Encyclopedia of Chemical Processing and Design: Volume 48 - Residual Refining and Processing to Safety: Operating Discipline (Chemical Processing and Design Encyclopedia) [1 ed.] 1584885653, 9780824724986, 0824724984

"Written by engineers for engineers (with over 150 International Editorial Advisory Board members),this highly laud

200 16 98MB

English Pages 496 [497] Year 1994

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Encyclopedia of Chemical Processing and Design: Volume 48 - Residual Refining and Processing to Safety: Operating Discipline (Chemical Processing and Design Encyclopedia) [1 ed.]
 1584885653, 9780824724986, 0824724984

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Encyclopedia of Chemical Processing and Design 48

EXECUTIVE EDITOR JOHN J. McKETTA The University of Texas at Austin Austin, Texas EDITORIAL ADVISORY BOARD

L Y L E F. A LB R IG H T Purdue University Lafayette, Indiana

JAM ES R. FA IR Professor o f Chemical Engineering The University o f Texas Austin, Texas

ER N EST E. LUDWIG Ludwig Consulting Engineers, Inc. Baton Rouge, Louisiana

JOHN HAPPEL Columbia University New York, New York

R. A. M cKETTA Chemical Engineer Purvin and Gertz, Inc. Houston, Texas

Encyclopedia of Chemical Processing and Design___________ John J. McKetta William A. Cunningham

e x e c u t i v e e d it o r a s s o c i a t e e d it o r

Residual Refining and Processing to Safety, Operating Discipline

MARCEL

d e k k e r , in c .

NEW YORK • BASEL • HONG KONG

Library of Congress Cataloging in Publication Data Main entry under title:

Encyclopedia o f chemical processing and design.

Includes bibliographical references. 1. Chemical engineering— Dictionaries Technical— Dictionaries. I. McKetta, John J. II. Cunningham, William Aaron. Tp9.E66 660.2'8'003 ISBN: 0-8247-2498-4

2.

Chemistry,

75-40646

COPYRIGHT © 1994 by MARCEL DEKKER, INC. ALL RIGHTS RESERVED.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage and retrieval system, without permission in writing from the publisher.

MARCEL DEKKER, INC. 270 Madison Avenue, New York, New York, 10016

Current printing (last digit): 10 9 8 7 6 5 4 3 2 1

PRINTED IN THE UNITED STATES OF AMERICA

International Advisory Board RAY C. ADAM

LUCIANO BENINCAMPI

NICHOLAS P. CHOPEY

Former Chairman of the Board N. L. Industries, Inc. New York, New York

Manager o f Public Relations CTIP— Compagnia Tecnica Industrie Rome, Italy

Editor-in-Chief Chemical Engineering Magazine McGraw-Hill, Inc. New York, New York

CARL W. ALBERS

LLOYD BERG

Senior Process Engineer M. W. Kellogg Houston, Texas

Professor Department of Chemical Engineering Montana State University Bozeman, Montana

FRANK CHRENCIK

NEIL S. BERMAN

V ice President American Petrofina Dallas, Texas

M. A. ALLAWALA Managing Director National Refinery Ltd. Karachi, Pakistan

HAMED H. AMER Chairman Agiba Petroleum Co. Cairo, Egypt

R. G. ANTHONY Professor, Department o f Chemical Engineering Texas A & M University C ollege Station, Texas

H. J. AROYAN Former Vice President Chevron Research Company Richmond, California

F. SID ASKARI President Technolog, Inc. Engineering and Industrial Consultants Tehran, Iran

DONALD L. BAEDER Former Executive Vice President— Science and Technology Occidental Petroleum Corporation Los A ngeles, California

WM. A. BAILEY Jr. Former Director, MTM Process Research and Development Lab Shell Development Company Houston, Texas

TRAVIS W. BAIN Vice President National Sales, Inc. Jackson, M ississippi

GAREN BALEKJIAN C. F. Braun Arcadia, California

CESAR BAPTISTE

Professor o f Chemical Engineering Engineering Center Arizona State University Tempe, Arizona

D. J. BLICKWEDE Former Vice President and Director o f Research Bethlehem Steel Corp. Bethlehem, Pennsylvania

M. J. P. BOGART Fluor Engineers and Constructors, Inc. Santa Ana, California

Z. D. BONNER Vice Chairman of the Board Tesoro Petroleum Corp. San Antonio, Texas

JOSEPH F. BOSICH Bosich Consultants Humble, Texas

WILLIAM H. BOSLER President Texas Consultants, Inc. Houston, Texas

ARCHIE BROODO President AID, Inc. Dallas, Texas

ARTHUR W. BUSCH Environmental Engineer Consultant Dallas, Texas

ROBERT C. BUTLER Administrative Assistant and Planning Manager, Petroleum Chemicals Division E. I. du Pont de Nemours and Co. Wilmington, Delaware

J. MORSE CAVENDER

Vice President Petroleos Mexicanos Mexico City, Mexico

President The Mactan Company Dusseldorf, Federal Republic o f Germany

LEON R. MARTINEZ BASS

PRAMOTE CHAIYAVICH

Sales Manager— Northern Mexico Zincamex, S. A. Saltillo, M exico

Chief Technologist The Tahi Oil Refinery C o ., Ltd. Bangkok, Thailand

ROBERT O. BATHIANY

S. D. CHELLAPPAN

Technical Planner Weyerhauser Company Tacoma, Washington

Process Engineering Manager Occidental Chemical Corporation Houston, Texas

Vulcan Materials Co. Birmingham, Alabama

R. JAMES COMEAUX

C. W. COOK Chairman, Executive Committee General Foods Corp. White Plains, New York

CHARLES F. COOK Vice President Research and Development Phillips Petroleum B artlesville, Oklahoma

EARL J. COUCH Research Associate Mobil Research and Development Corp. Dallas, Texas

JAMES R. COUPER Professor Department o f Chemical Engineering University o f Arkansas Fayetteville, Arkansas

HORACE R. CRAWFORD Senior Staff Scientist CONOCO Corp. Houston, Texas

ORAN L. CULBERSON Chemical Engineer Oak Ridge National Lab Chemical Technology Division Oak Ridge, Tennessee

DONALD A. DAHLSTROM Vice President, Research and Development Process Equipment Group Envirotech Corp. Salt Lake City, Utah

PERRY P. DAWSON Production Engineer Dow Chemical Co. Freeport, Texas

ELBERT M. DeFOREST Former Director o f Technology, Chemicals and Metals Vulcan Materials Co. Wichita, Kansas

ROBERT G. DENKEWALTER Corporate Vice President Technology Allied Corp. Morristown, New Jersey

iii

International Advisory Board

iv

J. P. de SOUSA

RALPH T. FERRELL

OM P. GOYAL

Publisher Chemical Age o f India Technical Press Publication Bombay, India

Senior Vice President, Corporate Development Vista Chemical Company Houston, Texas

Technagement Consultant New Bombay, India

JAMES D. D'lANNI

LOUIS FEUVRAIS

Former Director o f Research The Goodyear Tire and Rubber Co. Akron, Ohio

Directeur General Ecole Nationale Superieure D ’Arts et Metiers Paris, France

JUAN M. DIAZ Production General Manager Rohm and Haas M exico, S. A. C. V. M exico City, Mexico

WERNER DIMMLING Dipl-Chemist Friedrich Uhde GmbH Dortmund, Federal Republic of Germany

S. W. DREW Executive Director MCMC Technical Operations Merck & C o., Inc. Rahway, New Jersey

BARRETT S. DUFF Barrett S. D uff and Associates South Pasadena, California

R K. DUTTA Project Manager Chemical and Metallurgical Design Company, Private Ltd. New Delhi, India

WILLIAM F. EARLY Vice President Stone & Webster Environmental Services Houston, Texas

WALTER EMRICH Consultant Teterboro, New Jersey

E. FREDERICO ENGEL Member of the Board o f Management Chemische Werke Hiils AG Marl, Federal Republic of Germany

R E. G. M. EVERS Operations Manager Anzo Salt Chemical Delfzyl D elfzyl, The Netherlands

ALEXANDRE EVSTAFIEV Director, Division of Technological Research and Higher Education UNESCO— Paris Paris, France

GERALD L. FARRAR President Farrer Associates Tulsa, Oklahoma

F. M. FARRELL Technical Director 3M Company St. Paul, Minnesota

C. SHULTS FAULKNER President C. S. Faulkner, Inc. Houston, Texas

R. A. FINDLAY Former Director, Fuels and Lubricants, Research and Development Phillips Petroleum Company Bartlesville, Oklahoma

DALE FRIDLEY Manager, Intermediates Technology Division Exxon Chemical America Baton Rouge, Louisiana

ROBERT H. FRITZ President Loss Control Consultants, Inc. Alvin, Texas

GARY L. FUNK Director, Advanced Process Control Technology Brown & Root/ITI Division Houston, Texas

BILL F. GALLOWAY Plant Manager Quantum USI Division Port Arthur, Texas

DONALD E. GARRETT President Saline Processors Ojai, California

L. W. GARRETT, Jr. President Garrett A ssociates, Inc. San M ateo, California

ROY D. GERARD General Manager Westhollow Research Center Shell Development Company Houston, Texas

ION GHEJAN Department o f Chemical Engineering Institute o f Petroleum, Gas, and Geology Bucharest, Romania

JIM GILLINGHAM General Manager, Process Engineering Diamond Shamrock San Antonio, Texas

B. GENE GOAR Goard, Allison, and Associates, Inc. Tyler, Texas

MARCEL GOLDENBERG SAM IN Corp., Inc. New York, New York

WILHELM GRAULICH Director, Manager, Rubber Division Bayer AG Leverkusen, Federal Republic o f Germany

E. HENRY GROPPE Groppe, Long, & Littell Houston, Texas

GIANFRANCO GUERRERI INGECO Altech Group Societa per Azioni Con Sede in Milano Milan, Italy

KENNETH M. GUTHRIE Cost Consultant Marina Del Rey, California

NORMAN HACKERMAN Former President Rice University Houston, Texas

VLADIMIR HAENSEL Vice President, Science and Technology Universal Oil Products Co. Des Plaines, Illinois

HENRY E. HALEY V ice President Arthur D. Little, Inc. Cambridge, Massachusetts

R. L. HARVEL Project Manager Dow Chemical International Ltd. Tokyo,Japan

J. W. HAUN Former Vice President and Director o f Engineering General Mills, Inc. Minneapolis, Minnesota

TERUAKI HIGUCHI President Japan Fody Corp. Osaka, Japan

JOHN R. HILL, Jr. President and C hief Executive Officer Gifford-Hill & C o ., Inc. Dallas, Texas

PAUL E. HIME Former Vice President Operation & Technical Hoechst Celanese Chemical Group Dallas, Texas

HAROLD L. HOFFMAN Editor Hydrocarbon Processing Houston, Texas

NORBERT IBL Professor Eidg. Techn. Hochscule Zurich Techn.— Chemie Zurich, Switzerland

v

International Advisory Board

RUBEN F. INGA

W. S. LANIER

BRYCE I. MacDONALD

President Confederacion Interamerican de Ingeniera Quimica Lima, Peru

Project Manager Seadrift Expansion Projects Union Carbide Corp. Houston, Texas

Manager, Environmental Engineering General Electric Company Fairfield, Connecticut

JAMES R. JOHNSON

CLARK P. LATTIN, Jr.

Former Executive Scientist and Director, Advanced Research Programs Laboratory 3M Company, Central Research Labs Saint Paul, Minnesota

Former President The M. W. Kellogg Company Houston, Texas

Professor School o f Chemical Engineering Oklahoma State University Stillwater, Oklahoma

NAJI A. KADIR President Scientific Research Council Baghdad, Iraq

JOHN E. KASCH Former Vice President Standard Oil Indiana Escondido, California

RAPHAEL KATZEN Managing Partner Ralph Katzen Associates Cincinnati, Ohio

JOHN J. KELLY

ISIDORO LAZARRAGA-LEANZA

KLAUS MAI

C hief o f Engineering and Control Empresa Nacional del Petroleo Vina del Mar, Chile

Former President Shell Development Houston, Texas

JEAN Le BRETON

STANLEY D. MARTS

Managing Director Elf Aquitaine Indonesie Jakarta, Indonesia

Supply Specialist Shell Oil Company Houston, Texas

IRV LEIBSON

F. DREW MAYFIELD

Vice President Bechtel Corp. San Francisco, California

Drew Mayfield & Associates Baton Rouge, Louisiana

PIERRE Le PRINCE

Department o f Chemical Engineering University C ollege, Dublin Dublin, Ireland

Director of Refining and Engineering Center Instiut Francaise de Petrole Malmaison, France

HENNO KESKKULA

C. E. LETSCHER

Research Fellow Chemical Engineering Department The University o f Texas at Austin Austin, Texas

Caltex Petroleum Company New York, New York

C. J. LIDDLE

O. P. KHARBANDA

White Young & Partners Ltd. Herts, England

O. P Kharbanda & Associates Cost and Management Consultants Bombay, India

NORMAN N. LI

WLODZIMIERZ KISIELOW Professor o f Petroleum Technology, Director of Research Department o f Petroleum and Coal Centre of Polish Academy of Sciences Krzywoustego, Poland

ROBERT A. KLEIN President and C hief Executive Officer Continental Controls, Inc. Houston, Texas

MOHAN SINGH KOTHARI C hief Consultant Punjab Industrial Consultancy Organisation Chandigarh, India

G. R. KRUGER President Semarck, Inc. Houston, Texas

A. P. KUDCHADKER

R. N. MADDOX

Director, Chemical & Process Technology Allied Signal Engineered Materials Research Center Des Plaines, Illinois

DAVID C. K. LIN Senior Engineer Owens Corning Fiberglas Corp. Newark, Ohio

CHARLES E. LOEFFLER Technical Manager Celanese Chemical Company Pampa, Texas

T. N. LOLADZE Vice-Rector, Professor o f the Georgian Polytechnic Institute Tbilisi, USSR

STANLEY L. LOPATA Chairman o f the Board Carboline Company Saint Louis, Missouri

Professor o f Chemical Engineering and Dean o f Student Affairs Indian Institute o f Technology, Kanpur Kanpur, India

Chemical Engineer Consultant Austin, Texas

RALPH LANDAU

W. D. LUEDEKE

Former Chairman Halcon International, Inc. New York, New York

Former Planning Manager E. I. du Pont de Nemours Wilmington, Delaware

PHILIPS S. LOWELL

GUY McBRIDE Former President Colorado School o f Mines Golden, Colorado

CLYDE McKINELY Former Director, Allentown Labs Air Products and Chemicals, Inc. Allentown, Pennsylvania

RICARDO MILLARES President Papel Satinado, S. A. Mexico City, Mexico

ROBERT L. MITCHELL Former Vice Chairman o f the Board Celanese Corp. N ew York, New York

RICHARD MOLLISON General Manager Colpapel, S. A. Pereira, Columbia

DONALD D. MULRANEY C. F. Braun Co. Alhambra, California

CARLOS EPSTEIN MURGUIA General Manager and President o f the Board Industrias Guillermo Murguia, S. A. Naucalpan, Mexico

TAKAYUKI NATE Plastics Sales Department Tonen Petrochemical Co. Ltd. Tokyo, Japan

JAMES K. NICKERSON Research Associate Esso Research and Engineering Company Summit, N ew Jersey

ALEX G. OBLAD Distinguished Professor o f Chemistry Mining, and Fuels Engineering University o f Utah Salt Lake City, Utah

vi

International Advisory Board

H. E. O'CONNEL

H. KEN RIGSBEE

JOHN W. SHEEHAN

Former President Tenneco Chemicals Inc. Houston, Texas

Project Manager Phillips 66 Natural Gas Company Houston, Texas

ERNEST O. OHSOL

FRANK S. RIORDAN, JR.

Vice President, Manufacturing and Marketing Champlin Petroleum Company Kerrville, Texas

Consultant Ohsol Technical Associates Crosby, Texas

I. O. OLADAPO Dean of Engineering University o f Lagos Lagos, Nigeria

GORDON F. PALM President Gordon, F. Palm & Associates Lakeland, Florida

F. F. PAPA-BLANCO Advisor o f Educational Technology lnstituto Latino Americano de la Communicacion Educativa M exico City, M exico

DILIP M. PATEL Manager o f Process Design & Technology John Brown E & C, Inc. Houston, Texas

Director, Technology Planning Monsanto Textiles Company Saint Louis, Missouri

DENNIS F. RIPPLE Technical Manager, Process Technology Hoechst Celanese Corporation Corpus Christi, Texas

LOUIS R. ROBERTS

THOMAS C. PONDER Petrochemicals Editor Hydrocarbon Processing Houston, Texas

R. G. H. PRINCE Professor, Head o f Department Chemical Engineering University o f Sydney Sydney, Australia

Direttore D ivisione Resine Societa Italiana Resine Milano, Italy

ROBERTO RODRIGUEZ INTEVEP Caracas, Venezuela

GERHARD ROUVE Director o f the Institute for Water Resources Development Technical University Aachen Aachen, Federal Republic o f Germany

FRANCIS E. REESE Former Vice President and Managing Director International Monsanto Company Saint Louis, Missouri

AURELIO REITER Former Research Manager of Esso Standard Italiana Roma-Italy Rome, Italy

LARRY RESEN Larry Resen Associates W ilton, Connecticut

CARL I. SOPCISAK

PETER H. SPITZ President Chemicals System s Inc. New York, New York

SAM STRELZOFF Consultant Marlboro, Vermont

MARK B. STRINGFELLOW President & C hief Executive Officer Environmental Control Group, Inc. Maple Shade, New Jersey

JOHN H. SANDERS Vice President and Assistant General Manager Eastman Chemicals Division Eastman Kodak Company Kingsport, Tennessee

HIDESHI SATO

Professor of Chemical Engineering University o f N ew South Wales Kensington, Australia

ARTHUR L. SMALLEY, Jr.

Technical Consultant Synthetic Fuels Wheat Ridge, Colorado

General Manager Technical Information Office Technical Development Department Nippon Steel Corp. Tokyo,Japan

J. S. RATCLIFFE

Professor o f Chemical Engineering University o f Calgary Alberta, Canada

RICCARDO ROBITSCHEK

Project Manager The Pace Consultants, Inc. Houston, Texas Dow Chemical USA Texas Operations Industrial Chemicals Division Freeport, Texas

PHILIP M. SIGMUND

President Matthew Hall Inc. Houston, Texas

HUGH S. PYLANT

EDWIN L. RAINWATER

Designer Esso Engineering Services Ltd. Surrey, England

Director, Planning and Source Evaluation Texas Air Control Board Austin, Texas

MARCELLO PICCIOTTI Technical Promotion Manager TechniPetrol-Rome Rome, Italy

PIERRE SIBRA

GEORGE E. SCHAAL Manager, Research and Development Produits Chimiques Ugine Kuhlmann Pierre-Benite, France

GERT G. SCHOLTEN Managing Director Edeleanu Gesellschaft mbH Frankfurt/Main. Federal Republic o f Germany

DOUGLAS M. SELMAN Vice President Business Development & Technology Exxon Chemical Company Darien, Connecticut

M. L. SHARRAH Former Senior Vice President Continental Oil Company Stamford, Connecticut

Y. S. SURY CIBA-Geigy Chemical Corp. Saint Gabriel, Louisiana

MICHAEL W. SWARTZLANDER Staff Engineer Union Carbide Corp. South Charleston, West Virginia

T. SZENTMARTONY Associate Professor Technical University Budapest Budapest, Hungary

M. TAKENOUCHI General Manager o f Manufacturing Department Maruzen Oil C o ., Ltd. Tokyo, Japan

VLADIMIR TEPLYAKOV Head o f Membrane Research Center A. V. Tochiev Institute o f Petrochemical Synthesis The USSR Academy o f Sciences Moscow, Russia

SOONTHORN THAVIPHOKE Managing Director S. Engineering Services C o., Ltd. Bangkok, Thailand

ROBERT S. TIMMINS Core Laboratory Aurora, Colorado

International Advisory Board

vii

A. A. TOPRAC

T. J. WALKER

JACK W. WESTERFIELD

President Interchem-Hellas Athens, Greece

Former Production Manager Dow Chemical Europe S. A. Zurich, Switzerland

Manager, Project Engineering Diamond Shamrock San Antonio, Texas

YORGI A. TOPRAKCLOGLU

J. C. WALTER, Jr.

D. L. WILEY

J. C. Walter Interests Houston, Texas

Former Senior Vice President Union Carbide Corp. Danbury, Connecticut

Chairman of the Board of Directors Marshall Boya ve Vernik Sanayii A. S. Istanbul, Turkey

HERNANCO VASQUEZ-SILVA

THEODORE WEAVER

President Hernando Vasqez & Associates, Ltd. Bogota, Columbia

Director o f Licensing Fluor Corporation Los A ngeles, California

M. A. VELA

ALBERT H. WEHE

President VELCO Engineering, Inc. Houston, Texas

Chief, Cost and Energy U. S. Government Raleigh, North Carolina

JUAN JOSE URRUELA VILLACORTA

JACK C. WILLIAMS Former Vice President Texaco, Inc. Houston, Texas

MASAMI YABUNE Section Head, Technical Section Tonen Petrochemical C o ., Ltd. Tokyo,Japan

GUY E. WEISMANTEL

LEWIS C. YEN

Ingeniero Fabrica de Jabon “ La Luz, S. A .” Guatemala

President Weismantel International Houston, Texas

Manager, Technical Data M. W. Kellogg Company Houston, Texas

S. P. VOHRA

PAUL B. WEISZ

STANELY B. ZDONIK

Managing Director Bakelite Hylam, Ltd. Bombay, India

Distinguished Professor Chemical and Bio-Engineering University o f Pennsylvania Philadelphia, Pennsylvania

Vice President and Manager Process Department Stone and Webster Engineering Corp. Boston, Massachusetts

A. L. WADDAMS Former Manager, Marketing Services Division BP Chemicals International Ltd. London, England

Contributors to Volume 48 Gale D. Beckett

The Procter & Gamble Company, Cincinnati, Ohio: Roasting,

F luidized-B ed

Geof Brazier

International Marketing Manager, BS & B Safety System s, Tulsa, Okla­

homa: Rupture D isks

C. Thomas Breuer

Arthur D. Little, Inc., Cambridge, Massachusetts: R etrofitting,

C oal-F ired B oilers

Louis J. Cabano

President and CEO, Pathfinder, Inc., Cherry H ill, N ew Jersey: R etro­

fit, B asics

Stephen R. Cammarn

The Procter & Gamble Company, Cincinnati, Ohio: R oasting,

F luidized-B ed

Nicholas P. Cheremisinoff, Ph.D.

Exxon Chemical Company, Linden, N ew Jersey:

R heology and Rheom etry

John D. Constance, P.E.

Consultant, C liffside Park, N ew Jersey: R o o f Ventilation,

Sizing

Michel Davidson

Senior Engineer, Institut Frangais du Petrole, Rueil M alm aison, France: R esidual Refining and P rocessing

Kenneth F. Foley

Hercules Incorporated, M cQueeney, Texas: Resins, H ydrocarbon

Arvind G. Godse

Kuwait National Petroleum Company, Shuaiba, Kuwait: Rotating Equipm ent, Shop Testing

Charles H. Gooding, Ph.D., P.E.

Professor o f Chemical Engineering, Clem son Univer­ sity, C lem son, South Carolina: R everse O sm osis an d U ltrafiltration

H. Wade Goodner, P.E.

Consultant, H. W. Goodner Company, Anderson, South Caro­ lina: Risk M anagem ent o f E ngineered D esigns

H. L. Hsieh

Research and Developm ent, Phillips Petroleum Company, B artlesville, Okla­ homa: Rubber, Synthetic

G. J. Kauffman

Technical Safety Technology Leader, The Procter & G amble Company, Cincinnati, Ohio: Safety, Com bustion System s

George R. Kent, BSCE

Senior Engineer (retired), Chemical Engineering Department, Stone and Webster Engineering Corporation, Boston, Massachusetts: R otating E quip­ ment, Troubleshooting

F. Owen Kubias

Consultant, Rocky River, Ohio: Safety M anagem ent

Thomas L. Lange

The Procter & Gamble Company, Cincinnati,

Ohio:

R oasting,

F luidized-B ed

Jean F. Le Page, Ph.D.

Scientific Advisor, Institut Frangais du Petrole, Rueil M alm ai­ son, France: R esidual Refining and P rocessing

Alex Lifson

Senior Research Engineer, Southwest Research Institute, San Antonio,

Texas: Rotating Equipm ent, Vibration Lim its

Jerald Linsley, Ph.D., P.E. Return on Investm ent

Lead Engineer, Sim ulation Scien ces, Inc., H ouston, Texas:

Contributors to Volume 48

J. W. Oldfield

Nickel Developm ent Institute, European Technical Service Centre, Bir­ mingham, United Kingdom: R everse O sm osis and U ltrafiltration, A lloy C hoice

Arthur M. Patureau

Consulting Engineer, M cQueeney, Texas: R esins, H ydrocarbon

Jean C. Plumail, Ph.D.

Sales Manager, Procatalyse, Rueil M alm aison, France: R esidual Refining and P rocessing

Douglas A. Rausch, Ph.D.

Director o f Product Stewardship and Reactive Chem icals Program (retired), The Dow Chemical Company, M idland, Michigan: S afety, O p e r­ ating D iscipline

M. A. Saad

Technical Specialist, E. I. DuPont de Nemours and Company, Riyadh, Saudi Arabia: R everse O sm osis fo r W astewater Treatment

C. I. Sackincer

DuPont Far East, Inc., Bombay, India: R everse O sm osis, Seawater, P res­

sure Effects

Balaji B. Singh, Ph.D.

President, Chemical Market Resources, H ouston, Texas: R e ­ sources, C hem ical M arketing R esearch

Harold R. Simmons, P.E.

Group Leader, Southwest Research Institute, San Antonio, Texas: Rotating Equipm ent, Vibration Lim its

Anthony J. Smalley, Ph.D.

Group Leader, Southwest Research Institute, San Antonio, Texas: Rotating Equipm ent, Vibration Lim its

B. Tod

Nickel Developm ent Institute, European Technical Service Centre, Birmingham, United Kingdom: R everse O sm osis and U ltrafiltration, A lloy C hoice

P. H. Wagner, Ph.D.

Research and Developm ent, Phillips Petroleum Company, Bartles­ ville, Oklahoma: Rubber, Synthetic

Stanley M. Walas, Ph.D.

Department o f Chemical and Petroleum Engineering, U ni­ versity o f Kansas, Lawrence, Kansas: Rules o f Thumb E stim ating

Christianne Weismantel Guy Weismantel

Weismantel International, K ingw ood, Texas: Resins

Weismantel International, K ingwood, Texas: Resins

C. R. Wilder

Research and Developm ent, Phillips Petroleum Company, Bartlesville, Oklahoma: R ubber, Synthetic

Theodore J. Williams, Ph.D.

Professor o f Engineering, Director o f Purdue Laboratory for Applied Research Control, Purdue University, West Lafayette, Indiana: R obots in the C hem ical Industry

Adam Zanker, Ch.E., M.Sc.

Senior Research Engineer, Oil R efineries, Ltd., Haifa, Is­ rael: Residual, Visbreaking C osts; Resistivity, E lectrical, o f S olid an d L iqu id M etals; Rotating Bodies, C ritical Speed

Contents of Volume 48 Contributors to Volume 48 Conversion to SI Units Bringing Costs up to Date Residual Refining and Processing Jean F. Le Page, Jean C. Plumail, and M ichel Davidson

ix xiii xv 1

Residual, Visbreaking Costs Adam Zanker

112

Resins Guy W eismantel and Christianne W eismantel

116

Resins, Hydrocarbon Arthur M. Patureau and Kenneth F. Foley

132

Resistivity, Electrical, of Solid and Liquid Metals Adam Zanker

150

Resources, Chemical Marketing Research Balaji B. Singh

154

Retrofit, Basics Louis J. Cabano

176

Retrofitting, Coal-Fired Boilers C. Thomas Breuer

184

Return on Investment Jerald Linsley

198

Reverse Osmosis, Seawater, Pressure Effects C. I. Sackincer

203

Reverse Osmosis and Ultrafiltration Charles H. G ooding

209

Reverse Osmosis and Ultrafiltration, Alloy Choice B. Tod and J. W. O ldfield

221

Reverse Osmosis for Wastewater Treatment M. A . Saad

228

Rheology and Rheometry N icholas P. Cherem isinoff

235

Risk Management of Engineered Designs H. Wade Goodner

273

Roasting, Fluidized-Bed Stephen R. Cammarn, Thom as L. Lange, and G ale D. Beckett

306

Robots in the Chemical Industry Theodore J. W illiam s

318

Roof Ventilation, Sizing John D . Constance

330

Contents of Volume 48

Rotating Bodies, Critical Speed Adam Zanker

332

Rotating Equipment, Shop Testing Arvind G. Godse

336

Rotating Equipment, Troubleshooting G eorge R. Kent

348

Rotating Equipment, Vibration Limits Alex Lifson, Harold R. Sim m ons, and Anthony J. Sm alley

352

Rubber, Synthetic H. L. H sieh, P. H. Wagner, and C. R. Wilder

368

Rules of Thumb Estimating Stanley M. Walas

414

Rupture Disks G eof Brazier

431

Safety, Combustion Systems G. J. Kauffman

436

Safety Management F. Owen Kubias

442

Safety, Operating Discipline Douglas A. Rausch

472

Conversion to SI Units To convert from

To

Multiply by

acre angstrom are atmosphere bar barrel (42 gallon) Btu (International Steam Table) Btu (mean)

square meter (m 2) meter (m) square meter (m 2)

4.046 x 103 1.0 x 1 0 '10 1.0 x 102 1.013 x 105 1.0 x 105 0.159 1.055 x 103 1.056 x 103 1.054 x 103 3.52 x 10~2 4.187 4.190 4.184 1.333 x 103 98.06 0.457 1.745 x 10"2 1.0 x KT7 1.772 x KT3 3.888 x KT3 3.697 x 10“ 6 1.0 x KT5 1.60 x 10“ 19 1.0 x 10“ 7 2.96 x 10“5 0.305 2.01 x 102 4.404 x 10-3 3.785 x KT3 1.183 x 10”4 6.48 x 10“ 5 1.0 x KT3 7.457 x 102 9.81 x 103 7.46 x 102 50.80 45.36 2.54 x KT2 3.386 x 103 2.49 x 102 9.806 4.45 x 103 0.5144 5.559 x 103

Btu (therm ochem ical) bushel calorie (International Steam Table) calorie (mean) calorie (therm ochem ical) centim eter o f mercury centim eter o f water cubit degree (angle) denier (international) dram (avoirdupois) dram (troy) dram (U .S. fluid) dyne electron volt erg fluid ounce (U .S .) foot furlong gallon (U .S. dry) gallon (U .S. liquid) gill (U .S .) grain gram horsepower horsepower (boiler) horsepower (electric) hundred weight (long) hundred weight (short) inch inch mercury inch water kilogram force kip knot (international) league (British nautical)

newton/square meter (N /m 2) newton/square meter (N /m 2) cubic meter (m 3) joule (J) joule (J) joule (J) cubic meter (m 3) joule (J) joule (J) joule (J) newton/square meter (N /m 2) newton/square meter (N /m 2) meter (m) radian (rad) kilogram/meter (kg/m) kilogram (kg) kilogram (kg) cubic meter (m 3) newton (N) joule (J) joule (J) cubic meter (m 3) meter (m) meter (m) cubic meter (m 3) cubic meter (m 3) cubic meter (m 3) kilogram (kg) kilogram (kg) watt (W) watt (W) watt (W) kilogram (kg) kilogram (kg) meter (m) newton/square meter (N /m 2) newton/square meter (N /m 2) newton (N) newton (N) meter/second (m/s) meter (M)

xiii

xiv

Conversion to SI Units

To con vert from

To

league (statute) light year liter micron

meter meter cubic meter

mil

meter (m) meter (m) meter (m) newton/square meter (N /m 2) newton/square meter (N /m 2) ampere/meter (A/m ) newton (N) kilogram (kg) kilogram (kg) cubic meter (m 3) newton/square meter (N /m 2) cubic meter (m 3) kilogram (kg)

mile (U .S. nautical) m ile (U .S. statute) millibar m illim eter mercury oersted ounce force (avoirdupois) ounce mass (avoirdupois) ounce mass (troy) ounce (U .S. fluid) pascal peck (U .S .) pennyweight pint (U .S. dry) pint (U .S. liquid) poise pound force (avoirdupois) pound mass (avoirdupois) pound mass (troy) poundal quart (U .S. dry) quart (U .S. liquid) rod roentgen second (angle) section slug span stoke ton (long) ton (metric) ton (short, 2 0 0 0 pounds) torr yard

Multiply by (m) (m) meter (m 3) (m)

cubic meter (M 3) cubic meter (m 3) newton second/square meter newton (N) kilogram (kg) kilogram (kg) newton (N) cubic meter (m 3) cubic meter (m 3) meter (m) coulom b/kilogram (c/kg) radian (rad) square meter (m 2) kilogram (kg) meter (m) square meter/second (m 2 /s) kilogram (kg) kilgram (kg) kilogram (kg) newton/square meter (N /m 2) meter (m)

4.83 x 9.46 x

103 1 0 15

0 .0 0 1

1.0 x 10~6 2.54 x KT6 1.852 x 1 0 3 1.609 x 1 0 3 100.0 1.333 x 1 0 2 79.58 0.278 2.835 x 10“ 2 3.11 x KT2 2.96 x 10"5 1.0 8.81 x KT3 1.555 x KT3 5.506 x 10“4 4.732 x 10“4 0.10 4.448 0.4536 0.373 0.138 1.10 x KT3 9.46 x KT4 5.03 2.579 x 10"4 4.85 X KT6 2.59 x 1 0 6 14.59 0.229 1.0 x KT4 1.016 x 1 0 3 1.0 x 103 9.072 x 1 0 2 1.333 x 1 0 2 0.914

Bringing Costs up to Date Cost escalation via inflation bears critically on estimates of plant costs. Historical costs of process plants are updated by means of an escalation factor. Several pub­ lished cost indexes are widely used in the chemical process industries: Nelson Cost Indexes (Oil and Gas J .) quarterly Marshall and Swift (M&S) Equipment Cost Index, updated monthly CE Plant Cost Index (Chemical Engineering), updated monthly ENR Construction Cost Index (Engineering News-Record), updated weekly All of these indexes were developed with various elements such as material availability and labor productivity taken into account. However, the proportion al­ lotted to each element differs with each index. The differences in overall results of each index are due to uneven price changes for each element. In other words, the total escalation derived by each index will vary because different bases are used. The engineer should become familiar with each index and its limitations before using it. Table 1 compares the CE Plant Index with the M&S Equipment Cost Index. TABLE 1

Chem ical Engineering and Marshall and Sw ift Plant and Equipment Cost Indexes since 1950

Year

CE Index

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970

73.9 8 0.4 81.3 84.7 8 6 .1

88.3 9 3 .9 98.5 9 9.7 1 0 1 .8 1 0 2 .0

101.5 1 0 2 .0

102.4 103.3 104.2 107.2 109.7 113.6 119.0 125.7

M&S Index

Year

CE Index

M&S Index

167.9 180.3 180.5 182.5 184.6 190.6 208.8 225.1 229.2 234.5 237.7 237.2

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982

238.5 239.2 241.8 244.9 252.5 262.9 273.1 285 .0 303.3

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992

132.3 137.2 144.1 165.4 182.4 192.1 204.1 218.8 238.7 261.2 2 9 7 .0 314.0 316.9 322.7 325.3 318.4 323.8 342.5 355.4

321.3 332 .0 344.1 398 .4 444 .3 472.1 5 0 5 .4 545.3 5 9 9 .4 65 9 .6 721.3 7 4 5 .6 760.8 7 8 0 .4 789 .6 797.6 813.6

357.6 361.3 358.2

852.0 895.1 915.1 9 3 0 .6 943.1 XV

xvi

Bringing Costs up to Date

TABLE 2

Date 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992

Nelson-Farrar Inflation Petroleum Refinery Construction Indexes since 1946 (1946 = 100) Materials Component

Labor Component

M iscellaneous Equipment

N elson Inflation Index

1 0 0 .0

1 0 0 .0

1 0 0 .0

1 0 0 .0

122.4 139.5 143.6

113.5 128.0 137.1

114.2

149.5 164.0 164.3 172.4 174.6 176.1 190.4

802.8

144.0 152.5 163.1 174.2 183.3 189.6 198.2 208 .6 220 .4 231 .6 241 .9 249 .4 258.8 2 6 8 .4 280.5 2 9 4 .4 310.9 331.3 3 57 .4 391.8 441.1 4 9 9 .9 5 4 5 .6 585.2 6 2 3 .6 6 78.5 729.4 774.1 824.1 8 79 .0 9 5 1 .9 1044.2 1154.2 1234.8 1278.1 1297.6 1330.0 1370.0 1405.6

117.0 132.5 139.7 146.2 157.2 163.6 173.5 179.8 184.2 195.3 205.9 213.9

829.2

1440.4

832.8 832.3 824 .6

1487.7 1533.3 1579.2

797.5 827.5 837.6

201 .9 204.1 207.8 2 07 .6 207.7 205.9 206 .3 209 .6 2 1 2 . 0

216.2 219.7 224.1 234.9 250.5 265.2 277.8 292.3 373.3 4 2 1 .0 4 4 5 .2 4 71.3 516.7 573.1 629.2 6 93 .2 7 0 7 .6 712.4 735.3 7 3 9 .6 7 3 0 .0 7 4 8 .9

1 2 2 .1 1 2 1 .6

126.2 145.0 153.1 158.8 160.7 161.5 180.5 192.1 192.4 196.1 2 0 0 . 0

199.5 198.8 201 .4 206 .8 2 1 1 . 6

220 .9 226.1 228.8 239.3 2 54.3 268.7 2 7 8 .0 291 .4 361.8 415.9 423 .8 4 3 8 .2 474.1 515.4 578.1 6 4 7 .9 622 .8 656.8

2 2 2 .1

228.1 232.7 237 .6 243 .6 252.1 2 6 1 .4 2 7 3 .0 286 .7 304.1 329 .0 364.9 4 0 6 .0 438 .5 4 6 8 .0 522 .7 575.5 615.7 6 5 3 .0 701.1 7 5 6 .6 822.8 903.8

6 6 5 .6 6 7 3 .4 6 8 4 .4 703.1

976 .9 1025.8 1061.0 1074.4 1089.9 1121.5

732.5

1164.5

769.9

1195.9 1225.7 1252.9 1277.3

xvii

Bringing Costs up to Date

Table 2 shows the Nelson-Farrar Inflation Petroleum Refinery Construction In­ dexes since 1946. It is recommended that the CE Index be used for updating total plant costs and the M&S Index or Nelson-Farrar Index for updating equipment costs. The Nelson-Farrar Indexes are better suited for petroleum refinery materi­ als, labor, equipment, and general refinery inflation.

Since CB = CA(BIA)n

(1)

Here, A = the size of units for which the cost is known, expressed in terms of capacity, throughput, or volume; B = the size of unit for which a cost is required, expressed in the units of A ,n = 0.6 (i.e., the six-tenths exponent); CA = actual cost of unit A ; and CB = the cost of B being sought for the same time period as cost CA. To approximate a current cost, multiply the old cost by the ratio of the current index value to the index at the date of the old cost: Cb ~ CAIB/IA

(2)

Here, CA = old cost; IB = current index value; and IA = index value at the date of old cost. Combining Eqs. (1) and (2), CB = CA(BIA)n(lBIIA)

(3)

For example, if the total investment cost of plant A was $25,000,000 for 200million-lb/yr capacity in 1974, find the cost of plant B at a throughput of 300 mil­ lion lb/yr on the same basis for 1986. Let the sizing exponent, n, be equal to 0.6. From Table 1, the CE Index for 1986 was 318.4, and for 1974 it was 165.4. Via Eq. (3), CB = CA(B/A)n(IB/IA) = 25.0(300/200)° 6(318.4/165.4) = $61,200,000 JOHN J. McKETTA

Residual Refining and Processing Introduction A large proportion of residues from distillation do not receive refining treatment. These residues are used in formulations for heavy fuel oils used in industry, es­ pecially for producing electrical power. A chronological review of the processing undergone by the bottom of the bar­ rel reveals that some of the processes are almost as old as the refining industry itself. Delayed coking aims to produce coke for electrodes, deasphalting by pro­ pane to produce bright stocks, a heavy base stock for lubricating oil, and residue blowing to produce bitumens. Around 1960 came the development of viscosity breaking or visbreaking, which today has invaded a large proportion of all refineries. The sending of a larger and larger fraction of vacuum distillates for catalytic cracking and then hy­ drocracking introduced problems of viscosity and flow for heavy fuel oils that contained increasing amounts of vacuum residues. Visbreaking enabled the vis­ cosity of residues to be adjusted while producing nearly 20% atmospheric distil­ lates, making the entire process economically viable. In the 1970s, regulations in some countries forced users of fuel oils to limit pollutant emissions at industrial sites. There were two solutions: eliminate sulfur (and nitrogen) compounds from combustion fumes, or practice the catalytic re­ fining of residues in the presence of hydrogen. Both solutions were applied. This gave rise to a wave of catalytic hydrorefining units for atmospheric residues in most cases, with the essential aim of hydrodesulfurization with a minimum of conversion. The countries involved were mainly Japan, the United States, and Taiwan. The countries in the Persian Gulf sought to export their heavy fuels oils at a reasonable price to countries where regulations required a prior reduction in sulfur content. The 1978 oil crisis gave rise to a momentary lack of crude oil and a rapid rise in its cost. Many countries then tried to diversify their resources and energy sup­ plies and to cut back on energy consumption. These energy savings and recourse to natural gas, heavy oils, coal, and atomic energy brought about a drastic reduc­ tion in the consumption of petroleum products, especially heavy fuel oils. The result was a certain disequilibrium in the market for petroleum products and the more or less accentuated need, depending on the country, to convert part of the residues into light distillates for use in transportation. As shown in Table 1, this trend toward converting the bottom of the barrel should continue on a world­ wide level in the decades to come. In the face of this situation, new thermal and catalytic conversion processes have come into being. New process flows have been designed to diminish the pro­ duction of heavy fuel oils while increasing the flexibility of refineries at the level of crude oil supplies. Heavy oils, for which world reserves are of about the same order of magnitude as those of conventional crudes, deserve particular consideration. Outside Ven­ ezuela and Canada, they are little exploited today, but they open interesting per­ spectives for the future. Figure 1 shows how these heavy oils are situated in relation to conventional oil on a curve giving the viscosity at 20°C as a function 1

Residual Refining and Processing

TABLE 1

World Petroleum Market Evolution (10 6 t) U ses

1980

1985

1. Transportation 2. Fuelsa (developing countries) 3. Petrochemistry (LNG )b 4. Others" Total 5. % o f the market with no available substitutes at presentd

1,160 1,545 (230) 155 (2 0 ) 140 3,0 0 0

1,280 (250) 165 (30) 130

49

1 ,2 0 0

2,775 54

2 0 0 0

1 ,6 5 0 -1 ,8 5 0 9 2 5 -1 ,1 0 0 ( 3 5 0 -4 5 0 ) 2 7 5 -3 5 0 (5 0 -7 0 ) 1 5 0 -2 0 0 3 ,0 0 0 -3 ,5 0 0 69

“Refinery consumption included. bLiquids from natural gas. cBitumen, oils, paraffins, and solvents. d(TotaI - Fuels)/Total. Source: IFP Economics Department, 1986.

of the fraction that distills below 350°C. In this classification, a heavy oil is char­ acterized at 20°C by a viscosity higher than or equal to 20 cSt and a fraction dis­ tilling below 350°C lower than 0.25. At the same time, these heavy oils generally have a density higher than 950 kg/m3, and it is not possible to transport them by pipeline out of the production site in this state. To enable transport by pipeline and then processing by existing refineries, they must be converted, and the processes used are the same as those for the conversion of residues from conventional crudes. This can be seen quite clearly by comparing the characteristics of these two types of products, which have many similarities, as shown in Table 2. The viscosities, specific gravities, asphaltene content, Conradson carbon content, sul­ fur content, nitrogen content, and metal content of various typical heavy oils are of the same order of magnitude as those of the Aramco vacuum residue chosen as a reference. In applying the different processes and flows described here, we refer indis­ criminately, other than exceptions, to vacuum residues, atmospheric residues or topped heavy oils. As shown in Table 3, the technical solutions that have been proposed for re­ fining and converting distillation residues can be grouped in two categories: ( 1) treatment routes with carbon rejection, and (2) treatment routes without carbon rejection. The different processes for each of these categories will be examined successively. The last section is devoted to examining and comparing some pro­ cess arrangements belonging indiscriminately to one or another category. These arrangements have been selected because they have a particular technical or eco­ nomic advantage for processing a given residue or heavy oil within a given re­ fining context.

Processing Routes with Carbon Rejection In a refining process, these routes preferably apply to vacuum distillation resi­ dues. Such residues are a mixture of oils, resins, and asphaltenes, which are

3

Residual Refining and Processing

ELPAO GRENADE

BOSCAN

HEAVY CRUDES

MELONES

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

105 COLD LAKE B

10 *

^ROSPOMARE >LLOYDMINSTER ►CYRUS

103

SOUEDIE ARABIAN HEAVY KHAFJI MANDJI NIGERIA GULF NIGERIA MEDIUM MORGAN GASH SARAN ARABE MEDIUM ROMASHKINO OMAN OURAL ELALAM EN BASRAH AGHAJARI BREGA NIGERIA LIGHT ZUEITINA QATAR MARIN MURBAN UMMSHAIF ROSTAM ZAKHUM QATAR ARZEW BRUNEI HASSI MESSAOUD HASSI R MEL

I EOCENE

102

*

BUZURGAN

\

4

7 6 KUWAITV

CONVENTIONAL CRUDE

“T * 108

10

EKOFISK _ 1 1 \ 12 13 LIGHT ARABIAN % 14«

KIRKUK

X 'W

19 20

17

“ 2M 34 Yield in 350°C' (wt% )

0

I

I

i

10

20

30

FIG. 1

40

50

60

70

80

Viscosity of crudes as a function o f distilled fraction at 350°C.

chemical entities whose definition has remained highly empirical in practice be­ cause it is based simply on the nature of the solvents used to precipitate succes­ sively asphaltenes and then resins from their natural environment. The mixture of oil and resins extracted by the solvent is also known by the term “ maltene phase.” Figure 2a is a simplified illustration of the analytic definition of the different con­ stituents of a vacuum residue. The amounts of solvent (heptane, pentane, pro­ pane, and benzene) involved in these successive analyses are 30 times greater than those of the residue to be analyzed. Table 4 gives a detailed analysis of an Arabian Heavy vacuum residue and its constituents: oil, resins, and asphaltenes. We see that resins, especially asphaltenes, are the most hydrogen poor constituents and richest in carbon, sulfur, nitrogen, nickel, and vanadium. The difficulties encoun­ tered in the catalytic processing of these heavy feedstocks are always linked to the

4

Residual Refining and Processing

TABLE 2

Typical Properties o f Residues and Heavy O ils Venezuela

El Pao Boscan Density, °API 15/4°C V iscosity, cSt At 20°C (calculated)

7.1

10.5 0.995 1 2 0 ,0 0 0

At 100°C Pour point, °C Total sulfur, wt% Total nitrogen, wt% A sphaltenes, wt% Conradson carbon, wt% M etals V, ppm N i, ppm Wt% At At Wt%

(Orinoco)

135 17 5 .2 0 .6 5 13 15 1 ,2 0 0

150

D istilled 200°C 360°C on crude

4.8 18

Canada

France

M iddle East

Cold Lake B

Grenade

Vacuum Residue

sur L’Adour

Aramco

9.35 1.004

1.008

1 0 - 1 2

1 .0 2 0

1

,0

0 0 , 0 0 0

4 0 ,0 0 0

432 27 3.7 0.19

1 0 0

6 0 0 ,0 0 0 (280) 2 1

4 .7

8 . 2

0 . 2

8

15 375 80

15

1 1 .2

1 2 .6

15

240 70

0 . 6

2 . 0

14

19

45 5 5 .9 15.7

592 29 3 .9 0.27 6.45 16.4 6 6

15 —

19

presence of these metals as well as to the low H/C ratio of asphaltene and resin compounds. The processes used for the more or less selective elimination of these unde­ sirable compounds can be grouped into two sorts: ( 1) carbon rejection processes by solvent treatment, such as decarbonizing, demetallizing, deresinizing, and, more generally, deasphalting processes; in what follows we use the phrase “ deasphalting process” to designate these different operations; and (2) carbon rejection processes by heat treatment, of which the most widespread are coking processes. Besides this carbon (asphalt, coke, or pitch) rejection, in either case a residual oil or distillates are produced, which lend themselves better to subsequent catalytic conversion treatments, such as catalytic cracking or hydrocracking.

TABLE 3

Technical Solutions

Refining with Carbon Rejection Coking + Liquid product hydrotreatment + Hydrocracking or catalytic cracking Heavy solvent deasphalting + DAO hydrotreatment + Catalytic cracking (or hydrocracking)

Refining Without Carbon Rejection Thermal processes Visbreaking Hydrovisbreaking Solvent donor Slurry processes Catalytic processes on grain catalysts: H DM , H D S, HDV, H DAs, HDCC on fixed, m oving, and ebullating beds

5

Residual Refining and Processing

C 6 ASPHALTIC PHASE (CsAs) C 5 ASPHALTIC PHASE (CsAs) C 4 ASPHALTIC PHASE (C4 A 5 ) FIG. 2

Analytic definition and nomenclature o f the various hydrocarbon fractions contained in residues.

Deasphalting

The conventional deasphalting process has already been described in detail in this encyclopedia [1]. The same is true of its specific application to the demetalliza­ tion of residues in the case of the Demex variant proposed by Universal Oil Prod­ ucts (UOP) [2]. Therefore, we limit ourselves here to a review of the general

TABLE 4

VR Asphaltenes Resins

Analysis o f Vacuum Residue o f Arabian Heavy Resins

Carbon (Wt%)

Hydrogen (Wt%)

Sulfur (Wt%)

Nitrogen (Wt%)

Vanadium (ppm)

Nickel (ppm)

Asphaltenes (Wt%)

(Wt%)

83.3 82.3 8 2 .9

9 .8 7.3 9 .2

4 .9

0 .45 1.15

127 564

13.5

2 3 .9

0 . 6 8

278

42 195 73

7 .0 5.1









Residual Refining and Processing

conclusions issuing from these monographs so that to the discussion is focused on the following two aspects: ( 1) innovations and improvements made in the process in recent years, and (2) integration of the process in refining to minimize the pro­ duction of heavy fuel oils. D easphalting Process Since the advent of propane deasphalting for the production of lube base stocks, the basic principles of the process have remained the same: 3-10 vol solvent (pro­ pane, butane, pentane, and mixture of alcohols) is added to 1 vol residue under optimized conditions for each solvent, and the oil phase containing most of the solvent is separated from an asphalt phase containing the remainder of the sol­ vent. The solvent is recovered under conditions that are as economical as possi­ ble. The simplified process design in Fig. 3 is representative of all these processes. Moreover, the operating conditions are very similar among processes and fall within the ranges shown in Table 5 when the solvent is a hydrocarbon. However, deasphalting remains an evolving process that has been the target of improvements, even important innovations, in recent decades, especially when the goal is to reduce the production of heavy fuel oils. This evolution affects the deasphalting stage itself, the recovery of the solvent, and the production and con­ ditioning of increasingly harder asphalts with an R&B (Ring and Ball method) softening point to 210°C. Deasphalting Step. Improvement efforts have mainly been focused on the use of the heaviest solvents that give the highest yields of deasphalted oil (butanepentane mixtures, pentane, and light gasoline).

FIG. 3

Deasphalting process: So = solvent; E = exchanger; H = heater; F = flash tower; S = stripping tower.

7

Residual Refining and Processing

TABLE 5

Operating Conditions o f Deasphalting as Function o f Solvent

Operating Conditions

Propane

Butane

Pentane

Temperature range, °C

6 0 -9 0 3 .5 - 4 .5 6 -9

100-130 4 -5 4 -7

1 7 0 -210 4 -5 3 -5

Pressure range, MPa Solvent ratio

New designs have opened up possibilities of processing capacities of 2 x 1061/ year by a single unit. Vertical baffle extractors or rotating disk contactors can reach production capacities of only 6 x 105—7 x 105 t/year, which are sufficient for a lube oil unit but generally insufficient when vacuum residue conversion or the processing of heavy oils in the field is considered. The selectivity of operations has been considerably improved. The optimiza­ tion of operating conditions has led to an increase in yields of deasphalted oil while producing, even with heavy solvents, a good quality of deasphalted oil (DAO). The residual asphaltene (As) contents of such DAO are quite low (AsC7 < 0.05%, AsC5 < 0.3%) so that they can be hydrotreated at medium pressure with economically acceptable cycle times. These improvements are related to better understanding and management of the three physicochemical operating units involved in the deasphalting process: ( 1) flocculation and precipitation of the asphalt phase, (2) asphalt decanting, and (3) asphalt washing. Precipitation o f the Asphalt Phase. In crude oils and residues, asphaltenes may be present in different forms, that is, isolated molecules solvated by resins and pericondensed aromatic hydrocarbons, clusters of a few molecules, or colloidal micelles resulting from the agglomeration of asphaltene clusters and their pro­ cession of resins. These different asphaltene entities are in metastable equilibrium with the sur­ rounding maltene environment. The gradual addition of a light paraffin ends by breaking up this equilibrium with the precipitation of an asphalt phase. Under in­ dustrial deasphalting conditions (solvent content > 2, temperature close to the critical temperature of the solvent used), if precipitation is not immediate it is at least very fast. The operational parameters and variables with an appreciable in­ fluence are the nature of the solvent, the solvent content, and the temperature. The lighter the solvent, the higher is the yield of the asphalt phase and, hence, the better the quality of the oil phase extracted. For a given solvent and with operating conditions suitable for this solvent, an increase in the solvent-oil ratio generally leads to an increase in the DAO yield and quality. For a given solvent and a given solvent-oil ratio, the yield of oil diminishes but the quality of the oil is improved when the temperature increases. Variations in the oil-phase yield are all the more sensitive as the critical temperature of the solvent-oil mixture is approached. Pressure is not used as an operational variable in the deasphalting stage. Care is generally taken to see that it is higher than the critical pressure of the solvent. Decanting o f the Asphalt Phase. The decanting of the asphalt-phase particles in suspension in the solvent-oil environment is governed by Stokes’ law:

Residual Refining and Processing

2 r2 V = 9 ^ (do - d)g where V is the velocity of particles falling, cm/s; r is the particle diameter, cm; r\ is the viscosity of the solvent-oil phase, poise; g is the acceleration of gravity; and da and d are the respective specific gravities of the asphalt phase and the solventoil (or solvent-maltene) phase. In practice, the decanter must be designed so that the smallest particles can be decanted suitably. This implies the following corollaries: (1) the falling velocity of the smallest particles must be greater than the ascending velocity of the surround­ ing solvent-oil mixture; (2) the residence time of the solvent-oil-asphalt phase mixture must be long enough for there to be a clear-cut separation of the asphalt phase, including the smallest particles. In agreement with these considerations, if the size of the smallest particles is about 10 \im, their rate of decantation is about 1 cm/s. An increase in the solvent content and, to a lesser degree, in the temperature speeds decanting, as per Stokes’ law. Washing o f the Asphalt Phase. The asphalt phase occurs in the form of an emul­ sion in the solvent-oil mixture. It thus must be washed. This washing by a coun­ tercurrent of pure solvent has the following two roles: ( 1) to replace the solvent-oil environment by a pure solvent environment, and (2) to selectively repeptize the lightest resins to recycle them at the inlet of the decanter. For a feedstock and a given solvent, the operational parameters and variables governing this unit washing operation are as follows: ( 1) the operation is more effective when the solvent-asphalt phase ratio is high; (2) the amounts of repep­ tized resins increase as the temperature is lowered; and (3) with the washing zone must be carefully designed and suitably equipped to ensure proper contact be­ tween the asphalt phase and the solvent. In practice, these three unit operations— precipitations, decanting, and wash­ ing— are closely interdependent. The solvent used in the washing stage of the as­ phalt phase is always recycled with the dissolved oil and the repeptized resins at the level of the precipitation stage or at the inlet of the decanting zone. Figure 4 shows simplified flows for two variants of deasphalting. Figure 4a is more advantageous for deasphalting by the lightest solvents (C3 and C3 + C4) when the volume of the asphalt phase is of the same order of mag­ nitude as the volume of the oil phase. Part of the solvent (S, = 30-50%) is effectively mixed with the feedstock at temperature T0. Precipitation starts im­ mediately and continues inside the vertical columns above the washing zone, where the second fraction of the solvent rises (S2 = 70-50%), loaded with oil and resins extracted in the washing zone. To improve the quality of the DAO, an additional makeup of calories can be supplied at the outlet of the column (7^ < T2 < T3). Figure 4b is preferred for the deasphalting of residues and heavy oils by the heaviest solvents (C4 + C5, C5, and light gasoline), especially when there are high production capacities (1-2 Mt) [3]. Indeed, the need to have a maximum ascending velocity and a sufficiently long residence time, in keeping with Stokes’ law, means that feedstock capacities of 6 x 105 t/year in a vertical column cannot

9

Residual Refining and Processing

DAO

+ s*!

a

DAO + S’i

b

be exceeded. Above this capacity, the column becomes too large in diameter, which may cause problems of handling and transportation. For greater capacities, the operation must be done in several units, with the resulting multiplication of equipment. When the capacity is large and the solvent heavy, it is thus prefer­ able to use Fig. 4b, in which the horizontal decanter, with the same volume, solves the problems of the ascending velocity of the solvent-oil phase in the ver­ tical system. With heavy solvents, if S/R, T ] and T2 are optimized, the solvent-oil mixture need not be reheated at the outlet of the decanter to improve the quality of the oil.

Residual Refining and Processing

FIG. 5

Solvent recovery.

In solvent recovery, the pursuit of energy conservation has considerably mod­ ified solvent recovery in the last decade, at least from the solvent-oil mixture, where most of the solvent is entrained. Currently, the solvent recovery technique under supercritical conditions, as was claimed by the ROSE [4] (residuum oil su­ percritical extraction) process 20 years ago, has become both generalized and diversified. If we look at Fig. 5 [5], which illustrates the general principle of the method, we can see that solvent separation occurs in the supercritical zone, either at a pressure higher than the critical pressure of the solvent or at a pressure higher than the critical pressure of the DAO-solvent mixture, depending on the process. In both cases the energy cost of the vaporization operation is considerably reduced. It remains to determine the operating conditions that fulfill the following condi­ tions: ( 1) low oil content (21

operating conditions recommended for each, the characteristics (specific gravity, viscosity, and surface tension) and corresponding solvent power diminish from C5 to C4 and then to C3. The effect of temperature is associated with increasingly rapid variations in these same physical characteristics (including solvent power), as we approach the critical temperatures of the solvent and then of the solvent-oil mixture. At 150°C, for example, with a given vacuum residue, pentane leads to the rejection of a very hard asphalt with a good oil yield. At 210°C, asphaltenes and resins are rejected by the solvent-oil phase. Above this temperature the DAO yield decreases rapidly, and around 230°C, when we are operating in the supercritical range or even in the retrograde condensation zone, there is complete rejection of the oil phase. At a given temperature between 210 and 230°C, pentane can thus have environment properties similar to those of propane under propane deasphalting conditions, with comparable oil yields. The increase in the amount of solvent brings about an increase in DAO yield and an improvement in the quality of the oil. The viscosity and specific gravity of the DAO diminish, as do the contents of various impurities, such as S, N, Ni, AsC7, A sC5, and Conradson carbon. The temperature of the solvent at the inlet to the washing zone also has an appreciable influence on both the yield and quality of DAO. The lower the wash­ ing temperature, the more the resins are repeptized and the more the yield in­ creases. The existence of a positive temperature gradient between the washing and decanting zones causes a reflux of resins, which improves the selectivity of the operation. These effects of operating temperatures are illustrated by the data in Table 7. Figure 6 shows how the impurity contents of DAO vary (AsC5, Conradson car­ bon, and metals) and how the softening point of the asphalt varies as a function

Residual Refining and Processing

TABLE 7

13

Temperature Effects in C 5 Deasphalting o f a Rospomare Long-R esidue. Products Characteristics 3 B

A

1

T - 20

1

o

7 -4 0

T Asphalt washing

o

T° C

T Settling outlet

C

T - 60

DAO, yield wt%

5 4.7

5 6 .9

Specific gravity, 100°C, cSt Concarbon, wt%

0.995 38

0 .9 9 9 47 8 .4

C 5 asphalt, wt% C 7 asphalt, wt% S, wt% N i, ppm

0 .1 2

0 . 6

0.05 6 .0 4 7

0 .0 6

7.5

V, ppm Asphalt characteristics R& B, °C Specific gravity Y ield , wt% S, wt% Ni + V, ppm

50 159 1.143 4 5.3 9.95 675

6 . 2 0

58.5 1 .0 0 0

49 8.5 0.73 0 .1 1

6 .2 4

8

10

59

61

167 1.145 43.1 9.9 4 693

Feed

1 0 0

1.055 4180 28 .0 34.1 25 .2 8 .1

76 274

169 1.149 4 1 .5 1 .0 0 2

712

aAsvahl results. (Association for the Valorization of Heavy Oils, joint venture of ELF, Total, and Institut Frangais du Petrole, 1FP.)

of the DAO yield. The crude oil processed in this case is Safaniyah Arabian Heavy, and the variables involved are both the nature of the solvent and the op­ erating conditions.

Products: Uses and Further Processing We successively examine the two types of products of the deasphalting operation, deasphalted oils and pitches. Deasphalted Oils. The deasphalting stage is merely a stage in the preparation of feedstocks for subsequent conversion processing, such as catalytic cracking or hydrocracking. All deasphalted oils, moreover, are not always suitable for feeding to such conversion units, even when mixed with vacuum gas oil. A deasphalting unit must then be considered along with a conversion unit and an intermediate hydrorefining stage, as illustrated very schematically in Fig. 7. In catalytic cracking, the essential characteristics of the feedstocks taken into consideration are the amounts of nitrogen, metals, and Conradson carbon they contain. It is believed that the basic nitrogen content must be less than 2,000 ppm so that the acid function of the catalyst is not too inhibited. The Conradson carbon content is lower than 2% for conventional units but may be higher and even ap­ proach 7% if the unit is equipped with devices in the riser for reducing coke for­ mation and, in the regenerator, to burn more coke without damaging the catalyst. As for the metals that are irreversibly deposited on the catalytic cracking catalyst, with a tendency to destroy it, it is considered that above 20 ppm in the feedstock

DAO C5 INSOLUBLES Wt/%

DAO CONCARBON Wt/%

35

45

55

FIG. 6

75 □ Ca ■C5 • C4 + C5

Safaniyah vacuum residue deasphalting (Asvahl results).

65

< co 00 O Z

Residual Refining and Processing

15

Residual Refining and Processing

FIG. 7

Hydrorefining of DAO before cracking

catalyst consumption and the reduction in yields of liquid products are such that it becomes economically feasible to perform prior hydrorefining. The need for intermediate hydrorefining and the severity of this operation thus depend directly on the origin of the crude oil and, especially, on the deasphalting solvent. For example, if we refer to Arabian Light crude and to the characteristics of the vacuum gas oil (VGO) and of VGO + DAO mixtures given in Table 8, we arrive at the following conclusions: (1) the mixture of VGO with propane deas-

TABLE 8

Characteristics o f VGO, DAO, and Their Mixtures (Arabian Light Crude Oil)

C3

c4

c5

DAO

DAO

DAO

15.4 0 .9 5 9

18.7 0 .9 7 4 107

2 .4

9 .9 0.933 — 2 .8 0

0 . 1 0

0 .1 2

VGO 3 8 0 -5 1 0 ° Y ield on crude, wt% D ensity at 15°C, g/cm 3 V iscosity at 100°C, cSt Sulfur, wt% Nitrogen, wt% Basic nitrogen, wt% C5, asphalt, wt% C7, asphalt, wt% N i, ppm V, ppm Concarbon, wt% Saturates, wt% Arom atics, wt% Resins, wt%

19 0 .9 2 9 6



0.028

6 6

3.3 0.14 0.035

0 . 2

0 .1

0 .1

< 0 .0 5

< 0 .0 5

< 0 .0 5

< 1

< 1

< 1

< 1

0 .85 5 2 .0 4 8 .0 0 . 2

2.5 4 1 .4 55.8 2 .8

2

4 .2 0 . 2

0.045 0.3 < 0 .0 5

VGO + C 3 DAO

VGO + C 4 DAO

VGO + C5DAO

2 8.9 0 .9 3 0 — 2 .5 4

3 4.4 0.9 4 2 30 2 .8 0

3 7.7 0.951 56 3.3 0.15 — 0 .3

0 .1 1



0 . 1 2



0 . 2

0 . 2

< 0 .0 5

< 0 .0 5

< 0 .0 5

2

< 1

1

1

9

< 1

1

5

3 5 .6 24.1 7 1 .6

72

4 .3

8

11 2 0

1.4 4 8 .4 5 0 .7 1 .0

3.0 3 9 .0 5 8 .6 2 . 0

6 . 0

36.1 59.9 4.1

Residual Refining and Processing

phalted oil can be processed without major problems in conventional catalyticcracking units; (2) the mixture of VGO and C4 DAO is at the limit of the specifications required in conventional units but can be processed without prob­ lems in units suited for processing heavy feedstocks; and (3) the mixture of VGO and C5 DAO cannot be processed in conventional units and is at the limit of the specifications required for units suited for processing heavy feedstocks. An intermediate hydrorefining stage for DAO in some cases thus proves in­ dispensable so that catalytic cracking can be done without difficulty. What must be borne in mind, nonetheless, is that in all cases an intermediate hydrotreatment improves catalytic cracking yields of noble products and diminishes coke produc­ tion, whether VGO or pure C5 DAO, or both is used [7]. Table 9, for example, can be used to compare the results obtained in catalytic cracking with an Arabian Light C4 DAO before and after hydrorefining. We see not only an increase in yields but also an improvement in the quality of the products. Figure 8 shows the influence of hydrotreatment on coke production with varying conversions. In practice, there are no major problems in the hydrotreatment of deasphalted oils, and this can be done at hydrogen partial pressures between 8 and 10 MPa provided that the deasphalting operation was done correctly. C5 DAO are of course the most difficult to hydrotreat, especially when they come from heavy oils with high metal contents. In this case prior demetallization on a specific catalyst may be recommended. Likewise, in the hydrotreatment stage it may also be advantageous to do a cer­ tain amount of conversion to middle distillates, especially to diesel oil, particuTABLE 9

Catalytic Cracking o f C 4 DAO and C 4 DAO Hydrotreated Arabian Light C 4 DAO

Y ield on crude, wt%, 375°C + Sulfur, wt% Total nitrogen, wt% Basic nitrogen, wt% C 5 asphalt C 7 asphalt N i, ppm V, ppm Concarbon, wt% Saturates Aromatics Resins Catalyst-oil ratio Conversion, wt% Coke, wt% Fuel gas + H 2 S, wt%

10.7

9 .9

2 . 8

0 .1

0 .14 0.035

0 .0 6

0 .1

< 0 .0 5 < 2

< 2 .4 5 .6 24.1 7 1 .6 4 .3 6

74 8 . 2

4 .2

LPG Total gasoline

14.2 47 .3

RON clear

90.3 18.3 0 .9 6 4 7.8 5 .4

LCO dL C O Residue (350°C + ) S residue, wt%

C 4 DAO Hydrotreated

0 . 0 1 2

< 0 .1 < 0 .0 5 < 0 .1 < 0 .5 1 .2

35 63 2 6

81 6

1.9 14.7 5 7 .9 9 1 .4 14.4 0.975 5 .2 0 .5

17

Residual Refining and Processing

s? 5

Q -I

7

UJ

>-

LU

o

u

5

3

65

70

75

CONVERSION. Wt% FIG. 8

Comparison of FCC feeds [7].

larly when using a C5 DAO. Indeed, during this hydrotreatment, nearly 20% diesel oil can be obtained with a cetane number close to 40, whereas the light cycle oil issuing from subsequent catalytic cracking is at only about 20. Deasphalted oils may also be fed to a hydrocracking unit, if a refinery needs increased production of high-quality middle distillates for the market. Before turning to the hydrocracking stage, it is recommended that DAO be hydrorefined to give it characteristics (except molecular weight) similar to those of the VGO with which it will probably be mixed in the hydrocracking stage itself. This is particularly true for C5 DAO with high metal contents, AsC7, Conradson carbon, and basic nitrogen. C3 asphalts can be used in formulations for road bitumens or as constituents of industrial fuel oils. C4 asphalts are poorer in resins and lend themselves less well to bitumen formulations, and they are used as constituents of industrial fuel oils after being mixed with less viscous constituents such as light cycle oil (LCO), heavy cycle oil (HCO), decanted oils, and vacuum residues). C5 asphalts can either be used in the composition of liquid fuel oils after being mixed with other constituents or they can be used as solid fuels, such as coke or coal, in electrical power plants, lime kilns, or cement furnaces. C3 asphalts and even C4 asphalts may be subjected to a visbreaking treatment to increase refinery yields of light and middle distillates. Fuel oil yields of a re­ finery are reduced, but at the expense of the quality of this fuel oil.

Asphalt Pitches.

Residual Refining and Processing

Asphalts can also be gasified by partial oxidation with the aim of producing syngas or hydrogen for use in various refinery hydrotreatment units. Likewise, asphalts can be subjected to a coking operation to increase the distillate yields of the refinery. Figure 9 diagrams the different uses that can be considered for various types of asphalts. Carbon Rejecting Processes by Thermal Treatment

Deasphalting processes isolate heavy molecules, resins, and asphaltenes without modifying their chemical structure. Heat treatments, on the contrary, are based on a deep modification of this structure. By the end of the operation, an appreciable

____ r ---------------------C3 - A g

^

- Ag

----- >

---

Component for road bitumen

Feedstock to > visbreaking

- A C3 s CA - A 4 s - A C5 s

-------->

F e e d sto c k

to

-------->

---- >

c o k in g

Component for heavy liquid

fuels

C - A 3 s 4 - As 5

s

5

s

Use a s a

FIG. 9

s o lid

fu e l

Possible uses of various asphalts.

Residual Refining and Processing

TABLE 10

19

Carbon R ejecting Processes by Thermal Treatment11

Process and Trade Nam e

Licensers

Family

Status

Delayed coking [ 8 ]

FW; UOP; Lummus; Kellog;

C

I

ACTIV (asphalt coking treatment in

Conoco; KOA oil; Sinopec NIPPON M ining C o ., Ltd.

C

D

Osaka Gas C o ., Ltd.

c c

D

process) [14] Fluid coking [8 ] F lexicoking (and dual flexicok ing) [ 8 ] LR coking (Lurgi-Ruhrgas coking)

Exxon R&E Co. Exxon R&E Co. Lurgi

F F F

I I D (oil) I (coal)

[15] ART process (asphalt residue treating)

M W K ellog Co.

F

I

[16] KK process (coke fluidized-bed crack­

MITI (Japan)

F

D

Kobe steel, KOA O il, Idemitsu Kosan Nippon M ining Kurishima Engineering & A1

F

D

F F

D D

F F (R)

D I

MIW K ellog & Philipps Petr. Co.

F (R )

I

Stone & Webster Total France & IFP

F (R )

I I

vacuum) [2 0 a] Eureka [12, 13] CHERRY P (comprehensive heavy

Kureha-Chiyoda

I

ends reforming refinery petroleum

ing) [17] KKI process [18] HOT (heavy oil treating process) [19] ACC (allosite catalytic cracking pro­ cess) [2 0 b] Dynacracking process [21] RCC (reduced crude oil conversion process) [ 2 2 ] HOC (heavy oil cracking process) [23] S& W FCC process [24] R-2R (one riser, two regenerators)

Hydrocarbon Research, Inc. U D P Process D ivision, Inc.

F (R)

[24] aC = chamber; F = fluid; I = industrial; D = development; R = riser.

fraction of these molecules has become separated in the form of a carbon phase resulting from a series of successive reactions: dealkylation, dehydrogenation, and condensation. Different processes based on such heat treatment are classified in Table 10 and actually fall into two categories. In the first category, the treatment is purely by heat. The best known and most widely used of these processes are the coking processes, which have already been described in detail in this encyclopedia [8]. Other processes (Eureka, ACTIV, KK, KKI, Dynacracking, and so on) have merely been the subject of demonstra­ tion tests and are of limited industrial use. The second category includes residue catalytic cracking. The field of appli­ cation of such processes is relatively limited since most can process only feeds with a Conradson carbon content less than 10% by weight. In addition to the purely thermal transformation of the nonvolatilizable heavy fraction, there is also catalytic cracking of the volatilized fraction.

Residual Refining and Processing

There is no question of providing a detailed description of these different con­ version processes for heavy feeds. We limit ourselves to noting the great similar­ ities that lead them to be classified under the same heading, followed by the particularities that give each process its originality.

D ifferent Processes: Points in C om m on and Particularities Points in Common. All these processes are implemented in a relative low pressure range (0.05-2 MPa) and in a relatively high temperature range (430850°C). The liquid and gaseous fractions issuing from the cracking of heavy frac­ tions are rich in aromatic hydrocarbons and olefins. All these processes are endothermic. The transformations involved proceed by thermal activation and take place according to a free radical mechanism. Well before 400°C asphaltenes and resins can effectively give rise to free rad­ icals, which in turn intervene in various ways, as diagrammed in Table 11. These radicals, which are very complex in molecular structure, can ( 1) initiate their own cracking; (2) activate the molecules of the surrounding substrate; (3) react be­ tween themselves to give rise to new asphaltenes of higher molecular weight, which in turn may be activated and condensed, a free radical condensation that causes coke to be formed; (4) capture radicals issuing from the initiation or crack­ ing of lighter molecules from the surrounding substrate and slow the transforma­ tion of this substrate; and (5) be fixed by hydrogen issuing from a molecule with hydrogen donor power (naphtheno-aromatic molecules). Likewise, asphaltenes and resins can also play the role of donor solvent, either directly via hydrogen from naphthene rings coupled with their aromatic pericondensed rings or indirectly via hydrogen sulfide and mercaptans issuing from cracking of the C-S bonds included in lateral chains grafted onto aromatic nuclei and heterocyclics. This role of donor solvent also participates in slowing the cracking of light molecules by the capture of radicals that initiate or propagate it (secondary cracking). On the other hand, this hydrogen transfer from asphaltenes and resins to other molecules or radicals can only speed their condensation and demixing. For the interpretations that follow, it is useful to refer to a more concrete model of these asphaltenes and resins. From the results obtained with small-angle x-ray spectroscopy (SAXS) and small-angle neutron spectroscopy (SANS), it ap­ pears that these heavy molecules occur in their maltenic medium in a more or less “ open” two-dimensional form, consisting of a few pericondensed naphthenoaromatic rings linked by rather labile bonds involved in aliphatic chains or sulfur bridges [9]. Figure 10 shows a model developed for asphaltenes from Arabian crude [10]. Figure 11 illustrates the changes that can be observed in these mole­ cules by heat treatment [11]. The two-dimensional molecules with a high “ void volume” ( 1) are dealkylated to yield other molecules with about the same mo­ lecular weight but that are less porous (2). These new molecules then condense to larger and more compact molecules (3), which become increasingly difficult to solvate by the neighboring resins and aromatics (4). Above a certain conversion level, the molecules acquire a third dimension (4), which corresponds to the in­ cipient organization of asphaltenic micelles and the appearance of a pitch. De-

Residual Refining and Processing

TABLE 11

21

Possible Reaction M echanism s a heat (Free radicals) initiation

S° Abstraction o f H -------------------------- >



+

A° R° As°

cracking alkanes

S° A° R°

condensation recombination

A — R, As — As As — R, A — S

As°

S° + AH As° + AH R° + R'H

C— C C = C C = C

8 2 .6 145.8 199.6

^

H donor

^ ___ > / \ Bonding Energies (kcal/mol) C— H H— H C — O

9 8 .7 104.2 85.5

SH + A° AsH + A° RH + R'°

C — S -S — S -S — H

65 84 83

C — N C = N C = N

72.8 147 212.6

aS = saturated; A = aromatics; R = resins; As = asphaltenes.

pending on the operating condition is applied, a nematic discotic (disk-shaped) mesophase can be observed in the pitch (5). Further transformation leads to coke. All during these transformations, there is some selectivity in the breaking of the different chemical bonds. Only the most labile are broken, that is, those whose bonding energy is weaker, in keeping with the values given in Table 11, expressed in kilocalories. The majority of the sulfur atoms and all the nitrogen atoms of the heterocyclic type belonging to the resins and asphaltenes are in the carbon reject. Differences and Particularities. The processes listed in Table 10 can be clas­ sified into two different families even with regard to the technology: ( 1) the re­ action takes place in a soaking drum or a coking drum, the delayed coking family; and (2) transformation takes place in a fluidized bed of solids that are either inert (fluid coking) or catalytic (residue catalytic cracking).

Residual Refining and Processing

SULFUR BRIDGE

ALIPHATIC BOND

AROMATIC SHEET

PORPHYRIN FIG. 10

a = Dealkylation y = stacking up

E22 aaa

Structure of asphaltenes [11].

p = AS and R condensation s = appearence of mesophase

perricondensed naphthenics/aromatics alkyl chains FIG. 11

Condensation of asphalts and resins in pyrolysis.

23

Residual Refining and Processing Coking (soaking) drum

Gas

& Fr
125°C). Fluid Coking and Assimilated Processes. Whereas delayed coking pro­ cesses were initially designed to produce quality coke, not distillates, fluid cok­ ing, on the contrary, developed by Exxon 30 years ago, aims to maximize the distillate yield and to minimize the coke yield. Moreover, this coke does not have the qualities required for electrolytic or metallurgical uses, which require coke from delayed coking, the main outlet still being use as solid fuel. Figure 13 is the flow sheet for fluid coking. The fluidized bed is made up of coke particles issuing from the coking operation. Inside the reactor, the preheated feed is sprayed onto the coke particles. The mean temperature of the reactor is between 500 and 550°C. Actually, the real reactor is made up by the droplet of feed sprayed onto the moving coke particle. There is first a flash of the volatile fractions before the cracking and condensation reactions take place, as already mentioned. Carbon is deposited in onionskin form on the supporting particle. The flow of coke particles is then fed to the regenerator. A coke fraction correspond­ ing to about 5% of the carbon in the feed is burned to store the energy required for continuing the transformation. These hot particles are sent back to the reactor. The net coke is withdrawn from the regenerator. Table 13 gives, with reference to fluid coking, a few characteristics of flex­ icoking, LR coking, the ART process, and other processes in the same family. Flexicoking is an extension of fluid coking, to which a steam-air gasification step is added to produce a low heating value gas and subsequently a synthesis gas (dual gasification, as shown in Fig. 14). The coke is recycled to the reactor after having been heated by the low-Btu gas coming from the gasification. The LR coker operates according to the same principle as the fluid coker, but with mixing devices for the fresh feed with the heat carrier (coke). This device is a screw conveyer ensuring excellent dispersion of the feed within the solid as well as the fast flashing of the volatilizable fractions. This system can handle very vis­ cous feeds and even solids. The heat required for the operation is supplied by combustion of part of the coke inside the riser (Fig. 15) The ART process uses riser technology just like that in catalytic cracking. The solid is a specific inert solid (ART psuedo catalyst) on which is deposited the coke, metals, and sediments from the feed. Inside the burner, the coke is mostly converted to C 0 2, and the sulfur contained in the coke is converted to S 0 2 and partly to S 0 3. The flue gases are used to produce high-pressure steam before be­ ing scrubbed to remove SOv and fine particles. The catalyst can handle up to 4% metals (Ni + V). Like the LR coker, the ART process is based on very fast and selective vaporization of the volatilizable fractions.

FIG. 13

Fluid coking family processes.

§

o

Residual Refining and

aIFP-Total results. NA = not available

0 .5 - 3 .5 - 2 0 .5 - 2 0 .5 - 3

5 3 0 -6 0 0 4 8 0 -5 4 0 5 0 0 -5 6 0 5 0 0 -5 6 0

RCC HOC S& W FCC R-2R 0

NA NA NA

5 0 0 -5 5 0 4 2 5 -4 5 0 5 0 0 -5 5 0

.1 -2 3 -4 NA

2 - 1 0

HOT process ACC process D ynacracking

1 5 -2 0 15 -2 0 1 5 -2 0

4 9 0 -5 5 0 4 9 0 -5 5 0 4 9 0 -5 5 0

500 5 0 0 -5 4 0 7 0 0 -8 0 0 5 3 0 -5 8 0

(s)

0

t Range

(°C)

T Range

Flue gas + coke L ow -B T U gas L ow -B T U + synthesis gas Flue gas + coke Flue gas + steam Flue gas + steam Synthesis gas or low -B T U gas Flue gas, steam , H 2 Flue gas + steam M edium -BTU gas + Synthesis gas Flue gas -1- steam Flue gas + steam Flue gas + steam Flue gas + steam

Air Air + steam Air + steam Air Air Air Air + steam Air + steam Air Steam + 0 2

Coke Coke Coke Coke Calcined clay Coke Iron ore Laterite Allophane Alum ina ( + C 0 3 K 2) Catalyst Zeoliths in a matrix

Air Air Air Air

Final C Rejection

Coke Carrier

Coke Treatment

Carbon Rejection Processes o f the Fluid Coking Family

Fluid coking Flexicoking Flexicoking (dual G) LR coking ART process KK process KKI process

Process

TABLE 13

1

1 .3 -2 .1 .6 - 2 1 .4 -2 .5 1 . 1- 2 . la

0 .9 -1

1 .6

(0 . 8 )

0 .9 - 1.1 0 .8 - 1 0 .7 -0 . 8 1 .1 -1 .3

1 .1 -1 .3 1 .1 -1 .3 1 .1 -1 .3

Coke-C CR (Wt%/Wt%)





— —

— —





NA 8 0 -1 0 0 NA

1 2 0

1 2 0

1 2 0

Brom ine N o. (Naphtha)

-/c

0 . 8

2 0

1.05 1.05 1.05

32

3

2 .5 -3 . 6 3 .2 -3 .7 NA 5 -8 a

c

Residual Refining and Processing

Residual Refining and Processing

Gasification stage (air) Synthesis gas Gasification stage (steam)

Steam Air

Flexicoking Dual-gasification Flexicoking FIG. 14

Dual flexicoking.

The other processes in Table 13 (except for residue catalytic crackers) have not yet been industrialized and have merely been tested in demonstration units. The KK process also uses the fluid coking principle, but the temperatures in­ side the reactor are between 700 and 800°C. At such temperatures most produc­ tion is divided among gases (including C2H4, C3H6, and C4H8), gasoline, and pitch, with part of this last product recycled to the regenerator to provide the bed of coke particles with the thermal energy required for the transformation.

Air FIG. 15

Principle of the LR coker.

29

Residual Refining and Processing

The KKI (Kobe-Koa-Idemitsu) process aims to combine the coking of a residue and the reduction of crushed iron ore by the coke and its combustion products. This iron ore also carries the heat required for the reaction. The process thus mainly uses a reactor and an iron ore heater operating in fluidized beds. The partly reduced and sulfurized ore is entirely reduced afterward, either in a rotat­ ing furnace or in two successive fluidized beds. The first of these beds is fed with steam and oxygen for manufacturing the syngas used for the reduction. The HOT process also uses iron oxide as a carrier for heat, as well as for coke and sulfur. Three successive fluidized beds are used for the cracking, air com­ bustion of the coke deposited, and regeneration of the iron sulfide to oxide, which is recycled. The ACC process differs in two important ways. The particulate solid is an aluminosilicate of the allophane type, and the cracking operation takes place in two superimposed reactors. The coke deposited is burned in air in the regenerator. Dynacracking must be compared to flexicoking, but with the following dif­ ferences: the solid is alumina, and the reactor and regenerator are superimposed, as shown in Fig. 16. Regeneration is done with a steam-oxygen mixture to pro­ duce a hyrogen-rich gas. This gas is reinjected into the reactor so that cracking takes place under hydrogen partial pressure ranging between 0.5 and 1.5 MPa. The presence of hydrogen enables coke production to be reduced. Residue Catalytic Cracking. The processes are also shown in Table 13. When the feeds processed are heavy, these processes are fairly similar to fluid coking

FIG. 16

Dynacracking process reactor.

Residual Refining and Processing

and can be classified in the same family. However, the differences are great enough to deserve attention. The solid that carries heat and coke is not an inert solid but a veritable cata­ lyst that is generally composed of an ultrastable zeolite diluted in a suitably acid­ ified clay. Cracking transformation occurs mainly in a very brief time inside a riser that carries the catalyst from the regenerator to the disengaging-stripping section. Such residue catalytic cracking processes can handle only relatively light re­ sidual feeds with low metal and Conradson carbon contents. For a Conradson car­ bon level higher than 10, the carbon deposit becomes too great and there is danger of breaking the thermal equilibrium of the unit. For Ni + V > 30 ppm, the cat­ alyst becomes deactivated too quickly, zeolite loses its crystallinity, and yields decrease. To compensate for some loss of activity and selectivity, the daily makeup of fresh catalyst must be increased and catalyst consumption becomes prohibitive. Figure 17 schematically sums up the behavior of a heavy feed on a catalytic cracking catalyst. In the light of what has just been said about the pyrolysis of resins and asphaltenes, it is evident under the conditions of residue catalytic cracking that some thermal cracking of the feed cannot be prevented. This very fast pyrolysis involves the production of a pyrolytic coke at a level between 0.7 and 0.8 times the level of the corresponding Conradson carbon. The products re­ sulting from this thermal cracking of heavy molecules (unvaporizable) are then mixed with the vaporizable products initially contained in the feed. After this, at the level of the acid sites of the zeolite, they undergo an essentially catalytic cracking operation with the production of an additional amount of coke of the ordinary type. As in the coking fluid, the pyrolytic coke is deposited in a film on the outside surface of the solid grains, and the ordinary coke, like catalytic crack­ ing VGO, occurs dispersed inside the grains, more specifically at the level of the zeolite crystallites. To treat feeds that are richer and richer in Conradson carbon while producing acceptable octane-quality gasoline, conventional VGO catalytic cracking proLiquid + Gas

FIG. 17

Behavior o f heavy feedstock in residue catalytic cracking: (1) flash vaporization; (2) flash pyrolysis; (3) catalytic cracking.

Residual Refining and Processing

31

cesses had to be considerably modified, as did some of the operating parameters. These improvements have mainly concerned the following points: Decreasing pyrolytic coke production in the riser by better spraying of the feed inside the entrained bed of catalytic particles and by increasing the severity ( r 0 and F 0: temperature and heat flow) of the flash pyrolysis of unvolatilized heavy fractions (very brief contact time at high temperatures). Decreasing ordinary coke production by using ultrastabilized zeolites to diminish hydrogen transfer and the resulting coke production. Revamping the regeneration section to be able to burn a much larger coke pro­ duction without damaging the catalyst, especially its zeolitic components. This revamping had the following goals: (1) increase the volume of the regen­ erators, (2) increase the catalyst-oil ratio, (3) improve technology with regard to heat recovery, and (4) use additives to diminish and delay the harmful ef­ fects of metals (Ni + V) on the activity, selectivity, and stability of zeolitic catalysts. Perform ances In this survey of performance, residue catalytic cracking processes are dissociated from other pyrolysis processes. Product distribution in catalytic cracking is greatly modified by the interference of purely thermal reactions with catalytic reactions. Pyrolysis Processes. We do not analyze in detail the performance of all the processes listed in Tables 12 and 13, many of which do not yet have industrial references. We simply try to bring out some general laws that enable us to explain and, sometimes, forecast the overall yields of coke, liquids, and gases, depending on the feed processed, and particularly on the severity of the operation, no matter what process is considered. We then focus on industrialized processes for which results are available, that is, delayed cokers, fluid cokers, the ART process, and the EUREKA process.

Influence o f Operating Conditions. For a given feed, the relative importance of both cracking reactions and polycondensation of asphaltenes and resins depends mainly on the values posted for the different operational variables, such as tem­ perature, residence time, and pressure. On the basis of some data from Tables 12 and 13, we can see that the higher the temperature of the operation the lower is the coke-Conradson carbon ratio and the richer are the volatilizable products in olefinic and aromatic compounds. The bromine number of naphtha and the propene-propane ratio increase when the tem­ perature of the operation increases. At a given temperature, the longer the residence time, the more carbon, pitch, or coke reject is produced and the more secondary cracking is done, with an in­ crease in gas production. The residence time is defined here as the ratio between the volume of the reactor and the hourly throughput of liquid feed for coking drums, whereas for fluidized systems and risers residence time refers to the vol­

Residual Refining and Processing

ume flow of hydrocarbon vapor and steam and is measured in seconds, not hours. Qualitatively, the residence time of reaction fluids is nonetheless much longer in delayed cokers than in fluid cokers or, especially, riser cokers. In practice, it is very difficult to dissociate the respective effects of temper­ ature and residence time, just as it is difficult to give an exact value for each of these parameters. This is linked to the fact that the transformation is highly endothermic, that it takes place in an adiabatic system, and that it is accompanied by more or less extensive vaporization of the feed. The values for the operational variables that are accessible are usually as follows:

The temperature at the inlet to the reaction zone ( r 0). The residence time in relation to the liquid feed or to the vaporized feed 0). The temperature profile along the entire reaction zone (T = f(t)). In practice this variation in temperature depends on a great many factors, including the change in the carrier-oil ratios in risers and fluidized beds, as well as super­ heated steam injection in all cases. Generally speaking, for given T0 and t, less coke and gas are formed when the profile approaches a square profile. Such a profile is obtained when the heat flow F0 entering the reactor is much greater than that required to flash the vaporizable fraction and to supply the energy required for the endothermic transformation.

In summary, a coking operation is said to be severe when the values of F0 as well as T0 or t are high. For a given value of F0, an operation with a high tem­ perature and short residence time produces low yields of coke and gas and the volatilizable fractions are highly unsaturated. Such is the case with the ART pro­ cess. Simultaneous high T0 and t values are favorable to the appearance of sec­ ondary cracking without diminishing coke production, which even begins to increase above a certain value of T. It should be pointed out that these considerations concerning coke yield are the same as those that result from coal pyrolysis. Flash pyrolysis performed in a few hundredths of a second at high temperatures gives the best yields of volatil­ izable products. The cracked and volatilized molecules must be evacuated quickly to prevent them from recondensing upon contact with a liquid phase (pitch) or solid phase that is a veritable broth of free radicals. This effect of temperature on the coke yield is found for each process, as shown in Fig. 18a for the delayed coker [25]. An increase in pressure, at least when there is an increase in the partial pres­ sure of volatilized or cracked products, causes an increase in coke yield, as con­ firmed in Fig. 18b. The ACTIV process, which operates at reduced pressure, and the Eureka process, in which pitch is stripped by a high throughput of super­ heated steam, confirm this observation about the effect of pressure on coke yield (Table 12). To improve the quality of the volatilized fraction or to increase the coke yield, part of the heavy volatilized fraction may have to be recycled (fractionator bottom for the delayed coker or slurry from the scrubber bottom for the fluid coker). There is then an increase in coke, as shown in Fig. 18c for the delayed coker.

33

Residual Refining and Processing

a)

Coke Drum Temperature, °F

b)

Coke Drum Pressure, PSIG

C)

Combined Feed Ratio FIG. 18

Temperature (a), pressure (b), and recycle (c) versus coke yield.

Influence o f Technological Characteristics o f Processes. Whether the process is of the delayed or fluid type, the transformation to be made is the same and the performance depends above all on the operating conditions T0, t , and T = f (t ) the technique allows. Processes of the delayed coking type, working at low severity, produce more coke, but this production often remains the main goal of such processes. Using a furnace to transfer most of the heat required for the operation limits the Conrad-

Residual Refining and Processing

son carbon content to 25-30% of the feeds to be processed even if steam injection and improvements in furnace technology have enabled residence times in the hot­ test tubes to be reduced to values between 10 and 30 s. The linear velocity of the liquid has been increased, and the rate of coke deposition in the furnace has been limited. Yields of liquid products are of course lower than in processes of the fluid type, but they are less unsaturated and easier to hydrotreat. It must nonetheless be pointed out that when coke is not the desired product, the yield of volatilizable products can be increased by reducing the residence time of the vapor phase in the coking drum, either by decreasing total pressure or by steam injection. These general considerations are valid whatever the process, whether coke or a pitch for use as a coke precursor is produced. Delayed coking is the oldest and by far the most widespread coking process. It is also the one requiring the lowest investment per ton feed processed, and it is constantly being improved in all its aspects, that is, the furnace, controls, monitoring by microprocessor, and the de­ coking step. Processes of the fluid coking type operate at high severity. T0 is high, as is the thermal flow linked to the solid throughput on which coke is deposited. The aim is not to produce coke but generally to reject carbon and impurities (metals and sediments) with maximum yield of liquid fractions. The Conradson carbon content of the feeds processed must be greater than 5% for coke combustion to be able to produce the thermal energy required for vola­ tilization and then cracking. It may reach values higher than 40% to the extent that the softening point is not too high and the thermal flow is sufficient to ensure quick volatilization of the volatilizable fractions and to prevent the formation of bogging lime on the solid particles. In the same category, the LR coker is designed to process solid feeds, whereas the ART process is preferable for atmospheric residues with a view to preparing catalytic cracking feeds. Fluid coking and other processes of the same type have great flexibility. The carrier-feed ratio, regeneration temperature, and recycled fraction are easy to control, so that the liquid fraction yield can be modified at will. The difference in the ART or LR coker processes in relation to traditional fluid systems stems is that the reactors are concentration gradient (and tempera­ ture gradient) reactors, whereas both the fluid coker and delayed coker make use of a homogeneous concentration reactor. In a kinetic system of consecutive reac­ tions with apparently positive order, the gradient reactor both is more efficient per unit reaction volume and, in particular, can provide better selectivity of interme­ diate products. Influence o f the Nature o f Feeds. Each process is characterized by a coke yield that can be linked to the Conradson carbon content of the feed. This coke to Con­ radson carbon ratio varies from one process to another depending on the severity of the operation, but for each process it has a slight tendency to diminish when the Conradson carbon content of the feed increases, as shown in Figs. 19 and 20. The distribution of various kinds of impurities in the feed deserves particular attention. Metals (Ni + V) from the feed are generally more than 95% found in coke, whereas sulfur is only about 50% found for delayed cokers and about 25% for fluid cokers. As for the nitrogen initially contained in the feed, 60-80% of it

35

Residual Refining and Processing

CONRADSON CARBON FIG. 19

Coke yield as a function of feedstock Conradson carbon.

is found in the carbon deposit for processes of the delayed coking type and about 50% in processes operating at higher temperature (fluid or ART coking). This is so because the nitrogen of heavy heterocyclic compounds occurs in the form of particularly stable — C = N — bonds, as shown in Table 11. As an illustration, some typical results concerning delayed coking, the EUREKA process, fluid cok­ ing, and the ART process are shown in Table 14.

Product yields

Conradson Carbon Residue of LR-Coker Feedstock FIG. 2 0

LR coking or residual fractions: product yield structure [15].

1.1 62 90

122

70

5 5 .6

58

1.1

C 5-3 80°F , 3 8 0 -6 5 0 , 650°F + , C oke, wt%,

17.8 20.5 17.5 37

65

2 9 .5 5 5 8 .8 0 .8 50 82

C 5 -3 7 5 °F , 3 7 5 -6 6 0 , 6 6 0 - 1 ,004°F, Pitch, wt%,

0 .7 4 .1 4

2.1

1.5 12.3 C 5-350°F , 14.5 3 5 0 -6 5 0 , 18.0 650°F + , 2 6 .4 Net coke, wt%, 0 .7

4 ,8 0 0 27.7 269

11.1

M ixed M iddle East 932+ 7 .4 3 .9 5 ,0 0 0 20 338

Arabian H eavy 1 ,0 5 0 + 4 .5 5 .3 4 ,6 0 0 2 4 .2 255

Arabian Heavy 1,050+ 3 .0

6.0

Eureka Process [12]

D elayed C okinga

Flexicoking [77]

T yp ica l R esu lts in C ok in g P ro cesses

Feedstock Cut range, °F Gravity, °API S , wt% N , ppm Concarbon wt% M etals, ppm Y ield and product characteristics H 2 S , wt% C4~ , wt% Naphtha, (vol% ) G O , (vol% ) H G O , (vol% ) Coke (pitch) Naphtha °API S, wt% N , ppm Br, no. RON

TABLE 14

6 6 .5 0 .4 — 8 0 -1 0 0 ~ 80

0 .3 4 5 .4 7 12.1 C 5-4 0 0 °F , 2 1 .9 4 0 0 -6 6 0 , 4 1 .7 660°F + , Coke (burnt), wt%,

Arabian H eavy 667+ 12.6 4 .4 2 ,7 0 0 15 125

ART Process [78]

17 16.5 5 5 .3 10.7

aIFP data.

GO °API S, wt% N, ppm Br, no. Diesel index HGO °API S, wt% N, ppm Concarbon, wt% Metals, ppm Ultimate residue 10.1 5.8 4,600 4.8 1.7 Coke, wt%, 0.7 Gas mol% ch4 1 .2 14.9 h2 CO 17.5 co2 10.5 h 2o 5.2 51 n2

29.5 3.4 950 — —

Metals, ppm, N, wt%, 1.0

S, wt%, 6.4

Coke

13.2 5.3 4,100 — 4

30.2 3.9 800 45 36

S, wt%, N, wt%, Concarbon, wt%, C 7 asphalt, wt%, Metals, ppm,

Pitch

0.6

16.9 2.94 2,400

33 2.3 680 50 35

5.73 1.2 60.3 80.4 1175

Flue gas

11

5

11.4 4.3

10-15

27.4 1.4

vl

Residual Refining and Processing

Residual Refining and Processing

Residual Catalytic Crackers. These processes now apply only to products roughly meeting the following specifications:

Conradson carbon < 10% Ramsbottom carbon < 8 .5 % Metals (Ni + V) < 30 ppm Nitrogen < 3 ,0 0 0 -3 ,5 0 0 ppm

The feeds considered may be deasphalted oils, VGO from coking and, in partic­ ular, from ART process, atmospheric residues from light crudes, or mixtures of VGO with an atmospheric residue or a vacuum residue.

Influence o f Operational Parameters and Variables.

The discussion of the ef­ fects of a modification in the severity in coking processes ( T0, f, and F) can be repeated for residue catalytic crackers. The main difference is the presence of a zeolitic catalyst, which speeds and directs the cracking and rearrangement of the vaporizable fraction according to conventional mechanisms of acid catalysis in­ volving carbonium ions. In practice, consideration must be given to the presence of nitrogen and metals in the feed. The conversion, selectivity, and octane number of the gasoline depend on the type of catalyst as well as on its state of aging in the unit, and, especially, on its mean metal content, which may be higher than 10,000 ppm.

Influence o f Technological Characteristics.

Processes may be classified in two categories, depending on whether the regeneration procedure is done in one or two steps. In single-step processes, care is taken to operate at a moderate temperature so as not to deteriorate the zeolite in the presence of an appreciable amount of steam from combustion of the hydrogen contained in the coke. Heat is then extracted by internal regenerator-bed coils and by external coolers. Such practice is offered in HOC units. In two-step processes, 60% of the carbon and most of the hydrogen are burned below 700°C in the first step. Combustion is performed at a higher temperature (above 800°C) in the second step, under low steam pressure. This procedure is recommended for R-2R and S&W FCC. It avoids or at least delays the hydrothermal aging of zeolitic components. The high temperatures require the second generator to have a heat-proof lining and the cyclones to be situated outside the regeneration enclosure. This type of procedure also increases the T0 value in the riser and slightly diminishes pyrolytic coke production. Figure 21 illustrates this type of setup for an R-2R process. As in the ART process (or the LR coker), the catalytic bed is a circulating bed of catalyst, and the riser must be designed and equipped to reduce backmixing of the catalyst and reaction fluid as much as possible, to obtain proper spraying of the feed, to control the temperature profile, and to separate quickly and effec­ tively the products from the solid at the top of the riser. This set of precautions reduces coke and gas yields for a given conversion rate.

39

Residual Refining and Processing

FIG. 21

Residual catalytic cracking with dual regeneration (R — 2R = Total France; IFP).

Influence o f the Nature o f Feeds.

The Conradson carbon content has a direct

influence on coke production:

% coke = (0 .7 -0 . 8 ) Conradson carbon + ordinary coke = / (conversion, catalyst, Fq) with ordinary coke However, on the basis of data available in the literature, it is difficult to de­ termine the clear-cut influence of the feed composition on yields and on the char­ acteristics of different products, except for extreme feeds, either vary paraffinic or very aromatic. A very paraffinic feed gives an LCO with a cetane number that may be between 30 and 40. A very naphthenic-aromatic feed or especially a very naphthenic feed provides better octane for naphtha, a low coke yield, and a very small C4“ fraction. This last observation must be linked to the intervention of a volatilizable fraction that acts as a hydrogen donor, during both the pyrolysis step (coke) and the catalytic step (gas). Such an effect on the coke yield is found in coking processes [11]. In such a case, the coke yield may not conform to the usual correlations.

Residual Refining and Processing

For the distribution of impurities (metals, sulfur, and nitrogen) among the dif­ ferent products, the observations made concerning fluid coking remain applicable for residue catalytic crackers. Some typical results concerning the catalytic crack­ ing of residues are given in Table 15. The sets of results presented in Table 15 cannot be compared; they involve different technologies, feedstocks, and Concarb contents and are processed at different severities. It should be noted that in the three examples the C 3/C 4 ratio remains almost constant.

TABLE 15 Residue Catalytic Cracking: Typical Performance

0.3 0.15 4.8 19

0 .1 1

1.41 1 .0 2

0.97 1.30 2.40 3.80 6.75 0.80 1.27 3.50 5.77 5.30 8.54 C5-2 1 5 ‘3C 49.80 61.22 56.3 93.2 81.2 150 215-360°C 18.50 17.86 16.5 25 0.53 600 5.41 6.30

Wt% 0.06 0.83 0.75 0.53 0.77 2.42 0.82 2.16 3.17

Wt%

c5

1.41 4.30 1.30 3.56 4.83 °C 40.94

63 91.2 79.2

L5 4 J1 1

> 10.6 (17.10) J1 C5-205°C 43. 6 59 59.9 91 0.28 205-343°C 13,

1

15.9

2 1

29 0.38 400

2 . 8

7.14

7.99

6

.1

4.2 0.63 6.90 8 6

9.73

11 1

1 2 0

185-380°C 44.78 41.52

0.85

Vol%

►5.1

32.04

0

7.20 86.50

Vol%

L

Vol%

Wt%

17.7 3.0 0.19 7.5 36

2 1 . 2

0.3 0.15 4.8 19

5.6 4.3 5.6

16 o

aTotal-IFP, R-2R results. bFrom Ref. 79.

2 1 . 2

Maximum Gasolineb

00

Feed °API S, wt% N, wt% Concarbon, wt% Metals, ppm Products H2S Methane + H 2 Ethane Ethylene Propane Propene Butane Isobutane Butenes Gasoline Yields °API RON, clear MON, clear S, ppm LCO Yields °API Cetane number S, wt% N, ppm Slurry (yields) °API S, wt% Coke Conversion

00 /1

Characteristics

Maximum Middle Distillate 3

Maximum Gasoline 3

.8

41

Residual Refining and Processing

Products and F urther Uses Coke and Distillates from Pyrolysis Processes. Distillates issuing from pyrolysis processes are not directly usable. The octane properties of naphtha are poor, especially for delayed coking (research octane number RON 60 -6 5 ). The cetane properties of light gas oil are generally mediocre, especially when pro­ cessing is done at high severity (fluid coker). Likewise, all these products are un­ stable because highly unsaturated and relatively rich in sulfur, nitrogen, metals, and metalloids. Among these last impurities, mention must be made of silicon in naphthas and gas oils from delayed coking and, sometimes, arsenic. In vacuum gas oils nickel and vanadium are found at levels between 1 and 20 ppm, depending on the feed processed and the processing severity. All these products are generally hydrotreated. After hydrotreatment the naph­ tha is fed to the reformer and the LGO is simply mixed with the cetane pool of the refinery; the vacuum gas oil is catalytically cracked (or possibly hydrocracked). On the basis of results announced by Exxon [26], Fig. 22 gives the weight balance and refining scheme involving a fluid coker and the further treatment of the different products. Table 16 gives the results that can be expected from the separate hydrotreatment of naphtha and gas oils, as well as the different problems raised by the processing of these feeds, no matter what process produced them. These hydrotreatments are not easy operations. The denitrogenation and hydrogenation of aromatics require the use of high pressures (P > 100 bar for gas oils). The hydrogenation of olefinic and diolefinic compounds may require a specific catalyst if their concentrations are too high. Silicon, coming from the de­ composition of antifoam additives, is entirely captured by the catalyst and pro­ gressively lowers its hydrogenating function.

TABLE 16 Hydrotreatment o f Fluid Coking Products

Naphtha, C5-170°F

Characteristics

Feed

Gravity, °API Sulfur, wt% Nitrogen, ppm Bromine no., g/100 g Diene value, g/100 g

56.7 1.17 80

Middle distillate, 67% from fluid coking, 33% from HDT of heavy gas oil

Gravity, °API Sulfur, wt% Nitrogen, ppm Bromine no. MAV Aromatics, vol% Cetane index

Heavy gas oil (650-975°F) from fluid coking

Gravity, °API Sulfur, wt% Nitrogen, ppm Metals (Ni + V) ppm Concarbon, wt%

1 2 2

Products


450 °C) is about 1 min. The density of the average heat flux across the tube walls ranges between 25 and 30 kW /m 2*h. The soaker cracker uses a soaking chamber to complete the conversion already initiated in the heater. The temperature is between 30 and 40°C lower than in the coil cracker, but the residence time in the soaking chamber is relatively long, from 10 to 20 min (liquid feed at 15°C). Figure 26 is a simplified flow sheet of the two visbreaking variants, and Fig. 27 shows the changes in operating conditions in both cases [31]. Both processes require the application of an effective quench at the exit of the reaction zone to prevent coke buildup in the downstream equipment. This quenching can be done with several agents, depending on the requirements of the refinery: gas oil, vis­ breaking residue, and sometimes crude oil. Depending on the refining flow sheet, the distillation section may be designed to separate naphtha or atmospheric dis­ tillates only. The distillation section is sometimes equipped with a vacuum col­ umn or a vacuum flash to separate a VGO fed to catalytic cracking. It is very difficult to assign a decisive advantage to one or the other of these two variants.

Residual Refining and Processing

FIG. 2 6

Simplified schemes o f visbreakers: (1) furnace and (2) soaker, both equipped to produce some heavy gas oil.

In terms of yield, performance is comparable in both. A bare soaker, operating upflow without internal components, gives substantially the same yields as a coil reactor, if it is well designed (L/D = 6 ). By equipping it with horizontal sieve trays [32] or vertical walls, several reaction stages are created, which transform a steady-state concentration reactor into a reactor that approaches a plug-flow re­ actor, with the resulting consequences for a reaction with an apparent positive or­ der, one in the present case:

53

Residual Refining and Processing

0)

3 CO & E

S* c o e c

Q)

6

0k.»

* C o

+3■»

(0

I E

o u

In FIG. 27

Out Typical conditions along furnace and soaker [31].

Improved conversion per unit reaction volume, Less secondary cracking: slightly less light product (gas and gasoline) is produced for a given conversion. Less secondary condensation: at a given temperature, the tendency is to produce more asphaltenes than in a bare soaker, but the value of the flocculation threshold is slightly lower because these asphaltenes are less “ compact.” From one stage of the soaker to the other, backmixing is prevented by the in­ crease in the space velocities of the gas and liquid through the tray orifices. The coil reactor offers the advantages of any plug-flow reactor. It is efficient in terms of conversion per unit volume and theoretically selective in the sense that it should more effectively limit the subsequent secondary cracking reactions. This advantage does not appear in practice for the following reasons: The apparent activation energy of the cracking reactions of small molecules is higher than those of heavier homolog molecules. An elevation in temperature thus tends to favor secondary cracking.

Residual Refining and Processing

The fluid film on the tube walls is at a higher temperature than the average tem­ perature of the reaction fluid, and the residence time at this film is relatively long. This leads to overcracking, which tends to disturb yield. With respect to condensation reactions, the coil reactor theoretically offers a slight advantage over the soaker, because the value of E for cracking is much higher than for condensation reactions. This advantage is partly nullified by the development of overcondensation and overcoking reactions in the film. In terms of operability , operating costs, and investment cost, Table 21 summa­ rizes the respective advantages of each of the processes on the basis of compa­ rable conversions for a given feedstock. Process Performance. The influence of the operating conditions and of the technological characteristics of the process on product yields and quality have al­ ready been examined. Table 22 shows the variation in yields obtained in a soaker unit treating a Middle East vacuum residue when the soaking temperature is var­ ied. Table 24 illustrates this variation for a coil cracker operation with Safaniya vacuum atmospheric as feedstock. The influence of the type of feedstock remains to be examined. Vacuum res­ idues and conventional atmospheric residues substantially display the same yield pattern and the same variation in viscosity as a function of conversion. Figure 28 graphically illustrates this observation based on the treatment of six different feed­ stocks. However, it must be observed that these different feedstocks do not show the same tendency to flocculation and cannot be treated with the same severity. The appearance of flocculation is reflected in Fig. 28 by a halt in the viscosity TABLE 21

Advantages of Each Option Coil Cracker

Yields

Operability

Higher gas yield +0.5% wt maximum Slightly better stability of fuel Decoking operation is easier and shorter (burning) (run length 4 - 6 months)

Soaker Higher asphaltene yield +0.5% wt maximum Easier process control, longer run length ( 1 year), decoking prob­ lems more difficult to overcome especially with equipped soakers

Higher outlet T make possible a further heavy gas oil recovery Utilities Investments

No soaker and corresponding process control Can avoid building a vacuum column for some HGO production

10-15% less net fuel, 0.2% on feed Smaller furnace Less equipment for heat recovery Less pumping energy because of lower pressure drop As a whole 10% less investment than coil

55

Residual Refining and Processing

TABLE 22

Soaker Visbreaking Yields. Feed: Middle East Short Residues3 Yield (Wt%)

7TC) residence time H2S C, + C 2 C 3 + C4 C5-1 5 0 oC 150-250°C 250-375°C VGO Residues Severity FT

440 t 0 . 2 0

0.26 0.71 1.80 1.81 5.86 1 2 .0 2

77.54 4.5

450 t

455 1.5f

0.44 0.44 1.18 4.30 4.44 9.24 16.00 64.80 7

0.56 0 . 6 8

1.33 7.28 7.51 13.94 18.54 50.3 10

aFeed characteristics: density at 20/4°C, 1.014°; viscosity at 100°C, 736 cSt; 5.35 wt% S; 0.37 wt% N; 17.5 wt% Concarbon; 13.5 wt% C5 asphalt; 6.5 wt% C7 asphalt. IFP results.

drop. Note also that certain unusual feedstocks display special behavior: these in­ clude high-asphaltene feedstocks, such as asphalts from deasphalting operations, as well as certain feeds that are both asphaltenic and highly paraffinic. They tend to yield coke, and the visbroken fuels obtained are often unstable. The analytic approach presented in Fig. 24 can help to solve these problems. These unusual feedstocks also include the vacuum residue from catalytic hydroconversion of vac­ uum resids to increase slightly the gross yield of distillates. It is perhaps a pity to destroy the properties of an excellent fuel oil by subsequent heat treatment. If the catalyst can withstand it, it is preferable to increase the conversion in the hydrotreating step. Uses and Further Refining of Products. Table 23 lists the characteristics o f various products from the treatment of a Middle East vacuum residue in a soaker cracker. The characteristics of the gas and various atmospheric distillates are very similar to those of products obtained from delayed coking operations. These dis­ tillates must be hydrotreated for sulfur, nitrogen, and olefin removal before being further processed or sent to storage. Hydrotreating is generally carried out in ex­ isting units after mixture with straight-run feedstocks. The problems encountered in the treatment of these distillates are of the same type as those discussed con­ cerning the hydrotreating of delayed coking products. However, since they were obtained at a lower severity than in coking, they are less refractory. Moreover, they account for only a small proportion of the feedstocks sent to hydrotreating. After hydrotreating, the distillates are similar to the straight-run distillates contained in the original crude. The naphtha has an RON in the range 6 0 -7 0 and is fed to the reformer. The cetane number of gas oil ranges between 40 and 50 and is more or less correlated with the straight-run gas oil cetane number. To reduce the production of fuel oils, certain refineries separate a heavy gas oil or even a vacuum gas oil, which are sent to the catalytic cracker. LCO, HCO, and slurry are used as cutterstock to adjust the viscosity of the fuel oil. This flow

aIFP results.

Concarbon, wt% C 5 asphaltenes, wt% C 7 asphaltenes, wt% S, wt% N, wt%

1.040 1,640 32 28 26 5.9 0.46

7 1.030 1,360 26 23 17 5.55 0.42

17.5 13.5 6.5 5.35 0.37



66.5

81.4

1.014 736

and pa, see Fig. 24.

11.03 4.5

S 2 2 . 1 0

6.5

0.08

1.59 0.94 0.41 5 0.13

0 . 1 2

6

2

0.94 0.53 2

11.40 4

19.60 6.5

0.17

1.89 1.18 0.38 5 0.17

0 . 1 2

5.2

2 . 1 2

0.96 0.55 1

Residual Refining and Processing

Conversion W t%

Mean average boiling point °C FIG. 3 0

Sulfur distribution in visbroken products.

N oncatalytic Processes Under H ydrogen Pressure Compared to coking or visbreaking processes, these noncatalytic processes have not yet led to any important industrial applications. The latest oil crises revived interest in such processes, which seem to be relatively well suited for processing very heavy oils that are very rich in metals and sediments. All such processes can be classified in three categories: 1. 2.

3.

Hydrovisbreaking processes, which differ from visbreaking only by the judi­ cious use of hydrogen pressure. Hydrogen donor diluent processes in which processing occurs in the presence of a solvent capable of transferring hydrogen to the feed to be cracked; these processes also work under hydrogen partial pressure. Slurry processes working under hydrogen pressure but in the presence of powdered or colloidal additives; such processes stem directly from ones de­ veloped for coal liquefaction during World War II and an industrial unit operated on the VEBA COMBI process for more than 10 years, converting petroleum residues, before being shut down in 1964. In this process the additive was made up of iron oxide associated with powdered coal as a coke seed.

These three types of processes have the joint feature that the products cracked are highly olefinic and have a composition similar to those issuing from visbreaking

Residual Refining and Processing

61

0

Conversion

100 2

N.visb VR

1.5

Nfeed

1

N Distillates N Feed

0.5

0 0

500 Mean AV. Boiling point (°C)

FIG. 31

Nitrogen distribution in VB products.

and coking. The same free radical mechanism governs the transformation of the feed, and cracking and condensation reactions work in opposition, creating all the problems linked to the state in which asphaltenes are found in solution. An anal­ ysis still has to be made of how hydrogen, donor solvents, and additives can delay asphaltene condensation reactions, asphaltene flocculation, and coke formation, so that higher conversions can be achieved than with visbreaking. Processes of the slurry type could certainly also have been grouped with fixed-bed, moving-bed, or ebullating-bed catalytic processes like those described

TABLE 26

Visbreaking of a Middle East Short-Residue: S and N Content in Visbroken Fractions S (Wt%)

Conversion (Wt% at 550°C) Light gasoline Naphtha GO VGO Vacuum residue

2 2

1 .2 0

1.67 3.02 3.80 5.60

N (ppm)

34

56

1.31 1.64 2.73 3.82 5.95

1.30 1.53 2.62 3.85 6.45

34

2 2



— 5

1 0 0

477 4,320

15 1 2 0

527 4,350

56 10 2 0

140 753 5,110

Residual Refining and Processing

later. Indeed, it is undeniable that solid additives added to the feed have a certain catalytic activity, but their function consists mainly of delaying coke formation, not catalyzing hydrorefining reactions, such as hydrodesulfurization, hydrodenitrogenation, and aromatics hydrogenation. They actually have as many points in common with hydrovisbreaking and H-donor solvent processes as with hydrore­ fining processes. Hydrovisbreaking. This operation is performed under severity conditions (t and T) comparable to those used in visbreaking but under a hydrogen pressure that may be between 8 and 18 M Pa, depending on the feed to be processed and the conversion goal. As in visbreaking, the conversion of feeds takes place according to an apparent first-order kinetics and with an apparent activation energy of about 230 kJ/m ol. At isoconversion, the distribution of distillates and the characteristics of these dis­ tillates are very close in both cases. Gas, naphtha, and gas oils remain highly olefinic, and observations made for visbreaking concerning sulfur and nitrogen distribution remain valid in the presence of hydrogen. Table 27 illustrates these different observations. The two processes nevertheless display a number of differences that justify the application of hydrogen partial pressure as well as concomitant increases in in­ vestment and operating costs. These differences mainly concern the yields of res­ idues and their qualities (H/C ratio).

In hydrovisbreaking, some hydrogen is consumed (0 .2 -0 .4 wt% ), whereas in vis­ breaking, a very small amount of hydrogen is produced (—0.05 wt%). For a given conversion level, the hydrovisbroken (HVB) residue is always further from the flocculation threshold than the visbroken (VB) residue. For a given FT rating, hydrovisbreaking has the following advantages over visbreaking: (1) conversion into distillates is higher by 3 0 -5 0 wt%; and (2) the quality of the residue is better, the H/C ratio is higher, and the density and viscosity are lower; AsC 5 has a tendency to decrease as opposed to what is observed in visbreaking (Table 28 illustrates these observations).

For lack of sufficient data on the matter, it is very difficult to state clearly how hydrogen acts to slow or delay condensation reactions. However, hydrogen can be assumed to play the following roles: ( 1) a thermodynamic role through which it prevents excessive dehydrogenation of the pericondensed naphthenic and aro­ matic nuclei partly making up the heavy molecules; and ( 2 ) a chemical role as­ sociated with the fact that, simultaneously with hydrogen transfer from the heavy to the light components, a nonnegligible chemical consumption of hydrogen oc­ curs. However, how this hydrogen is activated remains to be explained. The following more or less reliable proposals have been made:

Direct activation of the H— H bond by collision with sufficiently energetic radi­ cals existing in the liquid medium [33].

3.60 30.57 28.15 35.37

2.32 1.32 2.33 3.14 4.70



S

— —

0 . 2

0.85

0 .1 1

70 19



Br no.

0.04



N

Specific gravity at 20/4°C C 5 Asphalt, wt% C 7 Asphalt, wt% Concarbon, wt% Viscosity at 100°C, cSt

Characteristics

1 2 .0

11.9 8.79 11.9

0.965

Syncrude (C5+ )

18.3 11.5 18.3 130

1 .0 2 0

350°C+

32.9 24.3 32.9 58,000

1.078

500°C+

aFeed characteristics: specific gravity at 20/4°C, 0.995; viscosity at 100°C, 95 cSt; 4 .2 0 wt% S; 0 .3 9 wt% N; 13.50 wt% Concarbon; 15.50 wt% C5 asphalt; 7.80 wt% C 7 asphalt. Asvahl results.

C4, < 150°C 150-350°C 350-500°C 500°C+

c4“

Yield

Synthetic Crude and Residue Characteristics

Hydrovisbreaking o f Cold Lake Topped Crude Oila

Yields (Wt%) and Characteristics o f Products

TABLE 27

Residual Refining and Processing

Residual Refining and Processing

TABLE 28

Characteristics of Hydrovisbroken and Visbroken Residues (375°C+ ): Feedstock, Topped Boscana

Yield 375°C+ Specific gravity at 20°/4°C Viscosity at 100°C, cSt S, wt% N, wt% Concarbon, wt% C 5 asphalt, wt% C7 asphalt, wt% H/C, wt%

Feed

Visbroken

Hydrovisbroken

17 0.998 198 5.25 0.7 15.2 19.5 12.3 0.127

31.76 1.046 842 5.09 0.95 24.3 29.6 26.7

35.94 1.040 511 5.24 1.06 24.2 24.2

0 .1 1 2

2 0

0.116

aAsvahl results.

Hydrogenation of very reactive pericondensed aromatics, which can then act as hydrogen donors. H2S can then act as a catalyst in such a donation [34]. Activation of hydrogen by nickel and vanadium sulfides issuing from partial as­ phaltene and resin demetallization. This activated hydrogen can in turn cap­ ture radicals in solution or participate in the partial hydrogenation of pericondensed aromatics playing the role of hydrogen donors. Hydrovisbreak­ ing should be an “ em bryo” of the Aurabon process, which is discussed later.

Figure 32 is a schematic of a hydrovisbreaking process that aims to transform a heavy oil into a transportable syncrude on a production field. Three specific applications give the scope of this process:

1.

The first application concerns the visbreaking of heavy oils to make them transportable without making too much of a change in their qualities as a source of road asphalt. For pipelining, Canadian and European recommen­ dations are 25 cSt at 50°C and the Venezuelan is 180 cSt at 50°C. Pipeline heating or dilution by a low-viscosity diluent are alternatives that have been applied to date, but in many cases hydrovisbreaking could apparently be a simple and attractive solution. Figure 33 shows how the viscosity of a Boscan

Recycle compressor

Makeup compressor

Residual Refining and Processing

FIG. 3 3

2.

3.

65

Comparison of viscosity of syncrudes: (1) visbreaking (Tervahl T) and (2) hydrovisbreaking (Tervahl H). Severity index = syncrude FT (Flocculation threshold) — feed FT.

heavy oil changes as a function of the severity applied comparatively in hy­ drovisbreaking and visbreaking. As a second application, it should be mentioned that a prior hydrovisbreaking step can be integrated with catalytic hydroconversion of residues as a simple way of increasing conversion yields [351, [36]. This processing scheme is tackled later. The third application concerns thermal hydrocracking at a very high hydro­ gen pressure (25 MPa) to ensure very deep conversion, as examined later with reference to slurry processes.

Hydrogen Donor Processes. The addition of a hydrogen donor solvent under hydrogen pressure ensures, without damage deeper conversion than could be ac­ complished in the presence of hydrogen alone. The use of donor solvents was mentioned as early as 1933 for coal liquefaction, but it was not patented for the processing of petroleum residues until 1947 [37]. The most important research was done by Varga and then, starting in 1950, by Esso Research and Develop­ ment, which developed hydrogen donor diluent cracking (HDDC) [38a] and hy­ drogen donor diluent visbreaking (HDDV) [38c] processes. This type of process seems to be well suited for the conversion of heavy oils, which are much easier to crack than residues from conventional crude oils. As shown in Table 29, these processes are differentiated by their operating conditions and by the nature of the hydrogen donor solvent. In all cases, however, this solvent is a cut rich in partially hydrogenated pericondensed aromatic hydrocarbons of the Tetralin, dihydroan­ thracene, or dihydrophenanthrene type, among others. These molecules have very labile hydrogen atoms capable of reacting with all radicals from the surrounding environment. Through this capture effect as well a dilution effect, the solvent diminishes the concentration of heavy hydrocarbon

4 -7 12-16

Lurgi

3

Exxon

Gulf Canada Corp.

3

P (MPa)

Exxon

Licenser

Characteristics of H-Donor Solvent Processes

H-donor diluent cracking (HDDC) [38a] H-donor diluent visbreaking (HDDV) [38c] Donor-refined bitumen (DRB) [39] Donor solvent hydrovisbreaking (DSV) [40]

TABLE 29

4 2 0 -4 9 0

4 1 0 -4 6 0

4 2 0 -4 6 0

4 5 0 -4 8 0

T (°C)

0

0

2 0 0 -5 0 0

3 8 0 -4 8 0

3 8 0 -4 8 0

0 .5 -1

.5

.5

2 2 0 -3 5 0

0

0

S olvent(S) Boiling Range (°C)

0 .5 -2

.2 -

.2 -

t (h)

0

.2 -

0

1 - 2

0 .5 -1

0 .5 -1

.4

S/Feed (m 3 /m3)

5 0 -1 5 0

5 0 -1 0 0

60

1 0 0

H 2 Consumed (m 3 /t)

5 0 -8 0

>60

2 0 -4 0

>50

Conversion of Feed (Wt%)

o> o>

Residual Refining and Processing

67

Residual Refining and Processing

TABLE 30

Some Reactions Involved in Hydrogen Donor Pro­ cesses: Feedstock, Asphalt (As)a

As° + As° o O

/ H As° + D ( XH

— -----------------------

As — As

»

^2

R° + olefin

— ---------------------->



— ----------------

>

AsH + D°

d2

As° + D° XH

----------- >

R° + D° XH

----------------------------

d3

AsH + D (aromatic)

>

RH + D

As° + D° XH

8 ----------->

As - DH

D° + D° XH XH

c, -----------»

/ H D - D( XH

D + D

------------ >

C2

D - D + H2

ac „ condensation of asphaltenes; c2, secondary cracking; d ,, d2, capture of asphaltene radicals by H°; d3, rupture of chain; g , solvent grafting; C ,, C2, solvent coupling.

radicals in the reaction medium and slows their condensation. At the same time, this labile hydrogen may capture radicals with the lowest molecular weight that from cracking reactions, thus diminishing secondary cracking and the ensuing gas production. Overall, the donor solvent slows conversion but makes it more selec­ tive. The operation must thus be done either at higher temperature or, preferably, with a longer residence time. At the same time as the acceptor compounds of the feed, the H-solvent donor can also react with itself or with the feed by coupling or grafting. These different possibilities are summarized in Table 30 for an asphalt feed. If we refer to the velocities of these different reactions, the best donor sol­ vent is that for which the k j k c, kdlk g, and k j k c ratios are the highest. Likewise, a direct role for hydrogen cannot also be dismissed, such as was given concerning hydrovisbreaking, when the hydrogen pressure applied is sufficiently high. Table 31 compares the performances of several donors in the ultim ate pyrolysis of an

TABLE 31

Ultimate Pyrolysis of the Boscan C 5 Asphalt in the Presence of Various Solvents3 Products (Wt%)

Solvent Nature Benzene T-DHN T-DHA T-DHP Benzene + H2S

Composition (Vol%)

Gas

Liquid

Coke

1 0 0

32

14 40 49 50 56

54 54 45 44 37

84 -1 6 8 4 -1 6 8 4 -1 6

6 6 6

7

a.P = 0.1 MPa; T = 500°C; asphalt, 15 wt%; DHN, dihydronaphthalene; DHA, dihydroan-thracene; DHP, dihydropyrene; T, tetraline. Source: From Ref. 34.

Residual Refining and Processing

asphalt of the Boscan type precipitated by pentane. The efficiency of H2S in free radical capture can be seen clearly [34]. Concerning the flow diagrams, all processes of the hydrogen donor type are similar, with the exception of the pressure. They all have a heater, a reactor, a fractionation section, and a hydrogenation section for a distillate that is recycled at the level of the heater. A simplified diagram of such a process is given in Fig. 34. For each process, the pressure applied must be sufficient to keep the donor solvent in the liquid phase. If pressures higher than 8 MPa are used, the action of the donor solvent can be coupled with that of the hydrogen, as described in the hydrovisbreaking operation. Such coupling enables the severity of the operation to be increased and high conversion to be obtained with isoviscosity of the residue. The transfer of an appreciable amount of hydrogen from the donor solvent to both the distillates and the cracking residue changes the yields of different prod­ ucts as well as their quality compared to what can be obtained by visbreaking and hydrovisbreaking:

Gas yields are relatively small compared to gasoline yields. The impurity contents (S and N) of the gases and distillates are slightly greater. The distillates are considerably less olefinic. These distillates, however, must undergo the same refining treatments as visbreaking or hydrovisbreaking products.

H2 recycle - Purification

FIG. 34

Gas + H2 purge

Simplified scheme of a solvent donor. Under hydrobreaking conditions, H-donor diluent: GO and/or VGO can be recycled from cracked feedstock.

69

Residual Refining and Processing

Residue yields and the quality of this residue are sensitive to the nature of the donor solvent, especially to the feed-donor ratio and the amount of labile hy­ drogen transported by the donor, as shown in Fig. 35. An increase in hydrogen partial pressure also increases distillate yields and improves the quality of the residues in terms of sulfur content, and the viscosity is lower than with simple thermal operations. This process is well suited for the processing of heavy oils and bitumens rich in sediments, such as bitumen from Athabasca. Table 32 illustrates the distribution of products that can be expected from applying this type of process. Slurry Processes Under Hydrogen Pressure. These processes are on the border between thermal and catalytic processes under hydrogen pressure. They are listed in Table 33. All these processes have many similar aspects:

They are designed to implement a powder additive with a more or less large grain size. They all consume hydrogen and are mainly exothermic, even if the distillates ob­ tained still contain olefins. They (most of them) make very high conversions. This feature is linked to the need to concentrate the additive in a minimum of hydrocarbon residue, either to eliminate it or to recycle it when it is costly. To achieve very high conversions, the operating conditions applied are very se­ vere, not only with regard to residence time and temperature but also con­ cerning the hydrogen partial pressure, which must be sufficiently high to prevent thermodynamically and slow kinetically the development of polymer-

H 2 addition, scf/bbl of donor FIG. 3 5

Effect of the feed-donor ratio on the quality of the cracked product: R = feed-donor ratio, multiring donor; MN1 = modified naphtha insolubles test, which gives the propensity of asphaltenes to flocculate [41].

Residual Refining and Processing

TABLE 32

Typical Results o f Thermal Cracking in the Presence of H-Donor Solvent with Athabasca Vacuum Residue as Feedstock, DRB Process Syncrude Characteristics

Specific gravity Sulfur, wt% Viscosity at 40°C, cSt C3“ , wt% Product distribution, vol% Naphtha C4-190°C Distillate, 190-360°C Gas oil, 360-504°C Residuum, 504°C+

Raw Bitumen 1.0057 4.3 11.500 (2.5) 2

17.3 30.8 49.9

Low Conversion

High Conversion

0.969 3.3 45

0.912 1.9

1 .8

12.7 21.5 46.2 19.7

6 . 8

9.6 27.0 17.4 48.4 6.9

aGas composition, wt%: 21.2 hydrogen sulfide; 22.8 methane; 0.5 ethylene; 23.2 ethane; 4.8 pro­ pylene; 24.7 propane; 2.8 other (H2 + CO + C 0 2). Naphtha: 0 .9 wt% S; 0.0017 wt% N; bromine no. 22. Diesel: 4 .4 wt% S; cetane no. 35. Source: From Ref. 39.

ization and polycondensation reactions of the pericondensed aromatic hydro­ carbons initially present in the medium or formed by dealkylation. Little information is available about the exact way these additives work. As shown in Table 34, they do not change the distribution of cracked products, which remain appreciably the same as in other thermal cracking operations, that is, do­ nor solvent, hydrovisbreaking, and even visbreaking. On the other hand, they help diminish the formation of pitch and stabilize the operating of the unit. This can be explained if we assume that these additives are capable of playing the follow­ ing roles: The solid additive can produce activated hydrogen that will intervene in control­ ling various free radical reactions, including condensation reactions leading to the formation of pitch and then coke. This intervention may be direct, in which case the additive acts as a solid donor of hydrogen, which captures and neutralizes the heavy radicals that are coke precursors. These heavy radicals are adsorbed on the surface of the additive. Coal, lignite added initially, or the coke that appears during the reaction also play the role of seeds for coke pre­ cursor deposition. Moreover, by a spillover effect it is possible for this carbon deposit to become loaded with hydrogen, which is activated by underlying metal sulfides, slowing coke formation. The solid additive may also participate in hydrogenating unsaturated compounds and in catalytically hydrogenolyzing some bonds involving S and N heteroa­ toms. This results in smaller light olefin contents and higher desulfurization and denitrogenation rates than with hydrovisbreaking. The few differences observed in the performance of these different processes are linked to the hydrogen pressure applied, to the nature and concentration of the

D P P P D

Intevep Veba UOP Idemitsu Kosan

Exxon R&E

Asvahl

Asvahl

M-Coke (Microcat) [47]

Hyvahl C [48]

Tervahl C [49]

aP, pilot plant; D , demonstration plant; I, industrial.

D

Iron-rich clay Ni + V sulfides N i, V, Co, Mo sulfides on carrier VIA (or VIII) soluble salts VIA or VIII soluble salts VIA soluble salt

D D

Veba Intevep

Petro Canada

Canmet [42]

Veba-LQ-Cracking, Veba-CombiCracking [43] HDH [44] Aurabon [45] MRH [46]

Nature Iron compounds + subbituminous coal Iron compounds + lignite (optional)

Developmental State3

0 . 0 1

No

Yes

0 .0 3 -0 .0 5
90 60

8 0 -9 5

>90

NA 4 2 0 -4 8 0 4 2 0 -4 8 0

Conversion (Wt%) t ( h)

T (°C)

Operating Conditions

Residual Refining and Processing

TABLE 34

Heavy Oil Thermal Hydrocracking: Comparison o f Product Spectra

Additive Liquid yield, wt% Naphtha Middle distillate Vacuum gas oil Gas selectivity, wt% C, C3 C4

A 3.7 wt%

B 3.0 wt%

Without

17.7 39.2 43.1

16.9 39.0 43.1

17.2 39.3 43.5

26.6 25.1 29.4 18.9

28.7 26.0 28.5 16.8

28.8 23.2 28.1 19.9

Source: From Ref. 43.

additive, and the technology of implementation. As in all preceding processes, it is mainly at the level of residue yield and quality that these differences appear. The higher the hydrogen partial pressure, the fewer coke precursors are formed and the higher the distillate yield. The higher the additive content and specific activity in terms of the rupture of free radical chains, the fewer coke precursors are formed, the lower the gas yield, and the higher the distillate yield. These observations can be explained if we con­ sider, as for coking and visbreaking, that the reaction medium is very rich in rad­ icals, whose content can be estimated by electron spin resonance [50]. The jr. radicals involving nonlocalized electrons of polynuclear aromatic hy­ drocarbons are very stable and continue to exist during the storage of fuel oils. These are the radicals that, by a fairly slow condensation reaction, gradually lead to the formation of pitch. For example, the adding of molybdenum naphthenate considerably diminishes the free radical content in the medium, resulting in the slowing of all reactions but especially of the condensation reactions. To achieve a given conversion, greater severity (t and T) is necessary than in visbreaking. The higher the additive content and the more this additive is active in hydro­ genation, the fewer olefins there are, the more sulfur and nitrogen in the distil­ lates, and the more the additive particles loaded with nickel and vanadium. The performance of the additive depends on its chemical composition, physicochem­ ical characteristics, including the grain size, and specific activity with regard to various hydrorefining reactions. It remains for us to mention the technological features that enable the additive to be implemented without problems of operability and to eliminate better the heat of the reaction. Figure 36 shows the simplified flow of a hydroconversion process in the presence of slurry. Depending on the variants, the reaction section may have one or several reactors and may be equipped or not for recycling a fraction of the catalyst, as shown in Table 33. This partial recycling of the additive means that it must first have been separated from the residue in which it was in suspen­ sion. Several separation techniques have been proposed, such as decanting, pre­ cipitation by a solvent, filtering, and centrifuging; their respective interest depends on the characteristics of the solid and its cost. Reactor design deserves particular attention. The reaction volume must be de­ signed to provide the residence time required to ensure the desired conversion.

73

Residual Refining and Processing

H2 recycle

I

Reactor section

Cold separator

Hot separator LJ

..Diesel oil Fractionator

Possible solid with drawal . Heavy gas-oil

H2 makeup . Pitch additive preparation

FIG. 3 6

additive separation and recycling

Hydrothermal cracking processes in the presence of catalytic additives.

Likewise, the LID (length/diameter) ratio of this reaction volume must enable the superficial linear velocities of the liquid and gas required for the proper operation of the process. The linear liquid velocity must be suited to the grain size of the solid and to its specific gravity so that this solid does not become accumulated in the reactor. The linear velocity of the hydrogen gas must enable homogeneous bubbling flow of small bubbles to be established while preventing, inasmuch as possible, any heterogeneous turbulent flow of large bubbles to ensure good dif­ fusion of hydrogen into the liquid phase. The products issuing from these processes have characteristics very similar to those shown by products issuing from hydrovisbreakers and H-solvent donor pro­ cesses. Table 35 gives a series of results obtained in the processing of an Arabian Light vacuum residue. The sulfur, nitrogen, and olefin content of atmospheric dis­ tillates is such that hydrorefining is necessary before subsequent use and process­ ing. Vacuum gas oil may be sent directly to catalytic cracking without prior hydrorefining provided that its nitrogen content is no greater than 3,000 ppm. The special feature of the VEBA COMBI process should be noted here, in that the entire gas fraction issuing from the hot separator is hydrogenated directly before being distilled. Once again, it is important to note that the evolution of gas and distillate yields as a function of the conversion hardly changes from one process to another or even from one feed to another, according to whether it is a residue coming from a heavy oil or a conventional crude. However, each feed may be characterized by its tendency to be cracked, and heavy oils, which, to geochemists, are immature, are always easier to crack than residues from conventional crudes. Table 36 il­ lustrates this similarity of behavior of different residues from heavy oils when the process implemented is the Canmet. This type of process is actually particularly well suited for the deep conversion of heavy oils rich in metals, sediments, asphaltenes, and resins, which risk quickly blocking trickle beds in conventional hydrotreating units and plugging the pores of granular catalysts implemented in these processes. Such processes can

74

Residual Refining and Processing

TABLE 35

Conversion of Arabian Light Vacuum Bottoms (552°C+), Canmet Conditions3 Product Inspection

Characteristics

Naptha

LGO

HGO

Specific gravity at 15.6/15.6°C S, wt% N, wt% C, wt% H, wt% Concarbon, wt% C 5 asphaltenes, wt%

0.747 0.16 0.03 85.4 14.4

0.871 1.82

0.971 2.70 0.34 86.3 10.3

0.11 85.6 12.3

< 1

Product Yields (Conversion = 85 Wt%)

Pitch

— 4.27 0.78 87.5 6.5 63.5 78

Feed + additive

1 0 0 .0

h2

1 .6 8

Total Products H2S

1 0 1 .6 8

2.45

c,-c 3

6 .2 1

18.23 C4-204°C (naphtha) 204-343°C (LGO) 27.62 3 4 3 -5 24°C (HGO) 32.43 524°C+ (pitch) 14.74 ________________________________________________________________________________________ 101.68 0.31

aFeed characteristics: specific gravity at 15.6/15.6°C, 1.019; 4.34 wt% S; 0.26 wt% N; 22.3 wt% Concarbon; 13.5 wt% C5 asphalt; 155 ppm metals; 85.02 wt% C; 10.17 wt% H; viscosity at 100°C, 1.066 cSt. Source: From Ref. 42.

achieve volume yields of liquid distillates (520°C~) approaching 100% provided that one can accept a hydrogen consumption varying between 2.2 and 4 wt% com­ pared to the feed, depending on the pressure applied. For such high conversions, despite hydrogen pressure and the additive, a certain amount of carbon reject can­ not be avoided whatever the process. Therefore, some use must be found for this TABLE 36

Hydrocracking of Various Heavy Oils: Similarities in Product Distribution (Canmet Process) Properties of Feedstocks

Properties Feed Gravity, °API Ash, wt% Conradson carbon, wt% C/H, wt ratio Viscosity at 99°C, cSt Metals, ppm (wt) Pitch, wt% Sulfur, wt% Nitrogen, wt% Product yields and quality Distillate gravity, °API Sulfur, wt% Nitrogen, wt% C ,-C 4, wt% C5-204°C , vol% 204-343°C , vol% 343-524°C , vol% 524°C+ , vol% Source: From Ref. 42.

Athabasca Bitumen (260°C+)

8 .1

0.590 14.9 8.168 213 162 53.8 4.74 0.42 28.2 1.75 0.26 6.3 28.9 44.3 26.6 4.1

Lloydminster Heavy Oil (343°C+)

Boscan Heavy Oil (343°C+)

Cold Lake Bitumen (404°C+ )

Laguna Vacuum Bottoms (454°C+ )

9.9 0.035 13.7 8.136 188 223 54.6 4.34 0.31

7.7 0.240 16.7 7.958 576 1,311 66.7 5.73

6.4 0.064 18.2 8.059 1.489 357 73.0 5.16 0.57

6.7

28.3 1.67 0 .2 1 6 .1

23.0 43.8 36.6 5.4

0 . 8

29.2 2.43 0.43 6.3 25.0 43.4 30.4 5.7

27.6 1.76 0.32 6 . 8

26.4 42.9 28.0 6 . 6

0 . 1 2

18.4 7.707 — 655 81.6 3.43 0.61 28.6 1.30 0.31 9.5 26.8 36.1 33.1 6.5

Residual Refining and Processing

75

drag stream, which is a pitch having a high sulfur, nitrogen, metals, and coke content. One last application of a slurry catalyst is worth mentioning. It concerns the addition of minute amounts of a very active colloidal additive to a heavy oil treated under hydrovisbreaking conditions with a view to making it transportable by pipeline. Through a comparison with hydrovisbreaking results, Table 37 shows the viscosity and density improvements that can be expected from adding catalyst to the feed for a value of $ 0 .2 0 /b b l. Comparison is made at the same flocculation threshold (FT = 7). The additive is left in the syncrude.

Catalytic Hydrotreatm ents These are not the processes that annually process the largest amount of residues, but they are the ones that have been and still are being studied the most. From the practical standpoint, this intensive R&D effort has resulted in the fact that more than a dozen hydrotreatment processes using granular catalysts are now available on the market, as is shown in Table 38. These processes can be classified in three categories, depending on whether the catalyst is implemented in a fixed, moving, or ebullating bed. It is true that adding on a catalyst leads to a much more complex chemistry than seen previously in the hydrovisbreaking or H-solvent donor processes. The presence of a catalyst initiates or speeds a wide range of competitive reactions that intervene at the same time or consecutively. These reactions are diagrammed in Fig. 37. Partial or total hydrogenation of aromatic nuclei and heterocycles (HDAr). This stage is useful for making aromatic and heterocyclic rings less stable. These aromatic compounds are known to be stabilized by resonance as, to a lesser extent, are compounds of the thiophene, pyrrole, and fur an types. This stage TABLE 37

Catalytic Hydrovisbreaking (HVB) of a Topped Boscan Crude3 G3S c,-c4. h2 s. nh3 IBP : ?50°C Jl ~ \ J ........... Crude from desalter

_ J _ . T oddIng > Mydro1 PP’ g |,__ *visbreaking 1--------------1 j i Bottom Cataly.st

Syncrude Characteristics (FT = 7) c ,- c 4 C5,