and heptane with the Hildebrand solubility parameter (𝛿) equal to the tar solubility parameter, 17.83 MPa0.5 , was used as hydrocarbon phase for the preparation of emulsions. An ultrasonic disperser was used to obtain emulsions. It was established that the average diameter of the globules of an aqueous solution emulsion (D, m) is related to the interfacial surface tension (𝜎, mN/m), the content of the aqueous solution in the emulsion (C, wt%), density of the precursor solution (d, 20 ∘ C, g/m3 ), and energy spent on emulsification (W), according to the equation: D = 60C𝜎∕Wd

(11.3)

All other conditions being equal, the size of the emulsion droplets and its stability increased with decrease in the interfacial tension at the boundary of two liquid phases. Asphaltenes, in the structure of which there were non-polar hydrocarbon groups and polar hydrophilic groups, including sulfur, nitrogen, oxygen atoms, exhibited surface-active properties comparable to the those of synthetic surfactants, such as Span-80 and Tween-80. Resins and hydrocarbon components of tar had weak surface-active properties. With the concentration of asphaltenes in a mixture of toluene/heptane increased from 0% to 5%, the interfacial tension coefficient at the water-dispersion medium boundary decreased from 33 to 3.4 mN/m. The average droplet diameter of the emulsion obtained by emulsifying water in a mixture of toluene-heptane in the presence of 5% asphaltenes was 330 nm. As a result of the research, it was established that HOF components (resins and asphaltenes) have surface-active properties, which made it possible to obtain stable inverse emulsions of dispersed catalyst precursors without use of synthetic surfactants. The emulsion method for the synthesis of suspensions of nanosized particles of metal oxides and sulfides was studied in (Khadzhiev et al. 2013a). To obtain nanosized MoO3 and Al2 O3 powders in pentadecane, aqueous solutions of APM and aluminum nitrate were emulsified with the addition of surfactants (sodium bis(2-ethylhexyl)sulfosuccinate and Span-80). The resulting microemulsion was subjected to heat treatment at 150–250 ∘ C. Varying the conditions of emulsification and heat treatment, the conditions for obtaining metal oxide suspensions with particle sizes less than 500 nm were determined. Using the emulsion method, ultra- and nanodispersed particles of multicomponent catalysts such as Mex Oy /Al2 O3 , Mex Sy /Al2 O3 (Me = Mo, Ni, Co, Fe) were synthesized by thermal decomposition of inverse microemulsions (ME) of aqueous solutions of precursors (Khadzhiev et al. 2013b). Pentadecane with additives of synthetic surfactants, sodium dioctylsulfosuccinate (AOT) and Span 80, or light catalytic cracking gas oil (LCCG) containing native stabilizing components (asphaltenes, resins, polycyclic aromatic hydrocarbons) were used as dispersion media. The initial precursors were APM, aluminum, iron, nickel, and cobalt nitrates. Aqueous salt solutions were dispersed in a dispersion medium using an ultrasonic disperser. Microemulsions were thermally treated for one hour in a stirred reactor at atmospheric pressure and temperature of 150–240 ∘ C,

11.5 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion

200 nm

150 nm

(a)

(b)

Figure 11.8 TEM images of binary nanosized particles with a shell-core structure, obtained by sequential introduction of precursors into a microemulsion in pentadecane with AOT addition: (a) MoO3 /Al2 O3 ; (b) FeS/Al2 O3 .

using N2 as bubbling agent. In experiments with sulfidation, in addition to an aqueous solution of the precursor, a solution of (NH4 )2 S was introduced as a sulfur donor. It was shown that the average diameter, structure, and morphology of synthesized 2- and 3-component ultra- and nanodispersed particles depended on the method of the precursor introduction, on the composition of MEs, and the conditions of their processing. For synthesis of spherical binary nanoparticles with a “shell-core” structure, sequential introduction of precursors appeared preferable, when the core phase of the particle (Al2 O3 ) was synthesized at the first stage, and then the component becoming the shell (Mex Oy , Mex Sy ) was generated. Figure 11.8a shows TEM images of binary Al and Mo containing particles obtained by successive introduction of precursors into a microemulsion in pentadecane with the addition of AOT. Figure 11.8b represents TEM image of FeS/Al2 O3 particles synthesized from emulsions in pentadecane with the addition of AOT (sequential scheme for the introduction of precursors, Al/Fe = 3 (by weight), S/Fe = 2.0 (mole), T = 240 ∘ C). The catalytic activity of polyfunctional nanosized catalysts synthesized in situ from reversed emulsions of aqueous solutions of precursors was studied in the hydroconversion of natural bitumen (Kadiev et al. 2013). Aqueous solutions of precursors (APM, nitrates of Al, Ni, Co, Fe, and the sulfiding component, ammonium sulfide) were emulsified in bitumen diluted with light catalytic cracking gas oil. In accordance with the results reported in (Khadzhiev et al. 2013b), two methods of the precursor introduction were used: simultaneous and sequential. In the sequential method, the microemulsion of the first component was subjected to heat treatment, then the second component was emulsified in the resulting product. With this method of synthesis, the product of thermal decomposition of the first component was inside the particle, and the second component was on the surface of the nanosized particle. Hydroconversion was carried out in a bubble flow reactor at T = 430 ∘ C, P = 7.0 MPa, H2 /feedstock = 550 nl/l, and v = 2.2 h−1 ; the sum of metals per tar was 0.05 wt%. Toluene-insoluble particles (TIP) were isolated from the hydrogenation product. According to XRD data, TIP contained graphitized compaction products, MoS2 , Co9 S8 , Ni3 S2 , Fe1–x S. Comparison of the hydroconversion results in the presence of a Mo, Al-containing catalyst obtained according to schemes with simultaneous and sequential introduction of components (Al/Mo = 0.14 wt) revealed that coke formation in the experiment with a sequential injection scheme was significantly lower (0.4%), and the fraction conversion 500 ∘ C+ was higher than for the simultaneous injection scheme. Kadiev et al. (2018b) carried out tar hydroconversion in the presence of iron-containing catalysts synthesized in situ. Water-soluble iron compounds (FeSO4 , Fe(NO3 )3 , Fe(COOCH3 )2 , 0.1–0.2% Fe

505

506

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

per tar), introduced with the aqueous phase of the inverse emulsion in the raw material, were used as precursors. The experiments were carried out on a flow-continuous hydroconversion unit with a vertical bubbling reactor. Under hydroconversion conditions (T = 445 ∘ S, R = 7 MPa), pyrrhotine Fe1-x S particles with average sizes from 50 to 100 nm were formed from iron sulfate and acetate, and hematite Fe2 O3 particles were formed from iron nitrate. Comparison of the experimental results performed with the use of Fe precursors with those in which the APM precursor was used showed pyrrhotine to inhibit the coke formation significantly worse than MoS2 . A suspension of nanosized hematite particles accelerated the cracking reactions and the formation of compaction products. There was studied the hydroconversion of the vacuum distillation residue of oil at a hydrogen pressure of 7 MPa, a temperature of 445 ∘ C, and a feed space velocity of 2 h−1 in the presence of suspensions of nanosized catalyst particles obtained in situ in the reaction zone from the reverse emulsions of aqueous solutions of precursors, the salts of molybdenum, nickel, cobalt and aluminum, as well as mixtures of these salts (Zekel′ et al. 2020). Based on the data of elemental analysis and 1 H NMR spectra, the structural parameters of residues from the vacuum distillation of the hydrogenation product were determined. The relationship between the structural and chemical parameters of residues, feedstock conversion, coke yield, and catalyst composition were established. The tar conversion and the structural parameter of vacuum residue unsaturation increased in the order of catalysts: Mo–Al, Ni < Mo–Co–Al < Mo < Mo–Co < Mo–Ni, Al. The coke yield increased in the same order. The structure and properties of molybdenum-containing catalyst nanoparticles formed from globules of APM aqueous solution emulsions under hydroconversion conditions were studied in (Khadzhiev et al. 2015). An aqueous solution of the APM precursor was emulsified (0.05% Mo and 2% water for feed) in the HOF samples (vacuum residues and natural bitumen). The prepared emulsion was subjected to hydroconversion in the vertical bubble-type flow reactor at 430–450 ∘ C, hydrogen pressure of 7.0 MPa, and space velocity of 2.2 h−1 . The resulting catalyst appeared highly active in the coking inhibition reactions. Depending on the hydroconversion conditions, the yield of solid particles ranged from 0% to 1.1%. It was shown that MoS2 synthesized in situ in a hydroconversion reactor had a layered structure. Individual slabs were 0.6–0.7 nm in thickness and have an average length of 9.8 nm. It was established that monolayer particles had an average size of 9.8 nm and were aggregated into spheres with the formation of multilayer particles and adsorbed on the surface of coke particles. The size of solid particle accumulations in hydrogenates was 300–400 nm. The thermodynamic analysis of the model precursor sulfidation reaction and the obtained experimental data showed that the complete molybdenum binding to sulfur occurred only under conditions of an excess of active sulfur (H2 S) with respect to molybdenum (Khadzhiev et al. 2017). According to the results of thermodynamic analysis, the formation of molybdenum sulfide in the MoO3 –S(H2 S)–H2 –C system can occur at temperatures above 350 ∘ C. At temperature of 420 ∘ C and hydrogen pressure of 7 MPa, a part of molybdenum can be present in the form of MoO2 even under the excess of hydrogen sulfide. To study the effect of HOF composition on the formation of ultrafine catalyst suspensions from an aqueous solution of an emulsion of the APM precursor, experiments were carried out on the hydroconversion of mixtures, including various amounts of tar, heavy catalytic cracking gas oil, coking gas oil, and solid petroleum paraffin (Kadiev et al. 2020a). The model mixtures differed in the elemental and group compositions. The experiments were carried out in an autoclave at hydrogen pressure of 7.0 MPa, temperature of 440 ∘ C, duration of 1 h, and in a bubbling flow reactor at 440 and 450 ∘ C and space velocities of 0.64–2.07 h−1 . The catalyst used in the experiments was synthesized in situ from an inverse emulsion of APM (0.05–0.1% Mo, 2% H2 O). From the products of hydroconversion by extraction with toluene, insoluble particles (ISP) were isolated, in which

11.5 Methods of Synthesis and Properties of Nanoscale Catalysts Used in Slurry Processes of HOF Hydroconversion

Figure 11.9 The plot of SMo : Mo ratio in catalyst particles vs. partial pressure of hydrogen sulfide in the reactor.

SMo : Mo (M)

the dispersed composition, elemental composition, content of sulfur associated with molybdenum (SMo ), and organic sulfur (Sorg ) were determined. The X-ray diffraction patterns of solid particles isolated from the hydrogenation products showed reflections corresponding to graphite (densification products), MoS2 , MoO2 , MoO3 . The ratio of reflection intensities depended on the feed composition. It was established that with increase in the partial pressure of hydrogen sulfide in the reaction gas, the degree of catalyst sulfidation increased (Figure 11.9). In turn, an increase in the portion of MoS2 in the catalyst composition caused a noticeable decrease in the coke yield in the experiment. The rate constants and activation energy for the reaction of sulfur removal from tar were calculated from the results of experiments on a flow-continuous set-up. At 440 and 450 ∘ C, the rate constants were 0.0263 and 0.0395 min−1 , respectively, and the activation energy was 178.9 kJ/mol. The hydroconversion of heavy oil was carried out in the presence of a suspension of MoS2 nanoparticles synthesized in situ from an emulsion of an aqueous solution of the APM precursor (Khadzhiev et al. 2018). Hydroconversion was carried out in a vertical flow bubbling reactor at hydrogen pressure of 7 MPa, temperature of 440–460 ∘ C, a feed space velocity of 1–3 h−1 , and a hydrogen/feedstock ratio of 1000 nl/l. The catalyst precursor emulsion in tar contained 0.05% molybdenum and 2% water in the feed. According to the research data, solid particles isolated from hydrogenation products had an average size of about 400 nm and consisted of graphitized compaction products and spherical MoS2 particles with a diameter of 30–70 nm (Figure 11.10). With the hydroconversion temperature increased and with the space velocity decreased, the yield of distillate fractions increased, reaching a value of 83.8% at 450 ∘ C. In the temperature range of 440–450 ∘ C, the coke yield did not exceed 0.02%. Analysis of the group hydrocarbon composition of the original oil and hydrogenate showed 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0

20 HM

Figure 11.10

0.05 0.10 0.15 0.20 0.25 Partial pressure of H2S (MPa)

10 nm TEM images of MoS2 particles isolated from heavy oil hydrogenates.

0.30

507

508

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

Table 11.2 Kinetic characteristics for the hydroconversion of 520 ∘ C+ fraction of heavy oil and its components in the presence of a suspension of MoS2 nanoparticles synthesized in situ.

Parameter

Heavy oil fraction +520 ∘ C

Paraffin-naphtene

T, ∘ C

440

450

440

450

440

450

440

450

440

450

0.0107

0.0173

0.011

0.0167

0.0125

0.0196

0.0095

0.0163

0.0062

0.0110

K, min−1 Ea , kJ/mol

216.8

178.9

Aromatic

192.8

Resins

231.4

Asphaltenes

247.0

that the content of paraffinic and naphthenic hydrocarbons increased, and the content of aromatic hydrocarbons, resins, and asphaltenes decreased at the hydroconversion process. The greatest decrease was observed in the content of resins. The data obtained in the work made it possible to determine the kinetic characteristics of the hydroconversion of heavy oil and components of its group composition (Table 11.2).

11.6 Principles of Sulfidation of Dispersed Molybdenum-Containing Catalysts Sulfidation of molybdenum-dispersed catalyst occurs in a reducing medium in situ due to hydrogen sulfide that resulted from decomposition of sulfur-containing components of the feed. This sulfidation method is often used in HOF hydroconversion processes using both oil- and water-soluble precursors. However, the mechanism of the process and its effectiveness have not been sufficiently studied. The formation of the sulfide phase of the catalyst can occur both as a result of the interaction of the products, resulting from precursor thermal decomposition, with sulfur-containing organic components of the feedstock, and as a result of interaction with the components of the gas phase, hydrogen, and hydrogen sulfide, formed at the feedstock thermal destruction. Oil contains sulfur-containing compounds of various classes: thiols, sulfides, thiophenes, thiophenes, and its derivatives. With increase in the boiling point of oil fractions, the portion of thiols and sulfides in them decreases and the content of organic compounds, including the thiophene group, increases. In high-boiling oil fractions, sulfur is included in the composition of dibenzothiophene derivatives: naphthothiophenes, naphthobenzothiophenes, etc. (Al-Rabiah et al. 2020). The sulfur atom in the molecules of dibenzothiophene derivatives containing alkyl groups is screened by hydrocarbon fragments of the molecules, thus significantly reducing the probability of both its interaction with the catalyst active site and its subsequent sulfidation (Bataille et al. 2000). Sulfidation of suspensions of molybdenum-containing nanosized particles resulted from precursor decomposition most likely occurs as a result of interaction with hydrogen sulfide formed at the hydrocracking of HOF. The mechanism of sulfidation for suspensions of nanosized catalyst particles synthesized from emulsions of APM aqueous solution with HOF components and sulfidizing additives was considered in the study by Kadiev et al. (2010) and Zekel et al. (2021). The study of sulfidation for the molybdenum catalyst synthesized from water-soluble APM precursor was carried out at tar hydroconversion in a unit with flow reactor at a space velocity of 0.4–1.2 h−1 , and a volume ratio of H2 (nl) : tar (l) equal to 500 (Zekel et al. 2021). The APM content (in terms of Mo) and water in the raw emulsion was 0.15% and 2%, respectively. A toluene-insoluble solid product (TIP) was isolated from the hydrogenate by extraction with toluene. According to the data of chemical and X-ray phase analysis, this product included molybdenum compounds

11.7 Formation of Coke-like Polycondensation Products and their Effect on the Structure and Catalytic Activity

(MoS2 , MoO2 ) and coke. According to a special technique, the forms of sulfur were determined in HOF particles: organic sulfur (Sorg ), which was part of the coke, and inorganic sulfur, appearing as sulfide, associated with molybdenum (SMo ). The temperature for beginning of tar thermal destruction accompanied by the formation of H2 S is 360 ∘ C, according to the data of thermo-gravimetric analysis. Based on these data, hydroconversion experiments were carried out in the range of 380–440 ∘ S. At temperatures of 380–400 ∘ C, increase in the concentration of H2 S in the gas led to increase in the SMo : Mo ratio. However, even at a reaction time of 150 min and a temperature of 400 ∘ C, the SMo : Mo atomic ratio does not exceed 1. The ln(SMo : Mo) – reaction time dependences at 380 and 400 ∘ C were linear, indicating the first-order reaction of dispersed molybdenum catalyst sulfidation with feed sulfur. The rate constants of sulfidation of dispersed molybdenum oxides with tar sulfur (through the stage of H2 S formation) at 380 and 400 ∘ C were 0.0044 and 0.0062 min−1 , respectively, and the activation energy of the process, calculated from the kinetic data, was 62.6 kJ/mol. Based on the data obtained, it takes more than four hours to achieve the atomic ratio SMo : Mo = 2 in the catalyst composition. As the temperature rises above 400 ∘ C, the content of hydrogen sulfide in the gas increases, but the fraction of MoS2 in the catalyst does not noticeably increase. Due to the increase in the coke amount in the TIP with an increase in temperature above 400 ∘ C, a decrease in the sulfidation rate constant takes place. Even at temperature of 440 ∘ C (1.16 vol% H2 S in gas) the SMo : Mo ratio in the catalyst particles does not exceed 1. Apparently, at high hydroconversion temperatures, the catalyst particles are blocked by coke formed, and the sulfidation rate decreases. To improve the sulfidation of a dispersed catalyst, it was proposed to use sulfidizing additives that form hydrogen sulfide at temperatures below that for the beginning of tar thermal decomposition, i.e. in the absence of coke formation blocking catalyst particles. The addition of sulfidizing reagents, such as elemental sulfur and thiocarbamide (at the rate of 0.2% sulfur per tar), made it possible to increase the SMo : Mo ratio in catalyst particles to 1.71, that corresponded to 85% molybdenum content in the form of MoS2 (Zekel et al. 2021).

11.7 Formation of Coke-like Polycondensation Products and their Effect on the Structure and Catalytic Activity of MoS2 Suspensions Dispersed MoS2 is highly active in the hydrogenation of radicals, formed at HOF thermal destruction. Nevertheless, when using this catalyst, compaction products are formed, including coke, the yield of which can be 0.2–5% by weight per feed. The presence of coke in catalyst particles affects the composition, structure, and catalytic activity of MoS2 . Kadiev et al. (2021) studied the composition, structural-morphological features, and particle size of the active phase of the catalyst (MoS2 ) synthesized in situ at heavy oil hydroconversion. The experiments were performed under severe conditions in a flow reactor at T = 450 ∘ C, P = 6.0–9.0 MPa, V = 1.0 h−1 , H2 /feedstock = 1000 nl/l, and catalyst concentration of 0.01–0.08 wt% in terms of molybdenum. It was shown that increase in the catalyst content and hydrogen pressure led to decrease in the yield of polycondensation products. Based on the results of studying the morphology and structure of catalyst particles, it was established that the main effect on the reduction in the coke yield in these experiments was due to an increase in the dispersion of MoS2 clusters. In the hydroconversion process, the components of the tar group composition are converted at different rates. As shown in Kadiev et al. (2010), Kadiev et al. (2018a), Khadzhiev et al. (2018), Al-Rabiah et al. (2020), Bataille et al. (2000), Kadiev et al. (2021), Zekel et al. (2021), the hydroconversion rate constants in the presence of Mo-containing catalyst synthesized in situ increase in the

509

510

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

order: asphaltenes < resins < paraffin–naphthenic hydrocarbons < aromatic hydrocarbons. Under the hydroconversion process, the content of asphaltenes increases due to the difference in hydroconversion rates in the reactor liquid phase. With the limiting concentration reached, asphaltenes aggregate with the capture of catalyst particles, accompanied by coke formation. The change in asphaltene concentration at hydroconversion is described by the equation (Kadiev et al. 2018a): lnCl.asph = B + τ(kfr.500+ − kasph )

(11.4)

where Cl asph is the content of asphaltenes in the reactor liquid phase; τ is the reaction time, kfr.500+ and kasph are the rate constants for the conversion of the 500 ∘ C+ fraction and asphaltenes, respectively. With kfr.500+ > kasph , the asphaltene concentration will increase to a value limited by the asphaltene solubility in the reactor liquid phase under hydroconversion conditions. In this case, aggregation and formation of coke will occur. Increasing both the pressure and catalyst concentration in the reaction volume increases the rate of asphaltene hydroconversion and reduces coke formation. A relationship was established between the HOF solubility parameter of and the coke yield at HOF hydroconversion of various compositions (Kadiev et al. 2018a): Scoke = 0.0408 [Sasph (𝛿asph − 𝛿media )]–0.0772, wt%

(11.5)

where Ccoke is the yield of coke, wt%; Casph is the content of asphaltenes in feedstock, wt%; 𝛿 asph and 𝛿 media are Hildebrand solubility parameters for asphaltenes and HOF, MPa0.5 , respectively.

11.8 Synthesis Conditions and Catalytic Activity of Catalysts Synthesized Ex Situ As seen from the above material, the synthesis of nanosized catalyst particles from water-soluble precursors directly in the feed (in situ) has certain limitations. In order to obtain stable inversed emulsions in the feed, it is necessary to have sufficiently high contents of resins and, in particular, of asphaltenes. To form the active (sulfide) form of the catalyst, the feed must contain components generating hydrogen sulfide at heating. However, as shown above (Zekel et al. 2021), the sulfidation of dispersed molybdenum catalyst synthesized in situ with hydrogen sulfide formed from the feedstock components is not sufficiently effective. Therefore, it was proposed in (Kadiev et al. 2017; Kadieva et al. 2019a) to synthesize concentrated suspensions of nanosized particles ex situ. Mo-containing catalysts were synthesized ex situ using aqueous solutions of APM and additives of sulfiding reagents. An aqueous solution containing the precursor and sulfiding additives was emulsified in tar (Kadiev et al. 2017).The heat treatment of the emulsion was carried out in two stages: the first stage at T = 150 ∘ C and the second stage at T = 350 ∘ C. Duration of treatment at each stage, the composition of the sulfiding agent, and the pH of the aqueous medium were varied at the synthesis. Synthesis products, the catalyst suspensions in tar, appeared as black viscous compositions after removal of conversion by-products (water, ammonia, hydrogen sulfide) from the reaction medium. Toluene-insoluble components (TIC) were isolated from the suspensions obtained ex situ and studied by X-ray diffraction analysis, flame atomic absorption, IR, and Raman spectroscopy. Depending on the synthesis conditions, HOF particles had sizes from 183 to 600 nm and contained from 13% to 50% Mo. The particle size was influenced by coke inclusions; it has had an inverse dependence on the molybdenum content. A catalyst sample free of MoO2 impurities was obtained using elemental sulfur as sulfidizing agent, taken in the ratio S : Mo = 17. According to

11.8 Synthesis Conditions and Catalytic Activity of Catalysts Synthesized Ex Situ

50 nm Figure 11.11

10 HM Ex situ synthesized catalyst particles.

the TEM data, it appeared as particles with an average size of 209 nm, consisting of graphitized compaction products with inclusions of spherical MoS2 particles (Figure 11.11). Experiments on hydroconversion in the presence of sulfided molybdenum-containing catalysts obtained ex situ were carried out on a unit with a flow reactor at constant parameters: temperature of 410 ∘ C, feed space velocity of 2.0–2.2 h−1 , hydrogen pressure of 7.0 MPa, hydrogen to feed ratio up to 1200 nl/l; a mixture of Urals tar (94 wt%) with heavy catalytic cracking gas oil (6 wt%) was used as a raw material. The catalyst loading with regards to the raw material was 0.05 wt (based on molybdenum). The catalyst activity was evaluated by the feed conversion per pass and by the mass of coke on the apparatus walls. Several catalysts synthesized ex situ were tested. In all the experiments, the conversion per pass was 43–44%, and no coke deposits were found (Kadiev et al. 2017). The influence of various sulfidizing additives on the catalytic properties of the Mo catalyst suspension synthesized ex situ was studied by Kadieva et al. (2019a). APM solutions were emulsified in tar with the addition of one of the sulfidizing additives, elemental sulfur (Cat (S)) or thiocarbamide, CS(NH2 )2 (Cat (TC)). The emulsions were subjected to thermal treatment in a hydrogen atmosphere to form the catalyst sulfide form and to remove water, ammonia, and excess sulfur. Elemental sulfur was introduced in the form of a powder; thiocarbamide was dissolved in a (NH4 )6 Mo7 O24 solution at molar ratio S/Mo = 3. Thermal treatment of the emulsions was carried out in a hydrogen atmosphere in two stages at 150 and 350 ∘ C with the exposure at each stage for five hours. The synthesized suspensions of Cat (S) and Cat (TC) catalysts contained particle accumulations of 209 and 223 nm in average size, respectively. In the Cat (S) and Cat (TC) catalyst particles isolated from the suspensions by extraction with toluene, the content of molybdenum was 38.6% and 46.6%, the S/Mo molar ratios were 1.9 and 2.1, respectively. According to TEM and HAADF STEM data, spherical MoS2 particles with average diameters of 80 nm (Cat (S)) and 120 nm (Cat (TC)) were included in the carbon matrix. In Cat (S) particles, oxygen entered the inner part of MoS2 clusters, with the outer layers of particles consisting of MoS2 (Figure 11.12). Cat (TC) particles had significantly lower oxygen content, the latter was uniformly distributed in the particle volume. The crystallite sizes of MoS2 Cat (S) and Cat (TC) were 38 Å and 17 Å, and the average numbers of layers were 6 and 3, respectively. Figure 11.13 shows X-ray diffraction patterns of the obtained TI samples. As it follows from XRD data, the samples studied do not contain bulk crystalline phases and have an amorphous-crystalline character. The X-ray diffraction patterns show peaks at 2𝜃 ≈ 14∘ , 33 ∘ –35 ∘ , 39 ∘ , and 59 ∘ , which are characteristic of various crystallographic modifications of molybdenum sulfide (MoS2 (2H, 3R), Mo3 S4 ). Comparison of interplanar distances (d002 ) calculated from the most intense reflection 2𝜃 ≈ 14 ∘ using the Wulf–Bragg equation and reference values of d002 for MoS2 (2H) (JCPDS

511

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

Mo

C

S

300 nm

O

300 nm

300 nm

300 nm

Figure 11.12 HAADF STEM images and corresponding elemental distribution maps for Mo, S, O, and C in Cat (S) – TI sample.

60 Cat (S) - TI Cat (TC) - TI MaS2-lines

Intensity (%)

512

40

20

0 10

Figure 11.13

20

30

40 2θ (°)

50

60

70

XRD spectra of Cat (S)-TI and Cat (TC)-TI samples.

# 01-075-1539), MoS2 (3R) (JCPDS # 01-089-2905), Mo3 S4 (JCPDS # 01-082-1709) suggests the presence of two forms of sulfides, MoS2 (2H) and Mo3 S4 , in the TI samples. The broadening of the peaks in the X-ray diffraction patterns indicates a weakly crystalline structure and a small size of the crystalline phases of molybdenum sulfide. Testing of the catalytic activity of the catalysts in the process of tar hydroconversion was carried out in the flow mode, at P = 70 atm., T = 425 ∘ C, H2 /feedstock = 1500 nl/l, feedstock space velocity of 0.4 and 1.4 h−1 , catalyst content in raw materials corresponded to 0.05% Mo. Along with Cat (S) and Cat (TC) catalysts, tar hydroconversion was carried out under the same conditions in the presence of a catalyst synthesized in situ. The results for Cat (S) and that synthesized in situ were similar (conversion of fr. 500 ∘ S+ were 55% and 48%, and the coke yields were 0.2% and 0.1%, respectively). With Cat (TC), the coke yield was 3.4%. The difference in activity between Cat (S) and Cat (TC) was attributed to the smaller particle size of the Cat (S) catalyst and the presence of single-layer MoS2 with a large number of edge regions according to the rim-edge catalytic activity model. Studying of the activity of suspensions of disperse catalysts synthesized “ex situ” from water-soluble precursors in hydrocracking reactions of a mixture of petroleum paraffin and catalytic cracking heavy gas oil (1 : 1) was carried out in (Maksimov et al. 2019). Heat treatment in the presence of a sulfidizing agent and hydrogen was used to obtain concentrated suspensions of metal sulfide nanoparticles from reversed emulsions of aqueous solutions of salts of Mo, Ni, Fe, and W stabilized in tar “ex situ.” The average diameters of the catalyst particles were 273–364 nm. Hydroconversion was carried out in an autoclave at hydrogen constant flow, temperature of 445 ∘ C, pressure of 7 MPa, for two hours, in the presence of 0.1% metals on the feed.

11.8 Synthesis Conditions and Catalytic Activity of Catalysts Synthesized Ex Situ

The studied catalysts affected the hydrocracking process in different ways. In particular, MoS2 , (MoS2 + Ni7 S6 ) and Ni7 S6 inhibited the processes of coke formation, as evidenced by the decrease in the yields of compaction products insoluble in toluene. On the contrary, suspensions of Fe1–x S and (NH4 )0.25 WO3 nanosized particles catalyzed cracking processes, that appeared to increase the coke formation and gas yield, as compared with the experiment without catalyst. The content of paraffinic and naphthenic hydrocarbons in the hydrogenate fractions decreased, while the content of sulfur, unsaturated hydrocarbons, vice versa, increased in the order: MoS2 < (MoS2 + Ni7 S6 ) < Ni7 S6 < Fe1–x S < (NH4 )0.25 WO3 . Mo, Ni, Fe, and W catalyst suspensions synthesized by Kadiev et al. (2019b) were tested in the experiments on the hydroconversion of heavy catalytic cracking gas oil with a high content of aromatic hydrocarbons. Hydroconversion was carried out in an autoclave at 425 ∘ C, constant hydrogen pressure of 7 MPa, and catalyst content in the reaction zone of 0.1% on the metal basis. The research results showed that Mo- and Ni-containing catalysts were more active in hydrogenation and hydrodesulfurization reactions. The catalytic activity, determined by the increase in the ratio of paraffin-naphthenic to aromatic hydrocarbons and the ratio of H/C in the hydrogenate fractions, decreased in the following order: MoS2 < (MoS2 + Ni7 S6 ) < Ni7 S6 < Fe1–x S < (NH4 )0.25 WO3 . The catalytic properties of synthesized ex situ nanodispersed suspensions based on MoS2 , (MoS2 + Ni7 S6 ), Ni7 S6 , Fe1–x S, and (NH4 )0.25 WO3 (the synthesis method and properties of the catalysts described by Maksimov et al. (2019)) were studied in hydroconversion of vacuum residue of West Siberian oil mixture (Kadiev et al. 2019a). Hydroconversion was carried out in an autoclave in the presence of synthesized catalysts (0.1% of the metal or the sum of metals per tar), at 425 ∘ C, hydrogen pressure of 7 MPa, with a constant supply of hydrogen, for two hours. Kinetic parameters of the process in the presence of synthesized “ex situ” suspensions of MoS2 (0.05% Mo per tar) were determined during the hydroconversion of tar in a flow-continuous unit with a vertical bubbling-type reactor at 430–440 ∘ C, hydrogen pressure of 7 MPa, space velocity of 0.68–3.33 h−1 . The activity of the catalysts used in the reactions of cracking and hydrogenation of the tar components was evaluated by the gas and coke yields, change in the group composition of the hydrogenate fractions, and by the content of sulfur and olefins in the hydrogenate fractions. The results were compared with those obtained without a catalyst. It was established that MoS2 and Ni7 S6 were active in the reactions of hydrogenation of tar components and prevention of the compaction products formation, whereas Fe1–x S and (NH4 )0.25 WO3 catalyzed cracking reactions. These conclusions are consistent with those of the study for the catalytic activity of disperse catalysts synthesized ex situ for the hydroconversion of catalytic cracking gas oil and solid petroleum paraffin (Maksimov et al. 2019; Kadieva et al. 2019b). The content of organic matter in solid particles, isolated from hydrogenates, increased and the H/C ratio decreased in the following order: MoS2 + Ni7 S6 < Ni7 S6 < MoS2 < (NH4 )0.25 WO3 < Fe1–x S. The average particle size and the amount of vanadium and nickel in the separated solid phase increased in the same sequence. Kinetic parameters for the following reaction were determined in the experiment: Feed (fraction 500 ∘ S+) → Products

(11.6)

It was established that reaction corresponded to the first order. The rate constants at 430 and 440 ∘ C were 0.8055 and 1.3582 h−1 , respectively, and the activation energy was 217.7 kJ/mol. The method proposed by Kadiev et al. (2016) was used to calculate the enthalpy of the hydroconversion process. The method is based on estimating the change in the enthalpy (ΔN ) of the components of the feedstock and reaction products, with the enthalpies of molecule formation presented as linear functions versus number of g-atoms of C, H, N, O, and S elements. The results

513

514

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

of computational studies showed that the proposed express-method allows to determine with satisfactory accuracy the total thermal effect of hydroconversion reactions, on the basis of experimental data on the elemental composition of raw materials and products. The enthalpy of tar hydroconversion calculated according to this method in a flow-continuous unit in the presence of MoS2 suspension synthesized “ex situ” was 490 kJ/kg.

11.9 Behavior of Vanadium and Nickel at HOF Hydroconversion Using Suspensions of Nanosized Catalysts The high concentration of metals, in particular vanadium and nickel, is a serious problem encountered in the processing of HOFs. Metal deposits can negatively affect the catalyst activity and cause their irreversible deactivation (Magomedov et al. 2015). The results of studying the behavior of vanadium and nickel in HOF hydroconversion in the presence of ex situ synthesized suspensions of nanosized catalyst particles are given in the study by Kadiev et al. (2020b). The distribution of vanadium and nickel at tar hydroconversion in the presence of preliminarily synthesized suspensions of nanosized particles of MoS2 , Ni7 S6 , (NH4 )0.25 WO3 , and Fe1−x S has been studied. The experiments were carried out in an autoclave reactor and in a flow-continuous hydroconversion unit at temperature of 350–440 ∘ C, hydrogen pressure of 7 MPa, and space velocity of 0.7–1.6 h−1 . In accordance with the catalyst activity in cracking reactions, the coke yield increased in the following order: MoS2 < Ni7 S6 < (NH4 )0.25 WO3 < Fe1−x S. The portion of metals in the composition of toluene-insoluble particles increased in the same sequence. The results of the HAADF-STEM study of the structure of HOF particles revealed that vanadium present in HOF was not included in the composition of the catalyst active phase. Taking into account the high thermal stability of vanadium and nickel porphyrinic complexes, it can be assumed that under conditions of minimal formation of compaction products (coke), metals (V, Ni) mainly remain in the liquid phase and are present in the form of porphyrin complexes. The map of vanadium atom distribution does not coincide with the distribution map for molybdenum and sulfur, but coincides with that for coke components: carbon and nitrogen atoms (Figure 11.14). This fact indicates the absence of interaction between MoS2 particles and vanadium from the feedstock.

HAADF

C

Mo

S

V

N

Figure 11.14 HAADF STEM image and corresponding maps for distribution of chemical elements for isolated toluene insoluble particles.

11.10 Kinetic Parameters of Heavy Oil Feedstock Hydroconversion in the Presence of a Suspension

11.10 Kinetic Parameters of Heavy Oil Feedstock Hydroconversion in the Presence of a Suspension of MoS2 Nanoparticles The kinetic parameters of heavy oil feedstock hydroconversion in the presence of MoS2 nanoparticle suspension synthesized in situ from emulsion of the APM aqueous solution were considered in the study by Maximov et al. (2020). The experiments were performed using samples of various HOF (vacuum residues, heavy oil, and natural bitumen) in a bubble flow reactor at 425–460 ∘ C, 7 MPa, space velocity of 0.7–3.3 h−1 , H2 /feedstock ratio = 1000 nl/l, in the presence of 0.05% Mo. The kinetic model was considered for two variants of the process: without noticeable coke formation (coke yield was less than 0.5%) and with coke formation. In the first variant, the conversion of residual fractions 520 ∘ S+ into hydroconversion products, including gas and the distillate fraction with the initial boiling point (IBP) of 520 ∘ S, can be described by the equation: Fraction 520 ∘ S+ → (gas + distillate fractions)

(11.7)

The second option describes feed conversion into separate components: gas, gasoline, diesel, oil fractions, and polycondensation products (coke). The used 12-parameter kinetic scheme with partial consideration of secondary reactions is shown in Figure 11.15. The dependencies ln[520 ∘ S+] = f (𝜏) calculated for the first kinetic model for all the studied samples of raw materials were linear (R2 > 0.98), thus indicating the first order of the hydroconversion reaction. The reaction rate constants at 440 and 450 ∘ C were in the range of 0.0048–0.0295 min−1 and 0.0082–0.041 min−1 , respectively, and the activation energies were in the range of 137–260 kJ/mol. Experimental results revealed that the hydroconversion activation energy increased with the increase in the asphaltenes content in the feedstock. This pattern is consistent with the previously obtained results of heavy oil hydroconversion in the presence of dispersed MoS2 , according to which the activation energy increased in the order: aliphatic hydrocarbons < aromatic hydrocarbons < resins < asphaltenes (Khadzhiev et al. 2018). The dependence of ln(k) on 1/T plotted from the experimental results for one of the HOF samples appeared linear. High activation energies and the slope of the curve indicated the reaction to proceed in the kinetic region. The calculation of the kinetic parameters of model No. 2 showed that the formation of polycondensation products (coke) under hydroconversion conditions led to G6 coke k12 k1

G1 feed k2

G2 Fr. 350–520 °C

k5 – k6 k10

Figure 11.15

G3 Fr. 180–350 °C

k4

k3 k7 – k8

G4 IBP–180 °C

k11

Kinetic model No. 2 of HOF hydroconversion.

k9

G5 gas

515

516

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

increase in the process rate and decrease in the activation energy. The obtained values of the kinetic indicators of hydroconversion appeared like those of HOF thermal cracking. The similarity of the kinetic parameters for hydroconversion and HOF thermal cracking can be explained by the catalyst low content in the reaction volume. This result indicates that the main amount of the feedstock is converted by the thermal cracking mechanism.

11.11 Conclusion The material presented in the review allows us to conclude that the hydroconversion processes in the catalyst suspension phase are characterized by the absence of by-products of low liquidity (coke, asphalt) and allows the maximum use of the petroleum feedstock potential. Further progress in the development of technology for upgrading heavy oils and bitumens is closely related to the search for new active catalytic systems.

References Ahmed, H.S., Shaban, S.A., Menoufy, M.F., and El Kady, F.Y. (2013). Effect of catalyst deactivation on vacuum residue hydrocracking. Egyptian Journal of Petroleum 22: 367–372. Akbarzadeh, K., Ayatollahi, S., Moshfeghian, M. et al. (2004). Estimation of SARA fraction properties with the SRK EOS. Journal of Canadian Petroleum Technology 43 (9): – A. PETSOC-04-09-02. Akbarzadeh, K., Alboudwarej, H., Svrcek, W.Y., and Yarranton, H.W. (2005). A generalized regular solution model for asphaltene precipitation from n-alkane diluted heavy oils and bitumens. Fluid Phase Equilibria 232: 159–170. Al-Rabiah, H., Al-Ostad, L., Al-Shamary, T. et al. (2020). Sulfur speciation of Kuwaiti crude oils fractions using two dimensional gas chromatography. Arabian Journal of Chemistry 13 (2): 3615–3623. Alshareef, A.H. (2020). The asphaltenes: definition, properties, and reactions of model compounds. Energy & Fuels 34 (1): 16–30. Ancheyta, J. and Speight, J.G. (2007). Hydroprocessing of Heavy Oils and Residua. Boca Raton: CRC Press Taylor and Francis Group. Ancheyta, J., Trejo, F., and Rana, M.S. (2009). Asphaltenes. Chemical Transformation During Hydroprocessing of Heavy Oils. Boca Raton: CRC Press. Taylor & Francis Group. Bataille, F., Lemberton, J.-L., Michaud, P. et al. (2000). Alkyldibenzothiophenes hydrodesulfurization-promoter effect, reactivity, and reaction mechanism. Journal of Catalysis 191 (2): 409–422. Bellussi, G., Rispoli, G., Landoni, A. et al. (2013a). Hydroconversion of heavy residues in slurry reactors: developments and perspectives. Journal of Catalysis 308: 189–200. Bellussi, G., Rispoli, G., Molinari, D. et al. (2013b). The role of MoS2 nano-slabs in the protection of solid cracking catalysts for the total conversion of heavy oils to good quality distillates. Catalysis Science & Technology 3: 176–182. Biswas, G., Gillis, D., Hurny, R. (2000) LC-MAX and LC-SLURRY high conversion residue hydrocracking technologies. https://www.mcdermott.com/getattachment/CLG/News-Resources/ Resources/Hydrocarbon-Engineering-LC-SLURRY-LC-MAX-RH-DG-BS.pdf.aspx?lang=en-US (accessed 07 February 2023). Bruix, A., Füchtbauer, H.G., Tuxen, A.K. et al. (2015). In situ detection of active edge sites in single-layer MoS2 catalysts. ACS Nano 9: 9322–9330.

References

Castaneda, L.C., Munoz, J.A.D., and Ancheyta, J. (2014). Current situation of emerging technologies for upgrading of heavy oils. Catalysis Today 220–222: 248–273. Chen, Z., Zhang, L., Zhao, S. et al. (2015). Molecular structure and association behavior of petroleum asphaltene. Structure and modeling of complex petroleum mixtures. Structure and Bonding 168: 1–38. Chianelli, R.R., Siadati, M.H., De la Rosa, M.P. et al. (2006). Catalytic properties of single layers of transition metal sulfide catalytic materials. Catalysis Reviews 48: 1–41. Del Bianco, A., Panariti, N., Di Carlo, S. et al. (1994). Developments in deep hydroconversion of heavy oil residues with dispersed catalysts. 2. Kinetic aspects of reaction. Energy & Fuels 8 (3): 593–597. Dmitriev, D.E. and Golovko, A.K. (2010). Transformations of resins and asphaltenes during the thermal treatment of heavy oils. Petroleum Chemistry 50: 106–113. Dohler, W., Kretschmar, D.K., Merz, L., and Niemann, K. (1987). Veba-Combi-Cracking. A technology for upgrading of heavy oils and bitumen. American Chemical Society, Division of Petroleum Chemistry 32 (2): 484–489. Fixari, B., Peureux, S., Elmouchnino, J., and Le Perchec, P. (1994). New developments in deep hydroconversion of heavy oil residues with dispersed catalysts. 1. Effect of metals and experimental conditions. Energy & Fuels 8: 588–592. Go, K.S., Lim, S.H., Kim, Y.K. et al. (2018). Characteristics of slurry-phase hydrocracking for vacuum residue with reaction temperature and concentrations of MoS2 dispersed catalysts. Catalysis Today 305: 92–101. Gray, M.R., Tykwinski, R.R., Stryker, J.M., and Tan, X. (2011). Supramolecular assembly model for aggregation of petroleum asphaltenes. Energy & Fuels 25 (7): 3125–3134. Jeon, S.G., Na, J.-G., Ko Ch. H. et al. (2011). Preparation and application of an oil-soluble CoMo bimetallic catalyst for the hydrocracking of oil sands bitumen. Energy & Fuels 25: 4256–4260. Jeong, H.-R. and Lee, Y.-K. (2019). Comparison of unsupported WS2 and MoS2 catalysts for slurry phase hydrocracking of vacuum residue. Applied Catalysis 572: 90–96. Joshi, J.B., Pandit, A.B., Kataria, K.L. et al. (2008). Petroleum residue upgradation via visbreaking: a review. Industrial and Engineering Chemistry Research 47: 8960–8988. Kadiev, Kh.M (2015). Development of design and construction of TATNEFT pilot plant for hydroconversion of heavy residues and natural bitumen. International Scientific and Practical Conference “Development of Deep Processing of Hydrocarbon Feedstock in the Russian Federation.” Kazan. Kadiev Kh.M., Gyul’maliev, A.M., Shpirt M.Ya., and Khadzhiev, S.N. (2010). Thermodynamic and quantum chemical study of the transformations and operation mechanism of molybdenum catalysts under hydrogenation conditions. Petroleum Chemistry 50 (4): 312–318. Kadiev Kh.M., Khadzhiev, S.N., and Kadieva M.Kh. (2013). Synthesis and use of polyfunctional catalyst nanoparticles for hydroconversion of natural bitumen. Petroleum Chemistry 53: 298–308. Kadiev Kh.M., Gyul’maliev, A.M., Kubrin, N.A., and Khadzhiev, S.N. (2016). A rapid method for calculating the enthalpy of hydroconversion of heavy petroleum feedstock. Petroleum Chemistry 56 (8): 697–702. Kadiev Kh.M., Khadzhiev, S.N., Kadieva M.Kh., and Dogova, E.S. (2017). Ex situ synthesis of sulfided molybdenum-containing ultrafine hydroconversion catalysts. Petroleum Chemistry 57 (7): 608–617. Kadiev Kh.M., Zekel, L.A., Kadieva M. Kh., and Khadzhiev, S.N. (2018a). Formation of polycondensation products in heavy oil feedstock hydroconversion in the presence of ultrafine catalyst: physicochemical study. Petroleum Chemistry 58 (7): 519–527. Kadiev Kh.M., Zekel′ , L.A., Gyul’maliev, A.M. et al. (2018b). Hydroconversion of oil vacuum distillation residues in the presence of iron-containing catalysts in situ synthesized from water-soluble precursors. Petroleum Chemistry 58 (14): 1165–1172.

517

518

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

Kadiev Kh.M., Gyul’maliev, A.M., Zekel, L.A. et al. (2019a). Hydroconversion of vacuum residue of a blend of Western Siberian oils in the presence of ex situ synthesized suspensions of nanosized catalysts. Petroleum Chemistry 59: 37–44. Kadiev Kh.M., Kadieva M.Kh., Zekel, L.A. et al. (2019b). The properties of water-in-oil emulsions of aqueous solutions of precursors for nanosized catalysts. Colloid Journal 81 (2): 90–97. Kadiev, Kh., M., Zekel, L.A., Gyul’maliev, A.M. et al. (2019c). Hydroconversion of oil vacuum distillation residues in the presence of ultrafine iron-containing catalysts synthesized from oil-soluble precursors. Petroleum Chemistry 59 (3): 498–503. Kadiev Kh.M., Zekel, L.A., Kadieva M.Kh. et al. (2020a). Effect of hydroconversion conditions on the composition and properties of an ultrafine Mo-containing catalyst formed in situ. Petroleum Chemistry 60 (10): 1154–1163. Kadiev Kh.M., Zekel, L.A., Kadieva M.Kh. et al. (2020b). Behavior of vanadium and nickel in hydroconversion of vacuum tower bottoms over nanosized slurry catalysts. Petroleum Chemistry 60 (9): 1009–1018. Kadiev Kh.M., Maximov, A.L., and Kadieva M.Kh. (2021). The effect of MoS2 active site dispersion on suppression of polycondensation reactions during heavy oil hydroconversion. Catalysts 11: 1–18. Kadieva M.Kh., Khadzhiev, S.N., Kadiev Kh.M., and Yakovenko, T.V. (2011). Synthesis of nanosized particles of components of petroleum refining catalysts in a hydrocarbon medium from inverted microemulsions. Petroleum Chemistry 51 (6). Kadieva M.Kh., Maximov, A.L., and Kadiev Kh.M. (2019a). Ex-situ synthesis and study of nanosized Mo-containing catalyst for petroleum residue hydro-conversion. Catalysts 9: 649. Kadieva, M. Kh., Zekel, L.A., Batov, A.E. et al. (2019b). Effect of various catalyst precursors on hydroconversion of heavy cracked cycle oil. Petroleum Chemistry 59 (12): 1278–1284. Kang, K.H., Kim, G.T., Park, S. et al. (2019). A review on the Mo-precursors for catalytic hydroconversion of heavy oil. Journal of Industrial and Engineering Chemistry 76: 1–16. Kapustin, V., Chernysheva, E., and Khakimov, R. (2021). Comparison of moving-bed catalytic tar hydrocracking processes. Processes 9: 500. Khadzhiev, S.N. (2011). Nanoheterogeneous catalysis: a new sector of nanotechnologies in chemistry and petroleum chemistry (a review). Petroleum Chemistry 51 (1): 1–15. Khadzhiev, S.N. (2016). Nanoheterogeneous catalysis: definition, state, and research prospects (review). Petroleum Chemistry 56 (6): 465–479. Khadzhiev, S.N., Kadiev, K.M., and Kadieva M. Kh. (2013a). Structural and morphological features of the formation of polyfunctional nanocatalysts in a reverse microemulsion medium. Petroleum Chemistry 53 (6): 374–382. Khadzhiev, S.N., Kadiev Kh.M., Yampolskaya, G.P., and Kadieva, M.K. (2013b). Trends in the synthesis of metal oxide nanoparticles through reverse microemulsions in hydrocarbon media. Advances in Colloid and Interface Science 197–198: 132–145. Khadzhiev, S.N., Kadiev Kh.M., and Kadieva, M.K. (2014). Synthesis and properties of nanosized systems as efficient catalysts for hydroconversion of heavy petroleum feedstock. Petroleum Chemistry 54: 323–346. Khadzhiev, S.N., Kadiev Kh.M., Zhigalina, O.M. et al. (2015). Structure and properties of molybdenum sulfide nanoparticles synhesized in situ in the hydroconversion process. Petroleum Chemistry 55 (8): 655–662. Khadzhiev, S.N., Kadiev Kh.M., Gul’maliev, A.M., and Kadieva, M.K. (2017). Properties and structure of nanosized catalyst systems based on molybdenum sulfides. Petroleum Chemistry 57 (14): 1277–1286. Khadzhiev, S.N., Kadiev, K.M., Zekel′ , L.A., and Kadieva M.Kh. (2018). Heavy oil hydroconversion in the presence of ultrafine catalyst. Petroleum Chemistry 58 (7): 535–541.

References

Kim, H. and Curtis, C.W. (1990). Reaction pathways of model coprocessing systems using molybdenum naphthenate and excess sulfur. Energy & Fuels 4: 206–214. Kim, K.-D. and Lee, Y.-K. (2019). Active phase of dispersed MoS2 catalysts for slurry phase hydrocracking of vacuum residue. Journal of Catalysis 369: 111–121. Kunnas, J. and Smith, L. (2011). Improving residue hydrocracking performance. Petroleum Technology Quarterly Q3 16 (4): 50–57. León, A.Y., Guzman, A., Laverde, D. et al. (2017). Thermal cracking and catalytic hydrocracking of a Colombian vacuum residue and its maltenes and asphaltenes fractions in toluene. Energy & Fuels 31 (4): 3868–3877. Lim, S.H., Go, K.S., Nho, N.S., and Lee, J.G. (2018). Effect of reaction temperature and time on the products and asphaltene dispersion stability in slurry-phase hydrocracking of vacuum residue. Fuel 234: 305–311. Magomedov, R.N., Popova, A.Z., Maryutina, T.A. et al. (2015). Current status and prospects of demetallization of heavy petroleum feedstock (review). Petroleum Chemistry 55 (6): 423–443. Magomedova, R.N., Pripakhayloa, A.V., Maryutina, T.A. et al. (2019). Role of solvent deasphalting in the modern oil refining practice and trends in the process development. Russian Journal of Applied Chemistry 92 (12): 1634–1648. Maksimov, A.L., Zekel′ , L.A., Kadieva M.Kh. et al. (2019). Assessment of the activity of dispersed catalysts in hydrocracking reactions of hydrocarbonaceous feedstock. Petroleum Chemistry 59 (9): 968–974. Manek, E. and Haydary, J. (2017). Hydrocracking of vacuum residue with solid and dispersed phase catalyst: modeling of sediment formation and hydrodesulfurization. Fuel Processing Technology 159: 320–327. Maximov, A.L., Kadiev Kh.M., Zekel, L.A. et al. (2020). Particular kinetic patterns of heavy oil feedstock hydroconversion in the presence of dispersed nanosize MoS2 . Pure and Applied Chemistry 92 (7): 1111–1121. McMillen, D.F. and Golden, D.M. (1982). Hydrocarbon bond-dissociation energies. Annual Review of Physical Chemistry 33: 493–532. Meyer, R. F., Attanasi, E. D., and Freeman, P. A. (2007). Heavy oil and natural bitumen. Resources in geological basins of the world. Open File-Report. Reston, Virginia: U.S. Geological Survey. 1084: 1–36. Mullins, O.C., Sheu, E.Y., Hammami, A., and Marshall, A.G. (2007). Heavy Oils and Petroleomics. New York: Springer. Nguyen, T.S., Tayakout-Fayolle, M., Lacroix, M. et al. (2015). Promotion effects with dispersed catalysts for residue slurry hydroconversion. Fuel 160: 50–56. Nguyena, M.T., Nguyena, N.T., Choa, J. et al. (2016). A review on the oil-soluble dispersed catalyst for slurry-phase hydrocracking of heavy oil. Journal of Industrial and Engineering Chemistry 43: 1–12. Nguyena, N.T., Parka, S., Junga, J. et al. (2018). Comparative reactivity between thermal and catalytic hydrocracking of vacuum residue: effect of asphaltenes. Journal of Industrial and Engineering Chemistry 61: 32–38. Oloruntoba, A., Zhang, Y., and Hsu, C.S. (2022). Review of fluid catalytic cracking (FCC) catalyst regeneration intensification technologies. Energies 15 (6): 2061. Panariti, N., Del Bianco, A., Del Piero, G. et al. (2000). Petroleum residue upgrading with dispersed catalysts: Part 2. Effect of operating conditions. Applied Catalysis A: General 204 (2): 215–222. Peng, C., Du, Y., Hu, Y., and Fang, X. (2018). Research and development of hydrocracking catalysts and technologies in China. Frontiers of Chemical Science and Engineering 12: 867–877.

519

520

11 Novel Technologies for Upgrading Heavy and Extra-Heavy Oil

Pevneva, G.S., Voronetskaya, N.G., Sviridenko, N.N., and Golovko, A.K. (2020). Effect of WC/Ni–Cr additive on changes in the composition of an atmospheric residue in the course of cracking. Petroleum Science 17: 499–508. Purón, H., Pinilla, J.L., Berrueco, C. et al. (2013). Hydrocracking of Maya vacuum residue with NiMo catalysts supported on mesoporous alumina and silica–alumina. Energy & Fuels 27 (7): 3952–3960. Rahimi, P., Dettman, H., Nowlan, V., and Del Bianco, A. (1997). Molecular transformation during heavy oil upgrading. Preprints-American Chemical Society. Division of Petroleum Chemistry. National Meeting, San Francisco, CA. Rana, M.S., Sámano, V., Ancheyta, J., and Díaz, J. (2007). A review of recent advances on process technologies for upgrading of heavy oils and residues. Fuel 86: 1216–1231. Rogel, E., Ovalles, C., Pradhan, A. et al. (2013). Sediment formation in residue hydroconversion processes and its correlation to asphaltene behavior. Energy & Fuels 27 (11): 6587–6593. Sahu, R., Song, B.J., Im, J.S. et al. (2015). A review of recent advances in catalytic hydrocracking of heavy residues. Journal of Industrial and Engineering Chemistry 27: 12–24. Sedghi, M. and Goual, L. (2010). Role of resins on asphaltene stability. Energy & Fuels 24: 2275–2280. Silva, R.C., Seidl, P.R., Menezes, S.M.C., and Teixeira, M.A.G. (2004). 1 H and 13 C NMR for determining average molecular parameters of asphaltenes from vacuum residue distillation. Annals of Magnetic Resonance 3 (3): 63–67. Somwanshi, S.B., Somvanshi, S.B., and Kharat, P.B. (2020). Nanocatalyst: a brief review on synthesis to applications. International Web Conference on Advanced Material Science and Nanotechnology (NANOMAT – 2020) 20–21 June 2020, Nandgaon Khandeshwar India. Journal of Physics: Conference Series 1644: 1–7. Speight, J.G. (2014). The Chemistry and Technology of Petroleum. Boca Raton: CRC Press. Taylor & Francis Group. Stratiev, D., Shishkova, I., Dinkov, R. et al. (2014). Current technology solutions in modern refining business. Oil and Gas Chemistry 1: 3–18. Stratiev, D., Nenov, S., Shishkova, I. et al. (2020). Commercial investigation of the ebullated-bed vacuum residue hydrocracking in the conversion range of 55−93%. ACS Omega 5: 33290–33304. Tankov, I., Stratiev, D., Shishkova, I. et al. (2017). Reactivity of heavy oils in catalytic and thermal cracking. Part II: SARA fractions and heavy oils. Oxidation Communications 40 (3): 1191–1208. Thian, T.C. (2008). Effects of catalyst morphology on hydrotreating reactions. Journal of Engineering Science and Technology 3 (2): 117–123. VHTL (2019). A new look at heavy oil. https://fluidoilcorp.com/wp-content/uploads/2019/02/VHTLWhite-Paper-2019.pdf (accessed 07 February 2023). Yu-Ran, L. (2002). Handbook of Bond Dissociation Energies in Organic Compounds. Boca Raton: CRC Press. Zekel, L.A., Gyulmaliev, A.M., Batov, A.E. et al. (2021). Sulfidation of a dispersed molybdenum catalyst with hydrogen sulfide formed from hydroconversion of petroleum feedstock. Petroleum Chemistry 61: 1096–1103. Zekel′ , L.A., Magomadov, E.E., Gyul’maliev, A.M. et al. (2020). Relation of structural and chemical parameters of tar hydroconversion residu to catalytic properties of nanosized catalysts based on Mo, Co, Ni, Al compounds. Russian Journal of Applied Chemistry 93 (3): 449–455. Zhang, S., Liu, D., Deng, W., and Que, G. (2007). A review of slurry-phase hydrocracking heavy oil technology. Energy & Fuels 21: 3057–3062.

521

Index a AAE see average absolute error (AAE) adsorbate concentration distributions 444 adsorbent particle 440 adsorption-chromatographic separation 88, 102 adsorption dynamics 435, 443 AET see atmospheric equivalent temperature (AET) alcohol-benzene resins 73 aliphaticity index 88 alkali/surfactant (AS) 162 alkali/surfactant/polymer (ASP) flooding 162 Al-Khafaji’s method 458 alkyl aromatic compounds 314 alkyl chains elongation 90 alumina nanoparticles 285 ambient mass spectrometry 141 4-amino-4-pyridyl diethoxylate (APE) 252 ammonium meta vanadate (AMV) 285 AMV see ammonium meta vanadate (AMV) anions 251 APCI see atmospheric-pressure chemical ionization (APCI) APPCI see atmospheric pressure photochemical ionization (APPCI) APPI see atmospheric pressure photoionization (APPI) aquathermal cleavage of, carbon-heteroatom bonds aquathermolysis reactions of, dibenzyl sulfide 412–415 metal stearate catalysts and simulation 415–421 in model compounds and thermochemical parameters 389–394

thermodynamic and kinetic parameters cyclohexyl phenyl amine 407–412 of cyclohexyl phenyl ether 401–406 of cyclohexyl phenyl sulfide 395–401 total reaction, mechanism of 394–395 aquathermal pyrolysis mechanism of, asphalthene fractions elucidation mechanisms of, pyrene oxidation 422–427 quantum chemical calculations 421–422 aquathermolysis 165, 168, 191, 192, 196, 240, 310, 311, 384 gasification 469 kinetic models 374–375 kinetic studies and modeling 373–374 mechanism 314 reaction order for 365–366 reaction rate coefficients and activation energies 366–373 aquathermolytic technology 221 aqueous sol-gel process 269 AR see atmospheric residue (AR) ARC see accelerated rate calorimetry (ARC) aromatic compounds 7, 287 aromatic hydrocarbons 91, 199 fraction 92 aromaticity 4, 7, 85 aromaticity coefficient 104, 105, 109 aromaticity index 106, 107, 204 of asphaltenes 82 aromatic sulfonic iron 218 Arrhenius constant 247 Arrhenius equation 336 Ashal’cha oil reservoir 81 Ashal’cha oil resins 80

Catalytic In-Situ Upgrading of Heavy and Extra-Heavy Crude Oils, First Edition. Edited by Mikhail A. Varfolomeev, Chengdong Yuan, and Jorge Ancheyta. © 2023 John Wiley & Sons Ltd. Published 2023 by John Wiley & Sons Ltd.

522

Index

Ashalchinskoe oil 45–48 asphaltene adsorption capacity 285 aggregation 45 condensing degree of 80 fractions 85, 86, 421 molecules 490 asphaltenes-to-aromatics (As/Ar) ratio 30 asphaltenes-to-resins (As/Re) ratio 30 asphaltenes-to-saturates (As/Sa) ratio 30 asphaltic-resin components (ARC) 489 Athabasca bitumen reactions 318, 329 atmospheric equivalent temperature (AET) 14, 17 atmospheric-pressure chemical ionization (APCI) 126–127 atmospheric pressure photochemical ionization (APPCI) 128 atmospheric pressure photoionization (APPI) 128 atmospheric residue (AR) 502 atmospheric solid analysis probe-mass spectrometry (ASAP-MS) 128–129 average absolute error (AAE) 338 2,2’-azobisisobutyronitrile (AIBN) 252

b batch reactor mass balance equation 327 Beal’s method 458 benzene resins 73 benzylic radicals 44 Bergman’s method 458 β-MnO2 nanocatalyst 286 bifunctional catalysts 238 bimetallic nanoparticles 293 binary magnesium-transition metal oxides 268 biomarkers 134 Boca de Jaruko oil 86 Bohr magneton 41 bond cleavage reactions 383 Brazilian coast sedimentary basin 120 Bronsted acid 247 Bronsted-or Lewis-type acid sites 255

c calcium compounds 8 carbene-carboid 102 fraction 86, 88

carbon-heteroatom bonds 80, 386 carbon monoxide (CO) 215 carbon nanotubes (CNT) 271–273 carbon number distribution 120, 206 carbon rejection process 309 carbon-to-hydrogen mass percent ratio (C/H) 7 Carr Purcell Meiboom Gill (CPMG) pulse sequence 71 C7 -asphaltene adsorption affinity 285 capacity 285 catalysts for hydroprocessing hydroconversion conditions, in slurry process 500 morphology of, catalysts 499–500 synthesized ex situ 510–514 and water proportions 443 catalytic aquathermolysis 173 catalytic composition 223 catalytic cracking 311 catalytic mechanism and kinetics kinetic models, for aquathermolysis process for gas prediction 326–329 liquid composition and gas generation 329–337 kinetic models for, in situ and ex situ hydrocracking detailed-lumping kinetic model 323–326 five-lump kinetic models 318–322 four-lump kinetic models 316–318 six-lump kinetic models 322–323 kinetic parameters and model assumptions aquathermolysis kinetic models 341–344 hydrocracking kinetic models 337–341 optimal set of 344–347 reaction mechanism, heavy crude oil upgrading aquathermolysis process, of heavy crude oils 313–314 hydrocracking of, heavy crude oils 311–313 CAtalytic Upgrading PRocess In Situ (THAI-CAPRI) 159–160, 264 cerium(IV) dioxide (CeO2 ) 292 cerium nanoparticles 293 C7+ fraction 455 Chamkalani stability classifier (CSC) 14 chemical ionization (CI) 126 chemical vapor deposition (CVD) 271

Index

chemical vapor synthesis (CVS) 266 C-heteroatoms bonds 210 Chinese lantern 428 structure 416 Chromates-Proton 20M 71 chromatographic separation 115 chromatography-mass spectrometry techniques 115, 128 chromatography methods gas chromatography fingerprinting 116–117 group analysis and simulated distillation 117–120 selective detection 121–122 liquid chromatography 123–125 CI see chemical ionization (CI) classical distillation methods 120 classification factor 325 13 C NMR spectroscopy 207 CO see carbon monoxide (CO) coal-based crude oils 117 coated and uncoated SiO2 nanoparticles 284 cobalt dichloride 243 C=O carbonyl groups 84 coke-like polycondensation, formation of 509–510 coke precursor formation 313 cold heavy oil production with sand (CHOPS) 154 cold production 154 colloidal instability index (CII) 13 commercial silica gel nanoparticles 285 condensation index 83 condensation reactions 206, 314 Conradson carbon residue 9 continuous lumping kinetic model 325–326, 359 continuous reactor 321 continuous wave (CW) mode 41 convection-diffusion parameters 445 conventional simulated distillation (CSD) 120 Co3 O4 nanoparticles 289 copper-based liquid catalyst 254 copper spinel ferrite (CuFe2 O4 ) 279 coprecipitation method 267–269 crude oil 2 crystalline silica gel nanoparticles 285 C–S bond cleavage 289 CSD see conventional simulated distillation (CSD)

CSS see cyclic steam stimulation (CSS) Cu-based catalyst 92 CuO nanoparticles 291 cyclic aromatic systems 65 cyclic solvent injection 162 cyclic steam injection (CSI) 263 cyclic steam stimulation (CSS) 154–155, 168, 309 cyclohexane-tetralin-decalin 107 cyclohexyl carbocation 406 cyclohexyl phenyl compounds 391 cyclohexyl phenyl ether (CPE) 386 cyclohexyl phenyl sulfide (CPS) 385, 386, 391

d Damkohler numbers 436, 444 DBT see dibenzothiophene (DBT) deasphalted oil (DAO) 13 Debye Scherrer formula 273 decarbonylation 314 delocalized π system 44 density functional theory (DFT) method 387 detailed-lumping kinetic model 323 continuous lumping kinetic model 325–326 diamondoids 136 dibenzothiophene (DBT) 6 dibenzyl (DB) 412 dibenzyl sulfide (DBS) 386 differential thermal analysis (DTA) 254 diffusional processes reagents-catalyst 238 direct analysis in real time (DART) 130 ionization 142 dispersed nanoparticulated iron(III) oxide 287 dispersion effect 239 distribution functions 436 double-bond equivalents (DBEs) distribution 124 downhole upgrading process 382 dynamic nuclear polarization (DNP) 65

e electromagnetic field (EMF) 40 electromagnetic (EM) heating 160, 288 electromagnetic interference 121 electron capture detector (ECD) 121 electronic energy of activation 395 electron ionization (EI) 126

523

524

Index

electron-nuclear double resonance (ENDOR) 40 pulsed 54–56 electron-nuclear modulations 49 electron paramagnetic resonance (EPR) 40, 62, 422 pulsed 48–51 resonator 41 spectroscopy basic principles of, petroleum 41–48 HYSCORE spectroscopy 51–53 pulsed ENDOR 54–56 pulsed EPR techniques 48–51 X-band EPR spectra 44 electron spin echo (ESE) 40 electron spin echo envelope modulation (ESEEM) 40 electrospray ionization (ESI) 128 EMF see electromagnetic field (EMF) ENDOR see electron-nuclear double resonance (ENDOR) energy demand 1 energy-dispersive X-ray spectroscopy (EDX) mapping 273 techniques 193 enhanced/effective solvent extraction incorporating EM heating (ESEIEH) 160 enhanced oil recovery (EOR) 40, 59 chemical EOR methods combination flooding 162 polymer flooding 161 solvent injection 162 surfactant flooding 161 gas EOR methods 162–163 hybrid EOR methods hybrid thermal-chemical methods 164 hybrid thermal-NCGs methods 164 hybrid thermal-solvent methods 163–164 microbial EOR methods 163 in situ upgrading 164–165 thermal EOR methods electrical heating methods 160–161 EM heating 160 ISC 156–160 steam injection 154–156 EOR see enhanced-oil-recovery (EOR) EPR see electron paramagnetic resonance (EPR) equation of state (EoS) parameters 453

equilibrium distribution functions 437 ESEEM see electron spin echo envelope modulation (ESEEM) 2-ethylhexanoic acid 243 expanding solvent-SAGD (ES-SAGD) 164

f FCC see fluid catalytic cracking (FCC) Fe-based catalyst 245 ferric oleate-based catalyst 243 FID see free induction decay (FID) field ionization (FI) 127 first-contact miscible (FCM) 163 five-lump kinetic models 318–322 fixed-bed reactor 192 flame-ionization detector 116, 117 flame photometric detector (FPD) 121 flow-continuous hydroconversion 506 fluid catalytic cracking (FCC) 10, 493 fluid model 453 FluidOil VHTL (viscositor heavy-to-light) 493 focused ion beam scanning electron microscope (FIB-SEM) 74 forward combustion 156 forward dry combustion 156 four-dimensional X-ray computed tomography 441 Fourier-transform infrared (FT-IR) spectra 79, 267 Fourier-transform ion cyclotron mass spectrometry (FT-ICR-MS) 131 Fourier-transform ion cyclotron resonance (FT-ICR) analyzer 129 four-lump kinetic models 316–318 four-reaction component-based kinetic model 331 FPD see flame photometric detector (FPD) fractionation method 88 Franco–Famenian carbonate complex 83 free hydrogen atoms 311 free induction decay (FID) 60, 72 free radicals (FR) 40 mechanism 313 scavengers 383 FT orbitrap mass spectrometers (FT-Orbitrap) 132 fuel deposition (FD) 158 full width at half maximum (FWHM) 131

Index

g gas chromatography (GC) 115 gas chromatography-mass spectrometry techniques 132 heavy oils, analysis of combined one-dimensional GC and MS method (GC-MS) 132–136 combined two-dimensional GC and MS (GCxGC-MS) 136–138 gas-liquid-solid phases 312 gasoline 244 Gauss–Newton technique 343 GC see gas chromatography (GC) Gibbs energy 390, 392, 397 Gibbs free energy 398, 409 gravity-drainage process 162 green power sources 382 gyromagnetic ratio 60

h halogenated hydrocarbons 121 Hamaca extra-heavy oil 287 HD see hydrogen donors (HD) HDN see hydrodenitrogenation (HDN) HDS see hydrodesulfurization (HDS) heavy and extra-heavy crude oils 2, 3 chemical properties from Canada 22, 23 carbon residue 8–9 from China 19, 21 elemental analysis 6–7 from Latin American countries 25, 27 metal content 7–8 from Mexico 19, 20 from Middle East countries 24, 26 molecular weight 9 vs. physical properties 27–34 from Russia 20, 22 from USA 24 from Venezuela 23, 25 composition SARA analysis 12–14 TBP distillation 14–19 physical properties API gravity 4 from Canada 22, 23 vs. chemical properties 27–33 from China 19, 21

density/specific gravity 4 from Latin American countries 25, 27 from Mexico 19, 20 from Middle East countries 24, 26 pour point 5 from Russia 20, 22 from USA 24 from Venezuela 23, 25 viscosity 4–5 heavy crude organosulfur compounds 242 heavy hydrocarbons 239 heavy oil feedstock (HOF) 489 catalysts for, hydroprocessing hydroconversion conditions, in slurry process 500 morphology of, catalysts 499–500 composition and properties of 490–491 hydrogenation process, for HOF upgrading 490, 493–494 nanoscale catalysts, in slurry process from oil-soluble precursors 501–503 from water-soluble precursors 503–508 processing, efficiency analysis 494–498 in suspension of, MoS2 nanoparticles 515–516 upgrading, with carbon part removal catalytic cracking 493 deasphalting 492 thermal cracking 492–493 vanadium and nickel behavior 514 heavy oil hydrocracking reactions 312 heavy vacuum gas oil (HVGO) 297 heteroatomic adamantanes 136 heteroatomic bonds 385 hetero-organic petroleum compounds 139 hexacyclic benzohopanes 134 high-resolution mass spectrometry 131–132 high-field asymmetric waveform ion mobility spectrometry (FAIMS) 144 highly oriented pyrolytic graphite (HOPG) 297 high molecular weight gas (HMWG) 333 high-performance liquid chromatography (HPLC) 9, 115 high temperature gas chromatography (HTGC) 120 high-temperature oxidation (HTO) 158 high temperature simulated distillation (HTSD) 120 HMWG see high molecular weight gas (HMWG)

525

526

Index

HOF see heavy oil feedstock (HOF) HOMO-LUMO gaps 400, 401 Hossain’s method 458 HPLC see high-performance liquid chromatography (HPLC) HTGC see high temperature gas chromatography (HTGC) HTSD see high temperature simulated distillation (HTSD) hybrid EOR methods hybrid thermal-chemical methods 164 hybrid thermal-NCGs methods 164 hybrid thermal-solvent methods 163–164 hybrid thermal-VAPEX 162 hydrocarbon-bearing reservoir 40 hydrocarbon footprint 190 hydrocarbon molecule 238 hydrocarbon-nonhydrocarbon (HC_NHC) systems 461 hydrocarbons, property estimation of critical parameters Kesler–Lee’s method 457 Twu’s method 455–457 dead oil viscosity, with temperature dependence 457–458 hydroconversion process 489 hydrocracking 311 activation energies 352–364 global reaction order, for residue conversion 347–349 kinetic models 374 reaction rate coefficients calculation 349–352 selectivity of 364–365 hydrodenitrogenation (HDN) 297 hydrodesulfurization (HDS) 27, 211–212, 297, 314, 323 hydrogenation-dehydrogenation reactions 311 hydrogenation reactions 310 hydrogen/carbon (H/C) ratio 309 hydrogen donors (HD) 169, 189 hydrogenolysis reactions 27 hydrolysis reactions 390, 392 hydrothermal-catalytic conversion 101 hydrothermal cracking reaction 246 hydrothermal transformations 80 hydrothermal upgrading (HTU) technique 168 hydrothermolysis 196 hydroxynitrate intermediates 275

HYperfine Sublevel CORrelation (HYSCORE) spectroscopy 40, 51–53

i immiscible gas injection 162 improved-oil-recovery (IOR) 40, 153 infrared spectroscopy (IR) method 79 IR-analysis of the aromatic hydrocarbon fraction 92 of the asphaltenes 95 of the resins 93 IR spectra A1-A2 subfractions 100 IR spectra A5-carbene-carboids asphaltene subfractions 101 IR spectral coefficients 98 initial boiling point (IBP) 515 injector temperature 116 in situ combustion (ISC) 247, 263, 415 THAI 158–159 THAI-CAPRI 159–160 traditional ISC 156–158 in situ upgrading evolved noncondensable gases CO2 and CO production 212–215 methane and C2+ generation 215–218 field tests 218–229 general aspects 168–170 hydrodesulfurization 211–212 hydrothermal upgrading process SARA fractions 171 water-gas shift reaction 171–172 minerals, as natural catalysts authigenic mineral formation 189 heavy oil extraction 193 kaolinite, formation of 190 kerogen cracking 198 reaction temperature, on HTU performance coke analysis, using FTIR-spectroscopy 210–211 evolved gas components analysis 201–202 liquid products analysis 202–210 material balance (products distribution) and pressure changes 200–201 viscosity and hydrothermal upgrading 186–189 water (steam) role, as green hydrogen donor deuterium exchange, possibility of 184–186 donating performance 174–179

Index

morphological and structural changes, of oil-soluble catalyst 182–184 upgrading performance 179–182 in situ upgrading, of heavy oils catalysts characterization 253–257 homogeneous catalysts 238–239 ionic liquids catalytic activity of 253 synthesis procedure 251–253 mineral catalysts activity of 250–251 synthesis procedure 247–250 oil-soluble catalysts activity of 244–246 synthesis procedure 243–244 water-soluble catalysts activity of 242–243 synthesis procedure 240–242 in situ upgrading process (IUP) 160 physical chemistry 383–385 insoluble particles (ISP) 506 inter-and intraparticle diffusion coefficients 435 intermediate carbocations 390 intrinsic reaction coordinates (IRC) 388 inverse Laplace transform (ILT) algorithms 71 ionic liquids catalytic activity of 253 synthesis procedure 251–253 ionic radii, of metals 406 ion mobility spectrometry (IMS-MS) 144 IR see infrared spectroscopy (IR) method iron acetyl acetonate catalyst 245 iron nanoparticles 288 iron(III) tris(acetylacetonate) 288 isomerization 314 isotope-tracing techniques 173

k Kang’s condition 438 kaolinite formation 190 kaolin-modified mesoporous-macroporous 298 Kendrick mass defect (KMD) 132 kerogen-containing formation 83 Kesler–Lee’s method 457 kinematic viscosity 436 kinetic modeling 310

kinetic models for, in situ and ex situ hydrocracking detailed-lumping kinetic model 323–326 five-lump kinetic models 318–322 four-lump kinetic models 316–318 six-lump kinetic models 322–323 kinetic parameters and model assumptions aquathermolysis kinetic models 341–344 optimal set of heavy oil aquathermolysis, study case 345–347 heavy oil hydrocracking, study case 345

l Langmuir adsorption equation 438 Langmuir adsorption isotherms 284 Langmuir model 284 laser desorption/ionization (LDI) 129–130 laser energy transfer 130 laser-induced acoustic desorption (LIAD) 129–130 lattice Boltzmann equations 436 Levenberg–Marquardt algorithm 340, 343 LF-NMR see low-field nuclear magnetic resonance (LF-NMR) Liaohe extra-heavy crude oil 345 Liaohe oilfield heavy oil 328 light catalytic cracking gas oil (LCCG) 504 linear regression 337 liquefied petroleum gas (LPG) 2 liquid-acid catalyst 321 liquid addition to steam for enhancing recovery (LASER) 164 liquid adsorption chromatography method 102 liquid chromatography 123–125 liquid products analysis changes, in SARA Fractions 205–206 elemental analysis and desulfurization, of oil samples 203–204 FTIR spectroscopy, of oils 204–205 SARA fractions GC analysis, of saturates 206 GC-MS analysis, of aromatics 206–207 MALDI-TOF measurements, of asphaltenes 208–210 NMR measurements, of resins 207–208 viscosity and API gravity, of oil 202–203 liquid pseudocomponents 323

527

528

Index

longitudinal (spin-lattice) relaxation time 49 low-field nuclear magnetic resonance (LF-NMR) 70 relaxation technique 74 low-temperature oxidation (LTO) 158, 247 LPG see liquefied petroleum gas (LPG) lumping sensitivity 466–468

m macroscopic density 436 macroscopic velocity 436 magnesium oxide nanoparticles 284 maltene fractions 427 mass spectrometry (MS) 114, 116 ionization methods APCI 126–127 ASAP-MS 128–129 CI 126 DART 130 desorption ionization techniques 130–131 EI 126 ESI 128 FI 127 LDI 129–130 PI 128 SPI 128 uHTMS Fourier transform orbitrap mass spectrometry 131–132 FT-ICR-MS 131 high-resolution mass spectrometry data 132 multireflection time-of-flight (TOF) mass analyzer 132 mass transfer mechanism 312 process 435, 440 mathematical regression technique 465 matrix-assisted laser desorption/ionization (MALDI) 129, 141 matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) technique 9, 422 matrix-matched calibration 121 medium pressure liquid chromatography (MPLC) 330 medium temperature oxidation (MTO) 158 mercapto-alcohols 314

mercapto-aldehydes 314 metallic precursors 296 metalloporphyrins 8 metal oxide nanoparticles catalysts preparation, in supercritical water 271 coprecipitation method 267–269 microemulsion method 270–271 sol-gel processing 269–270 metal-oxygen-metal bond 270 metal precursors 243 metal stearates 415 1-methyl-imidazole 252 micro carbon residue 9 microemulsion 267, 270–271 microwave (MW) radiation 41 mineral catalysts activity of 250–251 synthesis procedure blend of 248–249 single-mineral catalysts 248 synthetic catalysts, from minerals 249–250 minimum miscible pressure (MMP) 163 miscible gas injection 163 Mn2 O3 nanoparticles 286 Mo–Co clay-supported nanoparticles 298 Mo–Co metakaolin-supported nanoparticles 298 modified Jamal plot 14 modulation frequency 41 molecular dynamics (MD) simulations 388 molecular mechanics 388 molecular weight (MW) 9, 453 molybdenite 249 molybdenum catalysts 292 molybdenum octoate 323 molybdenum trioxide 249, 250 Mo–Ni nanoparticles 298 monometallic catalysts, in oil upgrading aluminum 284 calcium 285 cerium 292–293 chromium 286 cobalt 288–289 copper 291 iron 286–288 magnesium 284 manganese 286

Index

molybdenum 292 nickel 289–291 silicon 285 titanium 285 vanadium 285 Wolfram 293 zinc 291 zirconium 292 monomolecular nickel(II) stearate tetrahydrate 417 monomolecular reactions 390 mononuclear anions 251 Monte Carlo approach 340 Monte Carlo movement method 440 Mossbauer spectroscopy 273 MPLC see medium pressure liquid chromatography (MPLC) MRT see multirelaxation time (MRT) Mulliken population analysis 406 multicarbon number (MCN) 455 multiple-contact miscible (MCM) 163 multiple flame photometric detector (mFPD) 121 multiple reaction monitoring (MRM) 121 multiplet 61 multiple thermal fluids (MTFs) injection 164 multireflection time-of-flight (TOF) mass analyzer 132 multirelaxation time (MRT) 437

n n-alkane 44 nanonickel catalyst 289 nanoparticles, heavy oil in situ upgrading catalytic activity bimetallic and polymetallic catalysts 293–298 biogenic and complex organic supports 299 monometallic catalysts 282–293 supported catalysts 293–298 characterization 273–282 synthesis CNT 271–273 metallic nanoparticles 265–266 metal oxide nanoparticles 267–271 nanoparticulated molybdenum(IV) sulfide (MoS2 ) 292

nanoscale catalysts, in slurry process from oil-soluble precursors 501–503 from water-soluble precursors 503–508 naphthenic acid salts 8 naphthenic-aromatic compounds 314 natural bond orbital (NBO) isosurfaces 410 n-chloro-butane 252 Ni-based catalyst 92 nickel-based aquathermolysis catalyst 198 nickel-molybdenum interactions 253 nickel oxide (NiO) nanoparticles 289, 291, 295 nickel-sulfide nanoparticles 184 Ni-Mo/-activated carbon (Ni-Mo/AC) hydrotreating catalysts 296 Ni-Mo based liquid acid catalyst 242 NiO-coated Al2 O3 nanoparticles 284 NiO-ketjenblack nanoparticles 296 Ni-O stretching vibration 275 nitrogen-and oxygen-containing components 386 nitrogen-, oxygen-, and sulfur-containing (NSO) compounds 138 nitrogen-phosphorus detector (NPD) 121 NMR see nuclear magnetic resonance (NMR) noncatalytic aquathermolysis 81, 173 noncatalytic cracking 290 noncondensable hydrocarbons 179 nonlinear optimization 344 nonlinear regression 337 nonporphyrin metal compounds 8 nonsupported bimetallic catalysts 293 nonsupported polymetallic catalysts 294–295 n-pentane 331 nuclear magnetic resonance (NMR) 49 relaxometry 59 oil samples and individual SARA fractions 71–75 oil-saturated rock samples and individual SARA fractions 71–75 technical aspects of 70–71 spectroscopy 59 13 C NMR 64 1 H NMR 63 oil samples study and SARA fractions 65–70 phenomenon of 60–62 technical aspects of 62–65

529

530

Index

numerical simulation numerical validation lab-scale kinetic models 470–474 in static conditions 468–470 oil phase behavior, modeling cubic EoS and phase behavior 460–468 oil characterization 454–455 property estimation of, hydrocarbons 455–458 special data requirement 458–460 reaction scheme 453–454 upscaling laboratory-scale, to field-scale optimal grid type and size, selection of 479–482 upscaling laboratory data 477–479 upscaling steam processes 474–477 N-vinyl imidazole 252

o oil-bearing formation 219 oil distillation 501 oil fingerprinting 116 oil fraction composition 71 oil phase behavior, modeling cubic EoS and phase behavior binary interaction coefficients 461–462 lumping sensitivity 466–468 The Peng and Robinson EoS 460–461 tuning an, EoS 464–466 volume translation 462–464 oil characterization 454–455 property estimation of, hydrocarbons 455–458 special data requirement 458–460 oil-soluble catalysts activity of 244–246 synthesis procedure 243–244 oil-soluble Co-based catalyst 98 oil-soluble iron-based catalyst 96 oil-soluble Mo dispersed catalyst 322 oil-soluble molybdenum-containing precursors 502, 503 oil-soluble Mo precursor catalyst 323 oil viscosity 68, 154 with temperature dependence 458–460 olefin 313 1D grid-type model 471 optimization algorithm 343 organic feedstock 333

organometallic catalysts 251 organometallic precursors 266 organosulfur compounds 299 organosulfur model compounds 313 original oil in place (OOIP) 154, 291 oxidation coefficient 109 oxidative catalytic cracking process 103

p palladium(II) oxide (PdO) nanoparticles 295 paramagnetic species/centers (PCs) 40 PASH see polycyclic aromatic sulfur compounds (PASH) Peaceman formula 482 Peclet numbers 436 PERKIN ELMER spectrum 79 PES see potential energy surface (PES) petroleomics 115, 138 petroleum hydrocarbons 116 petroleum oil 1 petroleum vanadylporphyrin molecule 43 Petzold–Gear backward differentiation formulae (BDF) method 340 phenyl polydimethylsiloxane 116 phosphorus selective (pulsed) flame photometric detectors (FPD/PFPD) 121 photoionization (PI) method 128 pillared clay (PILC) 247 pillar technique 247 π-electron system 421 π-radicals 44 Planck constant 41 polycyclic aromatic sulfur compounds (PASH) 139 polycyclic aromatic sulfur heterocycle compounds (PASHCs) 184 polymer flooding 161 polymerization 206 polymetallic nanoparticulated catalysts 293 polynuclear anions 251 pore space heterogeneity 440 porous media, behavior catalyst catalyst distribution, in pore space 4D X-Ray CT 446–450 heterogeneity coupled with Damkohler numbers 444–445 with Peclet number 442–444 with porosity 445–446

Index

methods artificial digital models of 440 catalyst distribution 4D microtomography, 441–442 intraparticle diffusion model, validation of 440–441 mathematical model 436–440 potential energy surface (PES) 388 pour point 5 powder ammonium heptamolybdate 325 pre-exponential factor 353–358 Proton-20 71 pulsed flame photometric detector (PFPD) 121 pyrene 422 molecules 426 oxidation process 426

q quantum chemical approaches calculation methods, for molecular modeling 387–388 model compounds for, in situ process simulation 385–386 thermochemical and kinetic parameters 388–389

r radiofrequency (RF) pulse 54 reaction rate coefficients 353–358 reactor modeling 312 reference frequency 61 regularization parameter 71 relative hydrogen index (RHI) 73 resemblant reaction rate equations 322 reservoir simulation 312 residual simulated distillation (RSD) 120 residue fraction (RS) 324 resinous-asphaltene compounds 109 reverse combustion 156 Reynolds numbers 436 RHI see relative hydrogen index (RHI) Riphean–Vendian complex 95 Runge–Kutta method 340, 344

s SAGD see steam-assisted gravity drainage (SAGD) SALDI see surface-activated laser desorption/ionization (SALDI)

saturated and aromatic compounds, resins, and asphaltenes (SARAs) 59 composition 75 fractionation 10, 15–16, 171 scanning electron microscopy (SEM) 182, 255, 273, 423 adsorbed Fe3 O4 nanoparticles 281 iron oxide catalyst particles surface 280 NiO nanoparticles 280 SCW see supercritical water (SCW) batch reactor 286 second kinetic model 322 seed particles 265 selected ion monitoring (SIM) 116 short distance oil displacement (SDOD) process 158 silicon photodiode 121 simulated distillation 120 single-photon ionization (SPI) 128 SiO2 nanoparticles 285 six-lump kinetic models 322–323 slurry-phase hydrocracking 322 Soave–Redlich–Kwong (SRK) EOS 462 sodium dioctylsulfosuccinate 504 soft desorption methods 115 soft ionization method 115, 126 soft or low-energy ionization 126 sol-gel processing 267, 269–270 solid and gas fractions 326 solid-phase components 74 solvent-aided process (SAP) 164 solvent-enhanced steam flooding (SESF) 164 solvothermal methods 267 specific gravity (SG) 453, 457 spectral coefficients (SC) 79, 89 spectral methods 114 spectroscopic methods 238 spectroscopic splitting factor 41 spin-orbital interactions 41 spin-spin interactions 41 Spin Track relaxometer 71 spiral ion trajectory technique 132 stainless-steel autoclave 327 standard rate constants 399 standard split injector 117 Stankiewicz plot (SP) 14 steam-alternating-solvent (SAS) method 164

531

532

Index

steam-assisted gravity drainage (SAGD) 154–156, 168, 263, 309 steam-based injection 164 steam-based technology 154 steam drive see steam flooding steam flooding 154, 156, 309 steam injection process 222, 331 CSS 154–155 SAGD 155–156 steam flooding 156 steam-oil-ratio (SOR) 226 steam stimulation process 168 see also steam injection process stoichiometric coefficient 330, 332, 334, 335 sulfidation of, dispersed molybdenumcontaining catalysts 508–509 sulfone molecule 414 sulfoxide groups 86 sulfur 386 sulfur aromatic compounds 6 sulfur-based aquathermolysis kinetic models 368 sulfur-based kinetic model 331 sulfur compounds 96 sulfur removal model 324 sulphidity index 82 supercritical water (SCW) 40, 169, 267 superficial carbenium ions 238 supersonic flow velocity 126 supersonic molecular beam mass spectrometry (SMB-MS) 126 supported-nanoparticles, coated with metal nanoparticles alumina-supported nanoparticles 296 carbon-supported nanoparticles 296–297 complex inorganic supports 298 silica-supported nanoparticles 295–296 zeolites, as support 297–298 surface-activated laser desorption/ionization (SALDI) 129 surface adsorbate concentration 438 surfactant flooding 161 surfactant/polymer (SP) 162 surfactants 161 surplus steam 220 synthesized amorphous silica gel nanoparticles 285

synthesized silica nanoparticles 295 synthetic catalysts, from minerals dispersion 250 molybdenum trioxide, purification of 249–250 pulverization and calcination 249 synthetic oil-soluble iron-based catalyst 80

t target petroleum hydrocarbons 116 TBP see true boiling point (TBP) tetrahydrothiophene 326 aquathermolysis reaction 327 TG see thermogravimetry (TG) thermal cracking 299 thermal effect 263 thermal-enhanced oil recovery (EOR) methods 309, 466 ISC THAI 158–159 THAI-CAPRI 159–160 traditional ISC 156–158 steam injection CSS 154–155 SAGD 155–156 steam flooding 156 thermal treatment IR spectra of asphaltenes 83 of resins 82 spectral coefficients of asphaltenes 82 of resins 82 thermocatalytic treatment 91 thermogravimetric analyzer (TGA) 192 thermogravimetry (TG) 254, 422 thiophene 287 time-domain nuclear magnetic resonance (TD-NMR) 70 time-of-flight (TOF) mass analyzer 129 titanium oxide (TiO2 ) 285 toe-to-heel air injection (THAI) 158–159, 263 toluene-insoluble components (TIC) 510 toluene-insoluble particles (TIP) 505 total adsorption 438 total petroleum hydrocarbons (TPH) 116 TPH see total petroleum hydrocarbons (TPH)

Index

transmission electron microscope (TEM) 267 fresh NiMo/SiO2 282 nanonickel catalyst 282 trans-stilbene (SB) 412 formation 429 transverse magnetization, of protons 71 traveling wave ion mobility spectrometry (TW-IMS) 144 1-(3-trimethoxysilane) propyl-3-methylimidazole chloride 253 true boiling point (TBP) 10, 117, 325 conventional method 120 standardized methods 17, 18 trust-region-reflective algorithm 344 two-dimensional GC with nitrogen phosphorus detection (GCxGC-NPD) 122 two-dimensional nuclear magnetic resonance spectroscopy (2D NMR) 67, 123 Twu’s method 455–457 critical pressure 456 critical temperature 456 critical volume 457

vapor assisted petroleum extraction (VAPEX) 162 vapor-liquid equilibrium (VLE) 460 Veba combi cracking (VCC) 490 vibrating sample magnetometry (VSM) 267 1-(3-vinyl imidazolio-1-yl)-4-(4-amino-4pyridyldihexaethoxy monomethyl ether) butane 252 1-(3-vinyl imidazolio-1-yl)-4-(4-4-pyridyl tetradecanamido) butane 252 1-(3-vinyl imidazolio-1-yl)-12-(4-4-pyridyl tetradecanamido) dodecane 252 viscosity vapor pressure osmometry 9 VLE see vapor-liquid equilibrium (VLE) Voigt function 42, 73 volatile thermotolerant compounds 115 Volga-Ural oil 83 Volterra-type integrodifferential equation 325 Voronoi diagram 440 VPI see vacuum photon ionization (VPI) VSM see vibrating sample magnetometry (VSM)

u

w

uHRMS see ultrahigh-resolution mass spectrometry (uHRMS) ultradispersed catalysts 319 ultradispersed trimetallic catalyst 319 ultrahigh-resolution mass spectrometry (uHRMS) 115 Fourier transform orbitrap mass spectrometry 131–132 FT-ICR-MS 131 high-resolution mass spectrometry data 132 multireflection time-of-flight (TOF) mass analyzer 132 ultrasonic particle analyzer Zeta-APS 275 uncoated silica-fumed nanoparticles 285 undesirable compounds 6 up-flow tubular reactor 320 upgrading reactions 215

warm VAPEX 162 water-alternating-gas (WAG) 164 water-gas shift reactions (WGSR) 171–172, 184, 314, 328, 334 equilibrium 215 water-oil-rock system 238 water (steam) role, as green hydrogen donor deuterium exchange, possibility of 184–186 donating performance, evaluation elemental composition of, oil samples and SARA fractions 179 FTIR spectroscopy measurement, oil samples and SARA fractions 174–178 isotope analysis of, oil samples and SARA fractions 178–179 morphological and structural changes, of oil-soluble catalyst 182–184 upgrading performance, evaluation distribution of, n-alkanes in saturates 181–182 GC analysis of, evolved gases 179, 180 oil viscosity and elemental analysis 180–181 SARA analysis, of oil 181

v vacuum photon ionization (VPI) 128 vanadium-based species 8 vanadium pentoxide (V2 O5 ) 285 van Hoff–Arrhenius law 409

533

534

Index

water-soluble catalysts activity of 242–243 characterization 254 synthesis procedure 240–242 water-soluble iron compounds 505 water-soluble metal catalyst 239 water-soluble salts 239 Wolfram trioxide (WO3 ) nanoparticles Wulf–Bragg equation 511

x xantham-gum polymer 299 X-ray diffraction (XRD) pattern 267

X-ray diffractometer 328 X-ray fluorescence (XRF) spectrometer 328 spectroscopy 193 X-ray powder diffraction (XRD) technique 423 X-ray spectroscopy 277 292, 293

z Zeeman interaction 41 zeolites 297 zirconium oxide (ZrO2 ) 292 ZnO nanoparticles 291

WILEY END USER LICENSE AGREEMENT Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.