Congress on Research, Development, and Innovation in Renewable Energies: Selected Papers from CIDiER 2023 (Green Energy and Technology) 3031521706, 9783031521706

Citation preview

Green Energy and Technology

Mayken Espinoza-Andaluz · Ester Melo Vargas · Jordy Santana-Villamar · Ángel Encalada Dávila · Brayan Ordóñez Saca  Editors

Congress on Research, Development, and Innovation in Renewable Energies Selected Papers from CIDiER 2023

Green Energy and Technology

Climate change, environmental impact and the limited natural resources urge scientific research and novel technical solutions. The monograph series Green Energy and Technology serves as a publishing platform for scientific and technological approaches to “green”—i.e. environmentally friendly and sustainable—technologies. While a focus lies on energy and power supply, it also covers "green" solutions in industrial engineering and engineering design. Green Energy and Technology addresses researchers, advanced students, technical consultants as well as decision makers in industries and politics. Hence, the level of presentation spans from instructional to highly technical. **Indexed in Scopus**. **Indexed in Ei Compendex**.

Mayken Espinoza-Andaluz • Ester Melo Vargas Jordy Santana-Villamar • Ángel Encalada Dávila Brayan Ordóñez Saca Editors

Congress on Research, Development, and Innovation in Renewable Energies Selected Papers from CIDiER 2023

Editors Mayken Espinoza-Andaluz Center for Renewable and Alternative Energies Escuela Superior Politecnica del Litoral Guayaquil, Ecuador Jordy Santana-Villamar Faculty of Mechanical Engineering and Production Science Escuela Superior Politecnica del Litoral Guayaquil, Ecuador

Ester Melo Vargas Faculty of Social and Humanistic Science Escuela Superior Politecnica del Litoral Guayaquil, Ecuador Ángel Encalada Dávila Faculty of Mechanical Engineering and Production Science Escuela Superior Politecnica del Litoral Guayaquil, Ecuador

Brayan Ordóñez Saca Faculty of Mechanical Engineering and Production Science Escuela Superior Politecnica del Litoral Guayaquil, Ecuador

ISSN 1865-3537 (electronic) ISSN 1865-3529 Green Energy and Technology ISBN 978-3-031-52170-6 ISBN 978-3-031-52171-3 (eBook) https://doi.org/10.1007/978-3-031-52171-3 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable.

Preface

The effects of climate change have grown over time, such as the increase in global temperature, shifting weather seasons, and the rise of meteorological phenomena worldwide. Every change is accelerated by the significant greenhouse gas emissions released to the atmosphere using fossil fuels in the energy sector. Due to this, international alliances, for instance, the European Green Deal, have arisen to face the challenges of changing to a sustainable, clean, and economically viable energy matrix. In addition, one of the main concerns of the United Nations 2030 Agenda for Sustainable Development is to promote sustainable initiatives related to developing renewable energies, such as SDG7: Affordable and Clean Energy and SDG 13: Climate Action. Therefore, disseminating scientific and technical works that help meet the objectives of sustainable development and, in turn, help mitigate the effects of climate change is essential. Based on the success of CIDiER 2021 and CIDiER 2022, we are pleased to bring the selected papers from the third meeting of the Congress on Research, Development, and Innovation in Renewable Energies, CIDiER 2023, held on September 18–19, 2023, in Guayaquil, Ecuador. This book looks at biomass, wind energy, geothermal energy, hydrogen energy, tidal energy, computational modeling, and artificial intelligence applied to renewable energy studies. This 2023 Congress aimed to spread theoretical and experimental study results and applications related to relevant topics in renewable energies and generate a space of multidisciplinary interaction to strengthen and establish new contact networks in the academic and research community. Participants from all disciplines related to renewable energy and several Latin American countries contributed to this important event. This edited book examines some essential topics related to renewable energies, such as developing and studying new materials,

v

vi

Preface

process design and simulation, computational modeling, energy efficiency, and chemical and mechanical analysis. Finally, the editors of this book would like to thank the Springer editorial team and all contributing authors for their commitment and collaborative effort that have made this book possible. Guayaquil, Ecuador

Mayken Espinoza-Andaluz Ester Melo Vargas Jordy Santana-Villamar Ángel Encalada Dávila Brayan Ordóñez Saca

Contents

Part I

Artificial Intelligence

Demand-Side Management Integrating Electric Vehicles Using Multi-step Forecaster: Santa Elena Case Study . . . . . . . . . . . . . . Juan C. Guamán, Edwin Celi, Johnny Rengifo, Fernando Vaca, and Manuel S. Alvarez-Alvarado Long-Term Sustainable Energy Transition of Ecuador’s Residential Sector Using a National Survey, Geospatial Analysis with Machine Learning, and Agent-Based Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . Diego Moya, César Arroba, Christian Castro, Cristian Pérez, Dennis Copara, Alexander Borja, Sara Giarola, and Adam Hawkes Part II

3

23

Computational Modeling

Analysis of Incidence of Angle of Attack on Energy Efficiency of a Two-Dimensional Airfoil NACA 1412 . . . . . . . . . . . . . . . . . . . . . . . Luis Gonzaga-Bermeo, Carlos A. Cuenca, Jorge E. Game, and Bristol E. Carriel Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador: Simulation Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Juan Carlos Solano, Valeria Herrera, Ángel Ordóñez, Miguel Caraballo, and Aníbal Lozano The Role of Curved Buildings in Urban Wind . . . . . . . . . . . . . . . . . . . . Carlos Walter and Jorge Lässig

43

57

71

vii

viii

Part III

Contents

Miscellaneous

Advances in H-Type Darrieus Turbines for Urban Environments in Colombian Territory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carlos V. M. Labriola, Andres F. Galindo Rojas, Elizabeth A. López, and Javier A. Rosero García Assessment of Green Hydrogen Production from Hydropower in Ecuador . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ángel Recalde, Víctor Acosta, Giordy Ortiz, Ricardo Cajo, and Carolina Godoy

87

99

Feasibility of Shallow Geothermal Installations for Cooling Purposes in Tropical Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Mariana Villafán-Sierra, Daniela Blessent, Jacqueline Lopez-Sanchez, Carlos Ernesto Arrieta-Gonzalez, and Mauricio Gonzalez-Palacio Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines in Developing Countries with a Focus on Colombia’s Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Brandon Martínez, Carlos Arrieta, Ainhoa Rubio, Mario Luna, Hernando Yepes, Edwin Chica, Laura Velásquez, and Juan Pablo Gómez Montoya Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Garcés Anggie, Ramírez Charles, Juan Peralta-Jaramillo, Emérita Delgado-Plaza, Jorge Abad-Moran, Jorge Hurel, Guido Abril, and Ian Sosa Techno-economic Analysis for the Valorization of Palm Kernel Shell via Hydrothermal Carbonization and Anaerobic Digestion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Carolina Rueda, Sebastián Ponce, and Herman Murillo Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

About the Editors

Mayken Espinoza-Andaluz is a full professor at Escuela Superior Politécnica del Litoral (ESPOL) and co-supervisor of PhD students in heat transfer and energy systems at Lund University. He earned his PhD in heat transfer from the Department of Energy Sciences at Lund University. His areas of research include renewable energy, fuel cells, proton-exchange membrane fuel cells (PEMFC), solid oxide fuel cells (SOFC), physical and chemical phenomena, energy efficiency, Lattice Boltzmann methods (LBM), and computational modeling. He is a reviewer for several international journals, including the International Journal of Energy Research, International Journal of Hydrogen Energy, and Computers and Mathematics with Applications. He is the Second Vice-President of the Hydrogen Ecuadorian Association and the current official speaker about Hydrogen and Fuel Cells from ESPOL. Ester Melo Vargas received her bachelor’s degree in audit and management control in 2021 from the Escuela Superior Politécnica del Litoral (ESPOL), obtaining ad honors diploma and an award for her thesis project related to the management of medical supplies in Ecuador during the Covid-19 pandemic. From 2019 to the present, she has served as a research technician at the same institution. Her interest in research, technology, and gender inclusion in STEM has motivated her to participate in different technological innovation competitions and STEM projects sponsored by the US Embassy in Ecuador. This experience allowed her to intern in project engineering at Huawei Technologies (Guayaquil, Ecuador) in 2020. Her research interests include data analysis and fault detection applied to fuel cells, quality analysis, and lean manufacturing applied to productivity management. In addition, Ms. Melo is the author and co-author of scientific articles presented at congresses, conferences, and scientific dissemination meetings.

ix

x

About the Editors

Jordy Santana-Villamar received his bachelor's degree in mechanical engineering from Escuela Superior Politécnica del Litoral (ESPOL) in 2020, obtaining a diploma of distinction in research. He was a teaching assistant for four semesters during his student stage on subjects such as calculus, electromagnetics, physics, and statistics. Jordy is doing his master's degree in materials science and engineering and performing a graduate research assistant position at the Faculty of Mechanical Engineering and Production Sciences, ESPOL. He works on polymer exchange membrane fuel cell (PEMFC) characterization through different electrochemical techniques. His work also encompasses expanded graphite-resin composite materials development with secondary fillers for bipolar plates and the synthesis of biodegradable membranes based on chitosan. He has ten scientific papers published in high-impact journals. Furthermore, he has participated in international conferences and recently did a research stay at the Instituto de Carboquímica-CSIC in Zaragoza, Spain, addressing PEMFC stack conditioning for hydrogen fuel cell vehicles and large-scale vanadium redox flow batteries. Ángel Encalada Dávila received his bachelor’s degree in mechatronics engineering in 2021 from Escuela Superior Politécnica del Litoral (ESPOL). He works as a data scientist in an international data analytics company. He also has solid experience in the software industry for ATMs. Furthermore, he collaborates as a data scientist at the Control, Modeling, Identification, and Applications Research Group (CoDAlab) at the Universitat Politècnica de Catalunya (UPC), Barcelona, Spain. His research interests include structural health monitoring (SHM), fault prognosis applied to wind turbines, and computer modeling used to fuel cells. In addition, Mr. Encalada is the author and co-author of several scientific articles published in journals like Sensors, Energies, and Processes. He recently worked as an organizer of the third Congress of Research, Development, and Innovation in Renewable Energies (CIDiER 2023), Guayaquil, Ecuador. Brayan Ordóñez Saca received his bachelor’s degree in mechanical engineering in 2022 from Escuela Superior Politécnica del Litoral (ESPOL). He works as a research assistant in the same institution’s Laboratory of Renewable Energy Sources. Mr. Ordoñez also has experience in the production of energy through generator sets. Furthermore, he developed an investigation about modeling a hybrid dryer at a laboratory scale. He participated in a research project about studying the fluid dynamic behavior of a hybrid dryer. His research interests spin around fuel cells, specifically proton exchange membrane fuel cell, and their performance study applying electrochemical tests. On the other hand, now, he is working on expanded graphite-resin composite materials development with secondary fillers for bipolar plates. He recently went to work as an editor of the third Congress of Research Development and Innovation in Renewable Energies (CIDiER), Guayaquil, Ecuador.

Part I

Artificial Intelligence

Demand-Side Management Integrating Electric Vehicles Using Multi-step Forecaster: Santa Elena Case Study Juan C. Guamán, Edwin Celi, Johnny Rengifo and Manuel S. Alvarez-Alvarado

, Fernando Vaca

,

1 Introduction The transport sector has the highest dependence on fossil fuels of all industries. It represents 37% of CO2 emissions worldwide, so it is imperative to reduce them to mitigate the adverse effects of climate change [1]. Electric vehicles (EVs) represent an exciting alternative to replace conventional vehicles based on internal combustion engines and have been experiencing sustained growth in recent years [2]. Focusing on Ecuador’s context, the Organic Law of Energy Efficiency states that by 2025, all vehicles incorporated into the urban public transport service must be electricpowered [3]. The massive integration of EVs may have a negative impact on the electrical distribution network, affecting the load profile and the capacity of the system components. Consequently, this may lead to voltage imbalances and harmonic injection that could decrease power quality [4]. It is relevant to mention that in Ecuador, the Agency for the Regulation and Control of Energy and Non-Renewable Natural Resources (ARCERNNR), through regulation No. ARCERNNR -002/20, establishes reference values for power quality indices such as voltage level, rapid disturbances of voltage, voltage harmonic distortion, and voltage unbalance, which are mandatory compliance by distribution companies [5].

J. C. Guamán · E. Celi · F. Vaca · M. S. Alvarez-Alvarado Facultad de Ingeniería en Electricidad y Computación, Escuela Superior Politécnica del Litoral, Guayaquil, Ecuador J. Rengifo (*) Departamento de Ingeniería Eléctrica, Universidad Técnica Federico Santa María, San Joaquín, Santiago, Chile e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_1

3

4

J. C. Guamán et al.

Demand-Side Management (DSM) is a critical approach in modern power systems for optimizing energy consumption patterns, enhancing grid stability, and promoting efficient utilization of resources. Various techniques have been proposed and implemented to effectively manage energy demand, considering factors such as peak load reduction, cost savings, and environmental sustainability. For instance, load shifting involves redistributing energy demand from peak periods to off-peak times. Time-of-use pricing is a common approach where electricity prices are higher during peak hours to incentivize consumers to shift their usage. Authors in [6] provide insights into load-shifting strategies and their role in demand response programs. Another technique used for DSM is peak demand reduction, which aims to reduce the highest points of electricity consumption to avoid grid overloading. Demand response strategies, including direct load control and pricebased incentives, are implemented to curtail peak demand. Some authors [7] discuss various peak-shaving techniques and their impacts on the power system. Demand Flexibility and Energy Storage can also be used for DSM purposes, as demand flexibility involves adjusting energy consumption in response to grid signals. Demand response aggregators use this flexibility to balance supply and demand. Energy storage, like batteries, enables storing excess energy during off-peak periods for later use in [8] a review of demand response strategies and energy storage technologies in modern smart grids. Control-based DSM techniques such as model predictive control and fuzzy logic can also be applied, which optimize energy consumption considering various constraints and objectives. These techniques enable real-time adjustments based on system conditions. It is relevant to mention that [8] discusses advanced control methods in demand response applications. In addition, integrating renewable energy sources into DSM can help align energy consumption with clean energy availability. Smart grid technologies enable real-time monitoring and control of distributed energy resources. A study in [9] investigates the integration of renewables in DSM frameworks. Integrating Artificial Intelligence (AI) techniques into Demand-Side Management (DSM) has emerged as a transformative approach to optimize energy consumption, enhance grid efficiency, and facilitate the integration of renewable energy sources. AI offers sophisticated tools for real-time analysis, decision-making, and control in complex energy systems. Machine learning algorithms, such as neural networks, support vector machines, and decision trees, enable the analysis of historical data to predict future energy consumption patterns. These predictions aid load forecasting and scheduling, allowing utilities to allocate resources efficiently. Authors in [10] explore the integration of machine learning in DSM for demand prediction. Concerning reinforcement learning for dynamic control, this enables adaptive control strategies by learning optimal actions through environmental interactions. These techniques are suitable for managing dynamic energy consumption patterns and optimizing demand response actions. In reference [11], reinforcement learning-based DSM for peak demand reduction can be found. Genetic algorithms provide optimization solutions by mimicking natural selection and evolution processes. They are employed to determine optimal load-shifting strategies, demand response schedules, and energy resource utilization. A study conducted in [12]

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

5

applies genetic algorithms to solve DSM problems. AI-driven energy management systems leverage real-time data to optimize energy consumption based on user preferences, cost constraints, and grid conditions. These systems enable smart appliances and devices to manage their energy usage autonomously. In reference [13], AI-based energy management systems in DSM applications are discussed. Finally, deep learning models, such as convolutional neural networks and recurrent neural networks, are employed for anomaly detection in energy consumption. These models identify abnormal usage patterns that might indicate equipment malfunctions or inefficiencies. Authors in [14] explore deep learning techniques for anomaly detection in DSM. Focusing on EV demand management, this is a potential solution to enhance power quality and system reliability indices. An advantage of EVs is that they can act as distributed storage devices and deliver power to the network through the scheme known as Vehicle to Grid (V2G) [15]. V2G can offer auxiliary services, such as voltage regulation, frequency, load balancing, maximum power, and support for renewable energy resources. Despite its benefits, using V2G affects batteries’ life cycle and can accelerate their degradation. However, with an adequate battery charge and discharge strategy through a management system, said degradation can be minimized [16, 17]. For such management, it is necessary to configure the V2G scheme as bidirectional, i.e., it can supply or receive energy from the network [18]. In addition, the battery charge and discharge strategy must incorporate the State of Charge (SOC) analysis, which represents the available capacity expressed as a percentage of its nominal capacity [19]. For instance, in [20], the impact of plug-in EVs on the peak demand in a commercial microgrid is analyzed. A case study with four scenarios considering demand management for charging and discharging EVs is proposed. The Nissan Leaf is taken as a reference EV, whose battery has an available capacity of 16 kWh. It is concluded that implementing a management system is essential to reduce the peak demand in the network, which must incorporate a V2G or similar scheme. Potential congestion and stress problems in the components of a low-voltage residential distribution network in Australia are well described in [21], which focuses on voltage deviation and network losses. In such an investigation, an algorithm for the charging strategy of EVs and Particle Swarm Optimization (PSO) is established as demand management mechanisms for the optimal location of charging stations throughout the network distribution. To simulate the impact of EVs on the network, MATLAB and OpenDSS are used. The results led to a reduction of 15.32% and 39.38% in voltage deviation and losses, respectively, demonstrating the potential of demand management. The authors in [22] propose a multi-agent system comprising the EV users known as the aggregator and the distribution system operator. The role of the aggregator is to collect historical EV charging data, applying Support Vector Machines to achieve accurate forecasting of future EV charging demand. Such a study incorporates a typical distribution network of the United Kingdom (UK), 33/11/0.4 kV, and the results show that it is possible to reduce the peak demand of the EVs in the load curve. In [23], the implementation of a demand management system is investigated. The need for Smart Grids to

6

J. C. Guamán et al.

incorporate renewable energy sources and an Advanced Metering Infrastructure (AMI) that facilitates grid management is highlighted. The approach uses a threephase distribution network model at 0.4 kV, with a high penetration of EVs. Sixteen AMI measurement points to record the voltage and current values in each of the three phases. In addition, it incorporates an algorithm developed in Python that uses the AMI measurements as inputs and places the minimum normative values of voltage and current as constraints to implement the DMS. The results reveal that implementing the DMS allows an increase of almost 15% in EVs without affecting the current state of the network distribution components. Santa Elena’s (a coastal region in Ecuador) demand is transitioning, as in the next decade, the distribution network will present a high penetration of EVs. As a contingency plan, this paper proposes a DMS strategy applied to the representative feeder, obtained using an unsupervised classification method named “K-means.” The optimization is performed using CYME® executed from Python. This chapter contributes to a novel algorithm that reduces the computational burden by using a representative demand. Moreover, an optimization using PSO to perform DMS is performed, resulting in power quality metrics within the permitted limits given in the Ecuadorian standard No. ARCERNNR -002/20 [5]. Given the above, the rest of the chapter is structured as follows: Sect. 2 describes the process used to preset the data for the case study; Sect. 3 presents the demand-side management strategy used to optimize the demand curve; Sect. 4 presents a critical analysis of the results; finally, Sect. 4 brings the conclusions.

2 Methodology 2.1

Case Study Description

To simulate the incorporation of EVs into a representative distribution network of Santa Elena, information is obtained from 64 feeders of 13.8 kV of governmental company CNEL EP UN Santa Elena. The following parameters were collected: overhead and underground network length, number of users, active power, and reactive power demands. This information is used to classify the medium voltage feeders according to the characteristics described in Sect. 3.

2.2

Preset Data

Pre-processing data is a crucial step in data analysis and machine learning. It involves cleaning, transforming, and organizing the data to make it suitable for analysis and modeling. The process starts with “Data Cleaning” [24], which is employed to identify and handle missing data, either by removing the rows or columns with missing values or by inputting the missing values with appropriate

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

7

techniques (e.g., mean, median, or interpolation). In addition, it enables to detect and deal with outliers that may skew the analysis or modeling results. Outliers can be removed, transformed, or imputed based on the specific context. The next step is “Data Transformation” [24], which is used to scale the numerical features to a similar range to avoid the dominance of certain features during modeling. Common scaling techniques include Min-Max scaling or standardization (Z-score scaling). In this step, the conversion of categorical variables into numerical representations (e.g., one-hot encoding, label encoding) is performed with a view to be used by machine learning algorithms. This opens a pathway to create new features or derive meaningful information from existing features to enhance the model’s performance. This might involve extracting date features, grouping data, or continuous variables. The third step is “Data Reduction” [24], in which dimensionality reduction techniques like Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor Embedding (t-SNE) are employed to reduce the number of features while preserving important information. This is followed by “Data Normalization” [24], ensuring normal data distribution. Various normalization techniques, such as log or Box-Cox transformations, can be applied. Finally, the process finished with Data Splitting [24], which splits the data into training and testing sets to effectively evaluate the model’s performance. The training set is used to train the model, while the testing set is used to assess its generalization ability. Data pre-processing is a flexible and iterative process that depends on the specific characteristics of your dataset and the requirements of the analysis or modeling task. The choice of pre-processing steps will vary for different datasets and machine learning algorithms. Experimenting with different approaches and evaluating their impact on model performance are crucial to achieving the best results.

2.3

Elbow Method

The elbow method is a popular technique used for determining the optimal number of clusters in a clustering algorithm, such as K-means. The method is straightforward and involves plotting the variance (or within-cluster sum of squares) against the number of clusters used in the clustering process. The objective is to identify the “elbow point” on the plot, which represents the point of diminishing returns in terms of variance explained as the number of clusters increases. A step-by-step description of the elbow method is as follows [24]: 1. Choose the Range of Clusters: To apply the elbow method, you need to decide on a range of cluster numbers that you want to test. This range typically spans from a minimum number of clusters (e.g., 2 or 3) to a maximum number that makes sense given the dataset and problem. 2. Apply Clustering Algorithm: Run the chosen clustering algorithm (e.g., K-means) on the data for each cluster number in the defined range. The algorithm will

8

3.

4.

5.

6.

J. C. Guamán et al.

partition the data into the specified number of clusters and calculate the WithinCluster Sum of Squares (WCSS) or variance for each clustering result. Calculate WCSS/Variance: For each clustering result, calculate the WCSS (or variance) as a measure of how compact the data points are within each cluster. The lower the WCSS, the more homogeneous the clusters are. Plot WCSS/Variance vs. Number of Clusters: Create a line plot or a scatter plot with the number of clusters on the x-axis and the corresponding WCSS or variance on the y-axis. As the number of clusters increases, the WCSS tends to decrease, as smaller clusters can fit the data more tightly. The plot may show a decreasing trend, but the goal is to find the point where adding more clusters leads to only marginal improvements in WCSS. Identify the Elbow Point: Examine the plot to find the “elbow point,” which is the location where the WCSS curve starts to flatten out. The elbow point indicates the optimal number of clusters, as it strikes a balance between maximizing cluster homogeneity and minimizing the number of clusters used. Select the Optimal Number of Clusters: Choose the number of clusters corresponding to the elbow point as the optimal number for your clustering analysis. This number should provide a reasonable trade-off between complexity and meaningful separation of data points into clusters.

It is essential to note that the choice of the optimal number of clusters is not always straightforward, and sometimes, there may not be a clear elbow point. In such cases, additional techniques like silhouette analysis or domain knowledge may be helpful in determining the appropriate number of clusters. The elbow method is a useful starting point for exploring the number of clusters, but it should be used in combination with other validation methods to ensure that the clustering results are meaningful and valuable for the specific problem at hand.

2.4

K-Means

The K-means method is a popular unsupervised machine learning algorithm used for clustering data into K distinct groups or clusters. It aims to partition the data points into clusters in such a way that each data point belongs to the cluster whose center (centroid) is closest to it. The algorithm minimizes the Within-Cluster Sum of Squares (WCSS) or variance, making it sensitive to the Euclidean distance between data points and the cluster centroids. The algorithm starts with the selection of the number of clusters, then the centroids are randomly generated using a specific heuristic (e.g., K-means++ initialization) to serve as the starting points for the clusters. For each data point xi in the dataset, calculate the Euclidean distance between the data point and each centroid. The distance formula between a data point xi and a centroid cj is given by [24]:

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . . n

ðk Þ

ðk Þ

xi  cj

d xi , cj ¼

9

2

ð1Þ

k¼1

This is followed by the determination of the new centroids for each cluster by computing the mean value of the data points assigned to that cluster. The new centroid cj for cluster j is calculated as follows [24]: cj ¼

1 cj

ð2Þ

xi xi ECj

Where |cj| is the number of data points in cluster j. The assignment is repeated and updated at every iteration until one of the stopping criteria is met. Common stopping criteria include a maximum number of iterations or when the centroids no longer change significantly between iterations. The algorithm will converge to a final set of K centroids, and the data points will be clustered based on their proximity to these centroids. The K-means algorithm aims to minimize the Within-Cluster Sum of Squares (WCSS) or variance, which is given by [24]: K

xi  cj

WCSS ¼

2

ð3Þ

j¼1 xi ECj

The K-means algorithm is simple and efficient but sensitive to the initial centroid positions, and it may converge to suboptimal solutions depending on the initialization.

2.5

Optimization

The demand-side management is carried out using Particle Swarm Optimization (PSO) [25]. Two scenarios are proposed in which the arguments of the objective function for minimization vary. The first scenario is used as a benchmark and corresponds to the minimization of the objective function for flattening the demand curve through the decrease of the peak and valley filling of the feeder load profile, which mathematically is described as follows [26, 27]: T

T

Pg ðt Þ  Ploss ðt Þ þ

min t¼1

2

n

Pev ðk, t Þ  Pavr t¼1 k¼1

ð4Þ

10

J. C. Guamán et al.

where Pg represents the active power available at the vehicle charging power station, while Ploss represents the active power losses; corresponds to the predictive average of active power during period T; Pev is the active power consumed by EV. The second scenario is similar to the first scenario but also includes the reactive power losses of the grid located downstream from the electric power vehicle station Qloss. Therefore, the objective function can be described as follows [28]: T

ðPg ðtÞPloss ðtÞÞþ

w1 t¼1

T

þw2

T

2

n

Pev ðk , t ÞPavr t¼1 k¼1

ð5Þ

Qloss ðt Þ t¼1

The variables w1 and w2 are weighted factors that allow the integration between the active and reactive power, which are given in VA/W2 and VA/VAr. The constraints to be considered for the optimization problem are the following: T

Pg ðt Þ  Pload ðt Þ  Pev ðt Þ  Ploss ðt Þ ¼ 0

ð6Þ

Qg ðt Þ  Qload ðt Þ  Qev ðt Þ  Qloss ðt Þ ¼ 0

ð7Þ

t¼1 T t¼1 T

SOCf ¼

SOCo ðt Þ þ Pev ðt ÞΔt

ð8Þ

t¼1

SOCmin  SOCf  SOCmax

ð 9Þ

V max  V ðabcÞ  V min

ð10Þ

Where SOCo and SOCf represent the initial and final State of Charge of the battery. The constraints given in (5) and (6) follow the power-balanced equations in which the generated power (active and reactive) must satisfy the loads and power losses of the system [29]. Concerning the batteries, it is necessary to determine the current state of charge, which is quantified using (7). The restriction presented in (8) avoids further lifetime reductions. Hence, the SOC value should be kept within a specified range [30]. It is assumed that the EVs connected to this network are mainly used for public transportation (buses and taxis). Consequently, every EV will arrive at the charging station with initial states of charges SOCmin between 30% and 35%. Finally, restriction (9) states that the voltage at each phase (a, b, and c) in each node of the grid must be within the permitted range established by national standards [6].

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

2.6

11

DMS Algorithm

To implement the DMS in the “Libertad” feeder, it is required to integrate the CYME® electrical network analysis program with Python to automate the execution of power flows based on the location and forecast of the number of EVs. In addition, the objective function and the restrictions of the problem are considered to configure the electric station as a generator and supply energy to the feeder by applying the V2G scheme. Figure 1 presents the Flow Diagram to implement the DMS in the “Libertad” feeder. According to the objective function selected, the particles are the hourly active power or active and reactive power profiles. Thus, the optimization solution will define the V2G performance.

2.7

Representative Feeder Selection

The network feeders are clustered using the K-means unsupervised machine learning method. To cluster N points into K groups, K-means uses a distance measure and averages (means) to adjust the position of the clusters, where each cluster has a remarkable point called centroid [24]. The number of clusters required for the database is determined using the “elbow” criterion, which uses the average distance to the centroid as a function of K and finds where the rate of descent sharpens. The results (see Fig. 2) reveal that K is determined as equal to 4, and by running K-means in Python, the centroids of the four groups are determined as detailed in Table 1. Groups 1, 2, and 4 comprise feeders that supply electricity to rural areas or industrial customers. Since the most significant penetration of EVs will occur in the urban area, group 3 is taken to be representative urban feeders, and therefore, the “Libertad” feeder is established as the characteristic feeder.

2.8

Libertad Feeder: Voltage and Demand Profiles

The DMS algorithm uses the Libertad feeder’s active and reactive power demand to define the EV charging performance. Hence, forecast data are used during the optimization stage. A 1-week data collection (in quarter-hourly records) of the active and reactive load profiles of the Libertad feeder is used to train a Random Forest [31] model and obtain the percentage of coverage, which is reflected in Fig. 3; this figure shows the evaluation of the active and reactive power forecasting models with a coverage percentage prediction of 94.79% and 91.32%, respectively. As a result, Fig. 4 shows the 1-day prediction for the active and reactive power of the feeder with coverage percentages higher than 80% of the prediction models. By recording the feeder head-end voltage values, in the period of the actual demand values, a forecast of the line-to-neutral voltage parameters at the 13.8 kV

12

Fig. 1 Proposed demand-side management algorithm results and discussion

J. C. Guamán et al.

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

13

107

0 -2 -4

Score

-6 -8 -10 -12 -14 -16

0

2

4

6

8

10

12

Number of clusters Fig. 2 The elbow curve was obtained from K-means Table 1 Feeder classification results Group 1 2 3 4

Number of feeders 10 17 13 22

(a) Forecasting model prediction coverage

Representative feeder (Centroid) Punta Carnero Progreso Libertad Sacachun

(b) Forecasting model prediction coverage

Fig. 3 Forecasting model prediction assessment: (a) active power, (b) reactive power

bus of the feeder is obtained. For instance, Fig. 5 shows the 1-day forecasting of phase-to-neutral voltage on phase C, resulting in a prediction coverage percentage of 92.71%.

14

J. C. Guamán et al.

(a) Forecasting model prediction coverage

(b) Forecasting model prediction coverage

Fig. 4 Forecasting model prediction assessment: (a) active power, (b) reactive power Fig. 5 Forecasting of voltage line to ground of phase “C”

Fig. 6 Electric vehicle charging station profile

2.9

Expected Number of Electric Vehicles

To incorporate the demand for the charging stations, access to the charging curve data of a representative EV charging station in Guayaquil was obtained through the CNEL EP UN Guayaquil distributor. Figure 6 shows the weekly charging profile for the BYD power station in the Samanes Park sector, Guayaquil, Ecuador. The given

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

15

Fig. 7 Expected number of electric vehicles connected to the grid

data correspond to the charging of K9G electric buses used for public transportation, which present 324 kWh batteries with a charging time ranging from 4 to 5 h. By applying (1), the number of EVs expected to be charged during the period of analysis T is obtained. Notice that such expression also depends on the projected consumption of the Electric Vehicle Power Station (EVPS) and the total number of charging stations. Applying (1) to the Libertad feeder, the electric vehicles expected to be charged are shown in Fig. 7, expected number of electric vehicles connected to the grid. The given information in Fig. 7 is employed as input data for applying the demand-side management strategy, which is expected to bring the active and reactive power variables requested by the system to comply with the established restrictions by charging stations in use in each analysis period.

2.10

Location of the Electric Vehicle Power Station

To evaluate the impact of the demand of the electric station on the representative feeder and its subsequent management, the location of the EVPS is at the center of the primary feeder (node MTA_S_8257 see Fig. 8). The EVPS has a lagging power factor of 0.9 and is fed at low voltage by a 500 kVA, 13,800/440 V, three-phase transformer. For the modeling of the electric station in the CYME® power network analysis program, it was represented as a concentrated load, drives, and BESS; attributes and input data of these devices will be modified using Python for each period of the optimization analysis.

16

J. C. Guamán et al.

Fig. 8 Electric vehicle power station located at MTA_S_8257

Fig. 9 Active power behavior at the electric vehicle power station: (a) Scenario 1; (b) scenario 2

3 Results and Discussion It is assumed that the EVs connected to this network are mainly used for public transportation (buses and taxis). Consequently, every EV will arrive at the charging station with an initial SOC between 30% and 35%. By applying the proposed approach (considering both objectives’ functions) on the Libertad feeder, the active power management for the two scenarios previously described is presented in Fig. 9a, b, respectively. Both figures present three curves in three different colors. The curve in yellow corresponds to the charging station without energy management according to the local measurement given by the industry; the curve in blue is active

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

17

Fig. 10 Reactive power losses downstream from EVPS

power consumption using the energy management proposed; and the curve in cyan represents the average demand power used as a benchmark to compare the optimal power management. The load peaks in maximum demand intervals are reduced for both scenarios to reach the optimal solution. Moreover, between midnight and 05 h00, the number of EVs connected to the charging station allows it to reach the benchmark profile. While between 10 h00 and 23 h59, few EVs that require a charge are available. Hence, the load–demand curve remains almost the same. It is relevant to mention that between 05 h00 and 10 h00, the active power management shows a different performance produced by including the reactive power losses into the objective function. For the second scenario (Fig. 9a), the system reduces its ability to fill the valley with a higher rate of change than in the first scenario (Fig. 9b). Figure 11 reveals that voltage level at the head of the feeder, the number of EVs connected to the network, and random SOC between 30% and 35% of each connected vehicle, it is possible to develop demand management following with the restrictions previously described. In addition, it can be appreciated that the 1-day demand forecast (red line) presents peaks in the interval from 00 h00 to 05 h00, which decreases its value from 06 h00 to 20 h00, where, again in the period from 20 h00 to 23 h59, the consumption of the network increases due to the charging of EVs. Analyzing reactive power losses, Fig. 10 shows a slight decrease in reactive power losses, mainly in the hours when there are more EVs (hours 0–5, 23). Additionally, the reduction of reactive power losses causes a minor improvement in the voltage profile of the entire feeder, given the reactive power injection performed by the management algorithm to reduce reactive losses. Figure 11 shows the voltage profile of phase A at node MTA_S_85347, which is the most distant node electrically from the feeder head end.

18

J. C. Guamán et al.

Fig. 11 Voltage phase to ground profile at the charging station

4 Conclusions This chapter proposes implementing a demand-side management algorithm to include an electric vehicle charging center in the Libertad feeder of Santa Elena in Ecuador. The strategy uses classification and pattern identification methods successfully. The results allow identifying strengths, challenges, and opportunities to be considered for the integration of electric vehicles in a distribution network in the coastal region. For instance, the proposed approach classifies all the feeders of CNEL Santa Elena into four groups according to their main characteristics. The “Libertad” feeder turned out to be the most representative of the urban feeders in the region. Subsequently, the active and reactive power demands and the voltage profile at the feeder head were estimated. This regression used the random forest method with a coverage of more than 91% and 94%, respectively. The demand for electric vehicles was modeled based on the records observed in a load center located in Guayaquil, and the number of electric vehicles to be connected throughout the day was estimated in a similar way to the feeder demand. The results reveal the potential of the proposed algorithm to improve the load factor since the load peaks are reduced in the intervals of maximum demand and the valleys are filled in the hours of minimum demand, which benefits the distribution company since it maximizes the use of the installed capacity in the distribution networks. Including active and reactive power losses contributes to their slight reduction, besides improving the voltage profile due to the reactive power injection into the feeder. Despite the promising results, this study is not without limitations. The quality and availability of historical data used for training may influence the model’s performance and accuracy. Moreover, the DMS algorithm’s effectiveness may be

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

19

subject to fluctuations in EV adoption rates and changes in consumer behavior. Addressing these limitations opens avenues for future research to enhance the algorithm’s robustness and adaptability. For future research, incorporating realtime data streams, such as weather conditions, EV charging patterns, and grid load variations, could further improve the accuracy and responsiveness of the DMS algorithm. In addition, developing dynamic forecasting models that can adapt to changing EV adoption rates and demand patterns will enable more flexible and reliable predictions to be analyzed. Finally, explore advanced optimization techniques to enhance grid resilience and stability by effectively managing EV charging and load balancing across the distribution network.

References 1. IEA, Transport – Analysis.: https //www.iea.org/reports/transport (2023) 2. IEA, Global EV Outlook 2022 – Analysis.: https://www.iea.org/reports/global-ev-out look-2022 (2022) 3. Asamblea Nacional de Ecuador, Ley Orgánica de Eficiencia Energética. Registro Oficial SAN-2019-2438 (1993) 4. Das, H.S., Rahman, M.M., Li, S., Tan, C.W.: Electric vehicles standards, charging infrastructure, and impact on grid integration: a technological review. Renew. Sust. Energ. Rev. 120, 109618 (2020). https://doi.org/10.1016/J.RSER.2019.109618 5. ARCERNNR -17/2020, Calidad del Servicio de Distribución y Comercialización de Energía Eléctrica. Agencia de Regulación y Control de Energía y Recursos Naturales no Renovables, Quito (2020) 6. Afzalan, M., Jazizadeh, F.: Residential loads flexibility potential for demand response using energy consumption patterns and user segments. Appl. Energy. 254, 113693 (2019). https://doi. org/10.1016/J.APENERGY.2019.113693 7. Mariano-Hernández, D., Hernández-Callejo, L., Zorita-Lamadrid, A., et al.: A review of strategies for building energy management system: model predictive control, demand side management, optimization, and fault detect & diagnosis. J. Build. Eng. 33, 101692 (2021). https://doi.org/10.1016/J.JOBE.2020.101692 8. Groppi, D., Pfeifer, A., Garcia, D.A., et al.: A review on energy storage and demand side management solutions in smart energy islands. Renew. Sust. Energ. Rev. 135, 110183 (2021). https://doi.org/10.1016/J.RSER.2020.110183 9. Mohseni, S., Brent, A.C., Kelly, S., Browne, W.N.: Demand response-integrated investment and operational planning of renewable and sustainable energy systems considering forecast uncertainties: a systematic review. Renew. Sust. Energ. Rev. 158, 112095 (2022). https://doi. org/10.1016/J.RSER.2022.112095 10. Nasir, T., Bukhari, S.S.H., Raza, S., et al.: Recent challenges and methodologies in smart grid demand side management: state-of-the-art literature review. Math. Probl. Eng. 2021 (2021). https://doi.org/10.1155/2021/5821301 11. Zhou, Y., Zheng, S.: Machine-learning based hybrid demand-side controller for high-rise office buildings with high energy flexibilities. Appl. Energy. 262, 114416 (2020). https://doi.org/10. 1016/J.APENERGY.2019.114416 12. Ullah, K., Ali, S., Khan, T.A., et al.: An optimal energy optimization strategy for smart grid integrated with renewable energy sources and demand response programs. Energies. 13, 5718 (2020). https://doi.org/10.3390/EN13215718

20

J. C. Guamán et al.

13. Nutakki, M., Mandava, S.: Review on optimization techniques and role of artificial intelligence in home energy management systems. Eng. Appl. Artif. Intell. 119, 105721 (2023). https://doi. org/10.1016/J.ENGAPPAI.2022.105721 14. Elsisi, M., Su, C.L., Ali, M.N.: Design of reliable IoT systems with deep learning to support resilient demand side management in smart grids against adversarial attacks. IEEE Trans. Ind. Appl. (2023). https://doi.org/10.1109/TIA.2023.3297089 15. Habib, S., Kamran, M., Rashid, U.: Impact analysis of vehicle-to-grid technology and charging strategies of electric vehicles on distribution networks – a review. J. Power Sources. 277, 205–214 (2015). https://doi.org/10.1016/J.JPOWSOUR.2014.12.020 16. Ahmad, M.S.: Self-healing distribution grid: benefits, challenges and optimize way of implementation. In: Lecture Notes in Electrical Engineering, vol. 580, pp. 309–320. https:// doi.org/10.1007/978-981-32-9119-5_26/COVER (2020) 17. Papadaskalopoulos, D., Strbac, G., Mancarella, P., et al.: Decentralized participation of flexible demand in electricity markets – part II: application with electric vehicles and heat pump systems. IEEE Trans. Power Syst. 28, 3667–3674 (2013). https://doi.org/10.1109/TPWRS. 2013.2245687 18. Logavani, K., Ambikapathy, A., Arun Prasad, G., et al.: Smart grid, V2G and renewable integration. Green Energy Technol., 175–186 (2021). https://doi.org/10.1007/978-981-159251-5_10/COVER 19. Teleke, S., Baran, M.E., Bhattacharya, S., Huang, A.Q.: Rule-based control of battery energy storage for dispatching intermittent renewable sources. IEEE Trans. Sustain. Energy. 1, 117–124 (2010). https://doi.org/10.1109/TSTE.2010.2061880 20. Moghimi, M., Leskarac, D., Nadian, N., et al.: Impact of PEV behavior on peak demand reduction in a commercial microgrid. In: Proceedings of the 2016 Australasian Universities Power Engineering Conference, AUPEC. https://doi.org/10.1109/AUPEC.2016.7749380 (2016) 21. Prakash, K., Ali, M., Siddique, M.N.I., et al.: Bi-level planning and scheduling of electric vehicle charging stations for peak shaving and congestion management in low voltage distribution networks. Comput. Electr. Eng. 102, 108235 (2022). https://doi.org/10.1016/J. COMPELECENG.2022.108235 22. Rajakaruna, S., Shahnia, F., Ghosh, A.: Plug in Electric Vehicles in Smart Grids. Springer, Singapore (2015). https://doi.org/10.1007/978-981-287-299-9 23. Oliinyk, M., Džmura, J., Kolcun, M., et al.: Impact of electric vehicles and demand management systems on electrical distribution networks. Electr. Eng. 104, 667–680 (2022). https://doi.org/ 10.1007/S00202-021-01327-0/FIGURES/18 24. Nordin, B., Hu, C., Chen, B., Sheng, V.S.: Interval-valued centroids in K-Means algorithms. In: Proceedings – 2012 11th international conference on machine learning and applications, ICMLA 2012, vol. 1, pp. 478–481. https://doi.org/10.1109/ICMLA.2012.87 (2012) 25. Bonyadi, M.R., Michalewicz, Z.: Particle swarm optimization for single objective continuous space problems: a review. Evol. Comput. 25, 1–54 (2017). https://doi.org/10.1162/EVCO_R_ 00180 26. Ahmed, M., Abouelseoud, Y., Abbasy, N.H., Kamel, S.H.: Hierarchical distributed framework for optimal dynamic load management of electric vehicles with vehicle-to-grid technology. IEEE Access. 9, 164643–164658 (2021). https://doi.org/10.1109/ACCESS.2021.3134868 27. Hai-Ying, H., Jing-Han, H., Xiao-Jun, W., Wen-Qi, T.: Optimal control strategy of vehicle-togrid for modifying the load curve based on discrete particle swarm algorithm. In: DRPT 2011–2011 4th International conference on electric utility deregulation and restructuring and power technologies, pp. 1523–1527. https://doi.org/10.1109/DRPT.2011.5994138 (2011) 28. Tahir, B., Tariq, M.: Voltage profile enhancement due to large scale renewable integrated system with V2G. In: 15th International conference on emerging technologies, ICET. https:// doi.org/10.1109/ICET48972.2019.8994540 (2019)

Demand-Side Management Integrating Electric Vehicles Using Multi-step. . .

21

29. Alvarez-Alvarado, M.S., Rodríguez-Gallegos, C.D., Jayaweera, D.: Optimal planning and operation of static VAR compensators in a distribution system with non-linear loads. IET Gener. Transm. Distrib. 12, 3726–3735 (2018). https://doi.org/10.1049/IET-GTD.2017.1747 30. Rodriguez-Gallegos, C.D., Gandhi, O., Yang, D., et al.: A siting and sizing optimization approach for PV-battery-diesel hybrid systems. IEEE Trans. Ind. Appl. 54, 2637–2645 (2018). https://doi.org/10.1109/TIA.2017.2787680 31. Raúl, G., Guillermo, M.: Learning Scikit-Learn: Machine Learning in Python: Experience the Benefits of Machine Learning Techniques by Applying Them to Real-World Problems Using Python and The Open Source Scikit-Learn Library, p. 103 (2013)

Long-Term Sustainable Energy Transition of Ecuador’s Residential Sector Using a National Survey, Geospatial Analysis with Machine Learning, and Agent-Based Modeling Diego Moya , César Arroba, Christian Castro, Cristian Pérez, Dennis Copara, Alexander Borja, Sara Giarola , and Adam Hawkes

1 Introduction The Paris Agreement aims to limit global warming by keeping the average temperature increase below 1.5 °C compared to the beginning of the century. To fulfill the agreement, each participating country is required to submit Nationally Determined Contributions (NDCs) [1]. In 2019, Ecuador presented its first NDC at the 25th Conference of the Parties (COP25) held in Spain [2]. Within this NDC, Ecuador focuses on five key sectors for mitigation: (1) energy, (2) agriculture, (3) industrial D. Moya (✉) Technology Outlook and Strategy Division, Technology Strategy and Planning Department, Dhahran, Saudi Arabia Institute for Applied Sustainability Research, IIASUR, Quito, Ecuador Department of Chemical Engineering, Imperial College London, South Kensington, UK e-mail: [email protected] C. Arroba · C. Castro · C. Pérez Carrera de Ingeniería Mecánica, Facultad de Ingeniería Civil y Mecánica, Universidad Técnica de Ambato, Ambato, Ecuador D. Copara · A. Borja Institute for Applied Sustainability Research, IIASUR, Quito, Ecuador S. Giarola Department of Chemical Engineering, Imperial College London, South Kensington, UK School of Management, Milan, Italy RFF-CMCC EIEE, Milan, Italy A. Hawkes Department of Chemical Engineering, Imperial College London, South Kensington, UK © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_2

23

24

D. Moya et al.

processes, (4) waste, and (5) land use, land-use change, and forestry [3]. To achieve a climate-neutral planet, every commitment in these key sectors must be fully met, not only by developing countries but also primarily by the most developed countries. In Ecuador, the energy sector is the main contributor to CO2 emissions, accounting for 46.63% of the total emissions. As a result, five mitigation actions have been identified for this sector: (1) Incorporating, reformulating, and updating regulations to promote renewables; (2) Promoting safe and sustainable transportation; (3) Enhancing accessibility to renewable energies; (4) Strengthening energy efficiency and behavioral changes in consumption; and (5) Promoting research for energy efficiency [3, 4]. These actions would be conducted through initiatives such as the National Energy Efficiency Plan, Energy Efficiency Program, Renewable Energies, and Energy Efficiency in the Hydrocarbon sector [5]. Therefore, direct state intervention in energy subsector initiatives will be necessary to fulfill these actions in the first NDC. In 2020, the total energy demand in Ecuador reached 83 million Barrels of Oil Equivalent (BOE) [6]. This demand is distributed among seven sectors: (1) Transportation (45.4%); (2) Industry (17.4%); (3) Residential (15.7%); (4) Commercial and public services (6.4%); (5) Own consumption (4.7%); (6) Agriculture, fishing, mining (1.2%); and (7) Others (9.2%). Moreover, this same demand is broken down into 12 energy sources: (1) Diesel oil (31.4%); (2) Gasoline (26.5%); (3) Electricity (19.2%); (4) Liquefied gas (11.5%); (5) Fuel oil (3.3%); (6) Firewood (2%); (7) Sugarcane products (1.9%); (8) Non-energy (1.4%); (9) Petroleum (1.3%); (10) Natural gas (0.2%); (11) Jet fuel (0.7%); and (12) Other secondary sources (0.7%) [6]. This indicates that petroleum derivatives and electricity remain the primary sources of demand and consumption across different sectors. In 2020, the residential sector consumed 13 million BOE, accounting for 15.7% of the total energy consumption in urban and rural households in the country [6]. This consumption is divided into four sources: (1) Liquefied gas (51.8%); (2) Electricity (38.4%); (3) Firewood (9.7%); and (4) Natural gas (0.1%). Additionally, this consumption is focused on six end uses: (1) home heating (49%); (2) home ventilation (29%), and 22% for (3) water heating, (4) cooking, (5) lighting, and (6) appliances [7]. Energy consumption for heating is related to cold regions in the cold Andes region, while ventilation is associated with the warm climate in the coastal, eastern, and insular regions. In both cases, cold and warm regions, energy consumption for cooking primarily relies on liquefied petroleum gas [8]. In the literature, there is not a detailed assessment of the required sustainable transition of the Ecuadorian residential sector. This research aims to study the long-term energy transition of Ecuador’s residential sector using the results of a national survey, geospatial analysis of large spatiotemporal datasets, and agent-based modeling. This chapter is divided into four parts to fulfill this objective. Section 2 presents a literature review on the state of energy simulation in Ecuador. Section 3 provides the methodology details, while Sect. 4 analyzes and discusses the results in terms of agent characterization, demand, supply, consumption, and emissions. The conclusions of this research are summarized in Sect. 5.

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

25

2 Literature Review This literature review presents the state of the use of energy models for simulating long-term energy transition in Ecuador. Scientific articles related to previous studies on Ecuador’s energy planning have been reviewed for this purpose.

2.1

General Comparison of Studies in Ecuador

Many researchers have developed models that allow the study of different energy systems’ behavior, as shown in Table 1. While some researchers have focused on the Long-Term Optimization Model for Ecuador (TIMES-EC) to simulate alternative scenarios of electric capacity expansion until 2050 [9], others have focused on the Long-range Energy Alternatives Planning System (LEAP) to model energy sector pathways in Ecuador [8, 10, 11]. Other researchers have focused on the Ecuador Land Use and Energy Network Analysis (ELENA) model [12]. These models lack the specific features required for this research, such as using a national survey, geospatial analysis of large spatiotemporal datasets, and agent-based modeling for analyzing the energy transition in Ecuador’s residential sector.

Table 1 Simulation models applied in studies of Ecuador’s energy sector as of the date of this study (May 2023) References Carvajal and Li [9]

Model TIMESEC

ABM NO

GIS YES

IAM NO

Drivers IIASA population and GDP projections with SSP2 narratives Population GDP projections

Chavez et al. [10]

LEAP

NO

NO

NO

Castro et al. [8]

LEAP

NO

NO

NO

Economic, demographic, and technological factors

Espinoza et al. [11]

LEAP

NO

NO

NO

Economic, and demographic factors

Villamar et al. [12]

ELENA

NO

YES

YES

Population GDP projections

ABM agent-based modeling, IAM integrated assessment modeling

Period 2015–2050 every 5 years 2012–2030 every 2 years 2010–2030 every 5 years 2015–2035 every 5 years 2015–2050 every 5 years

26

2.2

D. Moya et al.

Agent-Based Models

Agent-Based Modeling (ABM) is a computational technique used in simulating energy scenarios to capture human behavior (agents) in different energy sectors. On the one hand, ABMs allow the creation of a collection of agents, a set of behavioral rules, and a simulation environment with different exogenous and endogenous constraints [13]. On the other hand, they provide the ability to geographically represent the dynamic behaviors of agents and their decision-making interactions with each other and their environment [14]. ABMs offer greater flexibility in evaluating energy policies by analyzing the population based on their specific characteristics (e.g., income, consumption level, demand, population density, location, etc.) [15]. These models have also been proposed as an alternative to guide innovation development in the context of energy efficiency, considering individuals’ primary motivations for choosing environmentally friendly technologies [16]. Overall, this review has found that the TIMES-EC, LEAP, and ELENA models do not utilize ABMs for studying the long-term energy transition in Ecuador’s energy sector. Instead, these models assumed a single representative agent that takes prices as given and makes rational choices with perfect knowledge of the market and rational expectations to maximize the utility subject to budget constraint.

2.3

Geographic Information Systems

The integration of Geographic Information Systems (GISs) into ABMs allows the identification of the geographic location of different agents with their respective characteristics. Cartography has provided essential support for various aspects of spatial information exchange related to the geographical world [17]. In Ecuador’s energy sector, the TIMES-EC model uses GIS to analyze electricity generation capacity in major regions and watersheds [9]. The ELENA model utilizes GIS to determine relative agricultural production costs for Ecuador and estimate energy demand [12]. However, GIS has more advanced capabilities that can be leveraged to analyze factors influencing demand, consumption, and technology acquisition through the application of Integrated Assessment Models (IAM).

2.4

Integrated Assessment Models

IAMs have a mathematical foundation that represents complex interactions between different spatial and temporal scales, processes, and sector-specific characteristics of the economy. By integrating social sciences, engineering, and climate economics, IAMs estimate the benefits of climate mitigation and compare them with the costs of ecological investments [18]. They are also a vital component in analyzing energy

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

27

system transitions for sustainable socioeconomic development [19]. The ELENA model is the only model that applies IAMs in Ecuador, helping to evaluate scenarios and validate policies within a decarbonization framework [12], while the TIMES-EC and LEAP models are optimization and forecasting models of the energy system [8– 10]. Unlike other models, the application of IAMs allows for the comprehensive integration of key societal and economic characteristics across different sectors. In energy models, both bottom-up and top-down approaches are applied. Studies using a bottom-up approach have more diverse objectives and methodologies, are technologically rich, have predominantly cross-sectional data, and limited time series data [20]. On the other hand, studies using a top-down approach often utilize aggregated historical time series data at the national or regional level [21]. In the case of Ecuador, the LEAP model presented by Chavez adopts a top-down approach [22]. In contrast, the LEAP model presented by Castro and the ELENA model utilizes a bottom-up approach [8]. The TIMES-EC model and the LEAP model presented in the study by Espinoza do not specify either of these approaches. Generally, the bottom-up approach tends to yield more precise results due to its methodology being closer to the requirements of IAMs.

2.5

Long-Term Considerations

In long-term estimations, it is essential to establish narratives and the period for which projections are needed. The TIMES-EC model is the only model that utilizes population and Gross Domestic Product (GDP) projections with Shared Socioeconomic Pathways (SSP2) narratives issued by IIASA [9]. In contrast, all the LEAP models and the ELENA model use narratives based on simple historical projections [8, 10]. The TIMES-EC model focuses its simulation on the period from 2010 to 2050, every 5 years [9]. Similarly, the LEAP model presented by Chavez simulates a period from 2012 to 2030, every 2 years [10]. The LEAP model presented by Castro focuses its simulation on the analysis from 2010 to 2030, every 5 years [8]. The LEAP model presented by Espinoza simulates the analysis from 2015 to 2035, every 5 years [11]. The ELENA model, on the other hand, focuses its simulation on the period from 2015 to 2050, every 5 years [12]. In general, all these periods delineate longterm estimation trajectories, crucial for techno-economic analysis and the direction of energy policies. The energy consumption trajectories that the world may follow are subject to change. Therefore, it is essential to consider different scenarios in energy system simulations. In the case of energy simulations in Ecuador, the TIMES-EC model presents five simulation scenarios: (1) Nationally Determined Contributions (NDC); (2) Dry Climate (DRY); (3) No Large Hydropower (NLH); (4) Costs and Prices (OVR); and (5) Diversified (DIV) [9]. Meanwhile, the LEAP model presented in the study by Chávez focuses on two simulation scenarios: (1) Reference and (2) Policy

28

D. Moya et al.

[10]. The LEAP model presented by Castro focuses on three scenarios: (1) Low Growth (LG); (2) Business-as-Usual (BAU); and (3) High Growth (HG) [8]. Additionally, Espinoza’s LEAP model presents two scenarios: (1) Historical and (2) Alternative (BAU, energy efficiency) [5]. Finally, the ELENA model considers six scenarios: (1) Minimum Cost (MinC); (2) Unconditional Nationally Determined Contributions (NDCu); (3) Conditional Nationally Determined Contributions (NDCc); (4) High Decarbonization (DDP High); (5) Low Decarbonization (DDP Low); and (6) High Decarbonization with Reforestation (DDP High_Refo) [12]. In general, the proposed scenarios focus on decarbonization, political considerations, growth, costs, and nationally determined contributions, which shape the outcomes in each sector. This diversity of narratives makes the methods and results less comparable, although comparisons can be inferred based on assumptions and input data such as population and GDP growth projections. Simulations conducted under different scenarios yield a variety of results based on the problem being addressed. The TIMES-EC model presents results such as capacity, generation, energy demand, and CO2 emissions for different scenarios [9]. Similarly, the LEAP model presented by Chávez shows results of capacity and energy consumption for the different scenarios considered [10]. The LEAP model presented by Castro presents results of energy consumption and generation in its various scenarios [8]. Meanwhile, Espinoza’s LEAP model generates results of energy demand and emissions in different scenarios [11]. Finally, the ELENA model provides results on primary energy supply, electricity generation, installed power capacity, and greenhouse gas emissions across the different scenarios considered [12]. The results presented by the TIMES-EC, LEAP, and ELENA models encompass various sectors: industrial, residential, commercial, transportation, agriculture, public, services, and others.

2.6

MUSE-RASA as an Alternative Analysis

The ModUlar energy system Simulation Environment (MUSE), ResidentiAl Spatially-resolved and temporal-explicit Agents (RASA) model, MUSE-RASA is a bottom-up Integrated Assessment Model (IAM) developed to simulate long-term alternative scenarios in all energy sectors [23]. These scenarios help identify plausible pathways for transitioning energy systems toward a low-carbon economy and the technologies involved [24]. It employs a modular, agent-based approach to simulate investment decision-making in each sector [25]. This framework focuses on investor motivations for adopting new energy technologies, allowing for a potentially more realistic representation of the energy market transition compared to transformation pathways in optimization models. The modular structure of MUSE-RASA is designed to enable transparent and flexible analysis of all energy market sectors as a whole or individually.

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

29

3 Methodology Figure 1 illustrates the three calculation processes of the MUSE-RASA model: (1) Input data, (2) Investment decisions within the residential sector, and (3) Outputs. The details of each of these processes are presented below. The methodology of this research has been grouped into three stages.

Fig. 1 MUSE-RASA framework applied in this study

30

3.1

D. Moya et al.

MUSE-RASA Input Data

In the first process, three types of inputs are presented: (1) GIS-based agents, (2) ABM characterized by a survey, and (3) Population and GDP projections based on IIASA SSP2 2010–2050 projections, carbon prices, and other prices (commodities). The initial input data consist of agents based on GIS and characterized in a national survey, which are stratified into five classes (Q1, Q2, Q3, Q4, Q5) and defined by Eq. 1. Details of this survey and agent characterization can be found in reference [26]. The ABM is used to stratify and characterize agents at the national level through a survey that considers aspects such as investment objectives, search rule, budget, decision strategy, and technology type when upgrading or acquiring technology for different end uses in the residential sector of Ecuador [26]. Table 2 presents the definition and metrics for characterizing each agent using Eq. 1. The parameters, definitions, and metrics used have been obtained from references [26, 27]. Finally, the primary exogenous input includes population and GDP projections issued by IIASA, based on the SSP2 socioeconomic pathways, which will guide the trajectory of the MUSE-RASA simulation [28] (Table 3). A = fObj, SR, DS, TT, B, TS, PPg

3.2

ð1Þ

Investment Decision in the Residential Sector of MUSE-RASA

In the second process, the decision-making scheme is observed when an agent decides to invest in technology within the residential sector module of MUSE-RASA. In the first stage, an estimation of future demand is made, considering historical trends and projections of macroeconomic variables (Population, GDP, urbanization). In the second stage, the estimation of future installed capacity is considered, considering the current installed capacity, decommissioning profile, and innovative technologies required to meet future demand. In the third stage, the estimation of total investment is made, with the primary focus on future demand, installed capacity, and utilization factor. All these estimates are based on the projections of macroeconomic variables projected by the SSP2 pathways. Finally, in the last stage, investment in technologies is estimated according to the parameters of each agent, focusing on the characteristics established by the survey for their selection. In this stage, potential assets are selected by excluding unviable options and classifying feasible options according to the investment objectives of each agent.

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

31

Table 2 Simulation models applied in studies of Ecuador’s energy sector as of the date of this study (May 2023) Refernces Carvajal and Li [9]

Chavez et al. [10]

Scenarios NDC DRY NLH OVR DIV Reference Policy

Approach Bottomup

Topdown

Castro et al. [8]

LG BAU HG

Bottomup

Espinoza et al. [11]

Historical Alternative

Bottomup

Villamar et al. [12]

MinC NDCu NDCc DDP high DDP low DDP High_Refo

Bottomup

Sectors Industries Strategic industries Commercial Residential Transportation Residential Transportation Industrial, public, and services Residential Transportation Industrial Services Others Residential Industry Transportation Commercial Public Industry Residential Commercial Transportation Agriculture

Results Capacity Generation Demand Emissions Capacity Consumption

Consumption Generation

Demand Emissions

Primary energy supply Electric generation Installed power capacity. Emissions

Scenarios refer to a hypothetical or projected representation of potential future energy production, consumption, and distribution over a certain time frame. Energy scenarios often consider various variables such as technological advancements, policy changes, economic conditions, environmental concerns, and societal shifts. These scenarios are used to understand the potential implications of different pathways in the energy sector, aiding in decision-making, policy formulation, and longterm planning for energy sustainability and security. The bottom-up approach starts from the individual components or elements and builds upward. The top-down approach starts from an overall view of a system and then delves into its components

3.3

Outputs of the Residential Sector of MUSE-RASA

In the third process, the MUSE-RASA model generates the following results: (1) Installed capacity, (2) Service demand and energy supply, (3) Fuel and electricity consumption, (4) Emissions, (5) Levelized cost of electricity, (6) Net present value, (7) Capital costs, (8) Fixed costs, (9) Variable costs, (10) Fuel costs, and (11) Emission costs. Thus, long-term projections (2050–2100) are presented for end-use energy services: space heating, water heating, space ventilation, cooking, lighting, and appliances. These results were obtained for a sustainable scenario, considering carbon prices for multiple investment objectives according to the survey.

32

D. Moya et al.

Table 3 Characterization of agents based on findings from the national survey Agent parameter Objectives

In Eq. (1) Obj.

Search rule

RB

Decision strategies

ED

Technology type

TT

Budget

B

Technology stock

TS

Percentage of population

PP

Definition A combination of economic, environmental, and technological aspects along with personal motivations.

A compilation of information on usable technologies and decisionmakers processing skills. The RB guides the search space (EB) of each agent, which encompasses all permissible technologies within the residential sector. There are two EDs for single or multiple objectives. The singleobjective ED focuses on the main objective, while the multipleobjective ED uses a merit-order approach where technologies are ranked based on the agent’s objective. Within MUSE, three possible multi-objective ED approaches are implemented. Two types of agents: New or upgraded. There is a distinction between upgrading or acquiring a new technology. These two types must be related to transfer their actions for future asset renewal. Maximum expenses and income that each agent can use in asset investment. Technological capabilities available in the base year, obtained through calibration to the energy balance and relevant data. Percentage of the population represented by spatial GIS characterization per agent.

Metric Emissions. Fuel consumption cost Efficiency ACE (annual cost equivalent) with a high discount rate. ACE with intermediate discount rate. ACE with a low discount rate. Similar Fuel type Existing All

Lexicographica Weighted sumb Epsilon-constrainedc

New. Upgraded.

USD.

Existing current technologies.

Percentage of the population found within each agent, classified through spatial analysis.

The agent selects all systems similar to the best based on the first criterion and then makes a final decision among these systems based on the second and third criteria b The agent weighs all three criteria equally when making a decision c The agent selects the system with the highest rating based on the first criterion only a

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

33

4 Results This section presents six sets of results: (1) geospatial characterization of agents, (2) national demand, (3) demand for space heating, (4) supply, (5) consumption, and (6) emissions by agent.

4.1

Geospatial Characterization of Agents

Figure 2 provides a summary of the results from the national survey combined with geospatial data to characterize five agents across the country. These results have been published in reference [26]. In summary, the survey allows assigning seven characteristics to each agent according to Eq. 1. Agent 1, ECU1, belongs to the quintile of the population with higher incomes, while Agent 5, ECU5, belongs to the one-fifth of the population with lower incomes. In Fig. 2, it can be observed that all agents, except ECU4, prioritize emission reduction as their first investment objective when selecting a technology for space heating. Objectives 2 and 3 highlight the heterogeneity among agents. For higherincome agents, ECU1 and ECU2, objective 2 is the initial capital cost (CAPEX), and objective 3 is the annual equivalent cost (AEC) and operation costs (OPEX), respectively. An interesting observation is the contrast between ECU1 and ECU5 in terms of the search rule. While ECU1 is willing to invest in all available technologies in the market that meet their investment objectives, ECU5 will only select technologies similar to those currently in use. This is particularly relevant because ECU5 represents approximately 51% of the Ecuadorian population.

Fig. 2 Geospatial characterization of agents in the residential sector of Ecuador

34

4.2

D. Moya et al.

Total Demand

Figure 3 presents the total energy demand distributed across six end uses for different types of agents in the residential sector of Ecuador. In Fig. 3a, the total energy demand in Ecuador is shown for the following end uses: (1) space heating (hspace), (2) water heating (hwater), (3) space ventilation (cspace), (4) cooking (cook), (5) lighting (light), and (6) appliances (appl). The energy demand distribution for the year 2020 shows that 37.36 PJ is allocated to heating space, 15.63 PJ to heating water, 0.15 PJ to cooling space, 10.02 PJ to cooking, 4.29 PJ to lighting, and 18.60 PJ to appliances, resulting in a total energy demand of 86.05 PJ. Moving forward to 2050, there is an uptick in the overall energy demand to 103.37 PJ. This distribution is as follows: 44.88 PJ for heating space, 18.77 PJ for heating water, 0.17 PJ for cooling space, 12.04 PJ for cooking, 5.16 PJ for lighting, and 22.35 PJ for appliances. Looking further ahead to 2100, there is a decline in the total energy demand, reaching 98.48 PJ. The breakdown of this demand is: 42.75 PJ for heating space, 17.88 PJ for heating water, 0.16 PJ for cooling space, 11.47 PJ for cooking, 4.91 PJ for lighting, and 21.29 PJ for appliances. Figure 3b shows the disaggregated energy demand by agents, represented by five income classes Q1, Q2, Q3, Q4, and Q5. In 2020, these agents have a demand of 1.14 PJ, 4.07 PJ, 8.93 PJ, 28.42 PJ, and 43.49 PJ, respectively. For 2050, the evolution of demand for these agents reaches 1.37 PJ, 4.89 PJ, 10.72 PJ, 34.15 PJ, and 52.24 PJ, respectively. In 2100, the demand would be 1.31 PJ, 4.65 PJ, 10.21 PJ, 32.51 PJ, and 49.74 PJ, respectively. Overall, it can be observed that the demand for different end uses of energy will experience growth until 2060 and then tend to decrease by the year 2100. The demand curve aligns with the population and GDP growth curves, based on the IIASA projections in the SSP2 narrative explained earlier.

Fig. 3 Energy demand by end users in the residential sector of Ecuador, using the MUSE-RASA model

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

4.3

35

Demand for Space Heating

Figure 4a presents the total energy demand for space heating in residential homes. It can be observed that in the years 2020, 2050, and 2100, the demand will be 37.34 PJ, 44.85 PJ, and 42.72 PJ, respectively. Figure 4b shows the disaggregated energy demand by agents classified into income classes Q1, Q2, Q3, Q4, Q5. The demand for each agent in 2020 was 0.49 PJ, 1.77 PJ, 3.87 PJ, 12.34 PJ, and 18.87 PJ, respectively. In 2050, the demand per agent would be 0.59 PJ, 2.12 PJ, 4.65 PJ, 14.82 PJ, and 22.67 PJ, respectively. By the year 2100, the demand would be 0.56 PJ, 2.02 PJ, 4.43 PJ, 14.11 PJ, and 21.60 PJ, respectively. Overall, it can be observed that agents ECU4 and ECU5 dominate the energy demand for space heating.

4.4

Supply by Technology

Figure 5 presents the energy supply technologies for space heating (hspace) both in total and disaggregated by agents in the residential sector of Ecuador. The MUSERASA model selects five technologies for this end use: (1) air source heat pump with

Fig. 4 Energy demand for space heating in the residential sector of Ecuador, using MUSE-RASA

Fig. 5 Energy supply by technologies for space heating in the residential sector of Ecuador

36

D. Moya et al.

natural gas backup (ASHeatPumpNG), (2) electric boiler (BoilerElectric), (3) biomass boiler (BoilerBiomass), (4) LPG boiler (BoilerLPG), and (5) kerosene boiler (BoilerKerosen). As shown in Fig. 5a, in 2020, a total of 37 PJ of energy is supplied, distributed as 16 PJ from LPG boilers and 21 PJ from electric boilers. In 2050, the supply will increase to 45 PJ, with 22 PJ supplied by electric boilers and 20 PJ by heat pumps. Biomass boilers represent a marginal supply of 3 PJ. In 2100, the supply decreases to 43 PJ, distributed as 21 PJ from electric boilers, 20 PJ from heat pumps, and 2 PJ from biomass boilers. Biomass boilers in agents ECU1 and ECU2 (higher-income agents), heat pumps in agents ECU3 and ECU4 (medium-income agents), and electric boilers in agent ECU5 (lower-income agents) will play a significant role in decarbonizing the residential sector of Ecuador in the medium and long term.

4.5

Energy Consumption

Figure 6 illustrates the fuel consumption for space heating in residential homes in Ecuador for five types of fuel and electricity: (1) diesel, (2) electricity, (3) biomass, (4) natural gas, and (5) liquefied petroleum gas (LPG). In Fig. 6a, it can be observed that in the year 2020, there was an approximate consumption of 50 PJ. This energy came primarily from electricity (60%) and LPG (40%). The projections for the sustainable scenario in 2050 show an approximate consumption of just over 80 PJ, mainly from electricity, natural gas, and biomass. In Fig. 6b, it can be observed that agents ECU1 and ECU2 are inclined to use technologies that utilize biomass as fuel, while agent ECU5 has a higher consumption of electricity. Agents ECU3 and ECU4 consume natural gas and electricity. Overall, a sustainable transition from LPG to electricity and hydrogen can be observed in the future. Regarding electricity consumption for space heating, in the year 2020, approximately 10 TWh (35 PJ) from hydroelectricity and 4 TWh (15 PJ) from biomass were consumed. In the year 2050, it is projected that approximately 11 TWh (38 PJ) of hydroelectricity and 10 TWh (35 PJ) of biomass will be

Fig. 6 Energy consumption for space heating in the residential sector of Ecuador

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

37

consumed. It can be observed that the highest consumption is by agents ECU4 and ECU5, with approximately 2.8 TWh (10 PJ) and 10 TWh (35 PJ) in the year 2050.

4.6

Emissions

Figure 7 shows the CO2 emissions from energy use for space heating in residential homes in Ecuador. Figure 7a illustrates that in 2010, there was 3.5 kt of CO2 emissions [ktCO2], which decreased to below 1 ktCO2 in 2015. In 2020, emissions exceeded 1 ktCO2, but by 2025, it is projected to be slightly above 0.5 ktCO2. It can be observed that from 2030 onward, the emissions are expected to reach zero in this sector. Within the agent-based model, Fig. 7b shows that agents ECU4 and ECU5 generate higher emissions, but they are projected to reduce to approximately 0.2 and 0.3 ktCO2 by 2025. It is important to note that the natural gas used by agents 3 and 4 in heat pumps serves as a backup and is not considered in the emissions calculation.

5 Conclusions This research has examined the long-term energy transition of Ecuador’s residential sector using the results of a national survey, geospatial analysis of large spatiotemporal datasets, and agent-based modeling in the MUSE-RASA framework. It has been observed that the national study of the residential sector transition can be disaggregated into five agents, primarily characterized by income level. Additionally, for each agent, investment objectives, search rules, decision strategies, spatially distributed population percentages, and a budget for investing in household energy technologies have been obtained. This geospatial characterization of agents enables more realistic input data for MUSE-RASA and facilitates a simulation closer to reality. The results are diverse and heterogeneous, but they provide insights into the

Fig. 7 CO2 emissions from the use of energy for space heating in the residential sector of Ecuador

38

D. Moya et al.

transition’s evolution for each agent and at a national level. These findings can be utilized by decision-makers in the public and industrial sectors to plan the country’s residential energy sector in the medium and long term. Localized energy policies can be implemented, and investment prioritization in the development and research of technologies suitable for a sustainable energy transition scenario can be conducted. Although the spatial characterization has been previously validated, and the survey is based on a proven methodology, the main limitation of this study is the validation and comparison of agent-based results. Currently, no literature is available for comparison with the presented results. Future research will address the costs of the transition. Acknowledgments Diego Moya, Christian Castro, César Arroba, and Cristian Pérez have been funded by UTA, DIDE research project, Award No. UTA-CONIN-2020-0296-R. Diego Moya has been also funded by the Ecuadorian Secretariat for Higher Education, Science, Technology and Innovation (SENESCYT), Award No. CZ03-35-2017, and supported by The Science and Solutions for a Changing Planet Doctoral Training Partnership, Grantham Institute, at Imperial College London. The Institute for Applied Sustainability Research (IIASUR) supports international research on global sustainability applied to the Global South. We acknowledge the important comments and suggestions made by the anonymous reviewers to improve the quality, clarity, and strictness of this chapter. This research was developed during the PhD studies of Dr. Moya at Imperial College London and in collaboration with the coauthors of this study. The edition and submission of this research paper have been developed during Dr. Moya’s position at Saudi Aramco’s TSPD-TOS team. Dr. Moya acknowledges the support and endorsement of Dr. Ali Al-Dawood to submit this chapter. The views expressed in this paper do not necessarily reflect Saudi Aramco’s official policies and do not reveal confidential data.

References 1. UNFCCC, Acuerdo de París. https://unfccc.int/files/meetings/paris_nov_2015/application/pdf/ paris_agreement_spanish_.pdf. Accessed 28 May 2023 2. Rabi, V.: COP25 El Camino a Seguir para Aumentar las Amiciones Climáticas de los Países del Hemisferio Sur Proyecto N°109140–001 (2020) 3. Ministerio del Ambiente Ecuador, Primera Contribución Determinada a Nivel Nacional para el Acuerdo de París bajo la Convención Marco de Naciones Unidas sobre Cambio Climático. https://unfccc.int/sites/default/files/NDC/2022-06/Primera%20NDC%20Ecuad or.pdf. Accessed 28 May 2023 4. Samaniego, J., Alatorre, J.E., Reyes, O., Ferrer, J., Muñoz, L., Arpala, L.: Panorama de las contribuciones determinadas a nivel nacional en América Latina y el Caribe. https://repositorio. cepal.org/handle/11362/44974 (2019). Accessed 28 May 2023 5. Muñoz Chumo, E.A., Balderramo Vélez, N.R., Pico Mera, G.E.: Eficiencia Energética en Función del Desarrollo del Plan Maestro de Electrificación (PME) en Ecuador. Revista de Investigaciones en Energía, Medio Ambiente y Tecnología: RIEMAT. ISSN: 2588-0721. 3(2), 1 (2018). https://doi.org/10.33936/riemat.v3i2.1624 6. Ministerio de Energía y Minas Ecuador, Ministerio de Energía y Minas: Balance Energético Nacional (2022) 7. Ríos, A., Guamán, J., Vargas, C.: Análisis de la Implementación de una Estrategia de Reducción del Consumo Energético en el Sector Residencial del Ecuador: Evaluación del

Long-Term Sustainable Energy Transition of Ecuador’s Residential. . .

39

Impacto en la Matriz Energética. Revista Técnica ‘Energía’. 15(1) (2018). https://doi.org/10. 37116/revistaenergia.v15.n1.2018.328 8. Castro Verdezoto, P.L., Vidoza, J.A., Gallo, W.L.R.: Analysis and projection of energy consumption in Ecuador: energy efficiency policies in the transportation sector. Energy Policy. 134, 110948 (2019). https://doi.org/10.1016/j.enpol.2019.110948 9. Carvajal, P.E., Li, F.G.N.: Challenges for hydropower-based nationally determined contributions: a case study for Ecuador. Clim. Pol. 19(8), 974–987 (2019). https://doi.org/10.1080/ 14693062.2019.1617667 10. Chavez-Rodriguez, M.F., et al.: Fuel saving strategies in the Andes: long-term impacts for Peru, Colombia and Ecuador. Energ. Strat. Rev. 20, 35–48 (2018). https://doi.org/10.1016/j.esr.2017. 12.011 11. Espinoza, V.S., Sebastian Espinoza, V., Guayanlema, V., Martínez-Gómez, J.: Energy efficiency plan benefits in Ecuador: long-range energy alternative planning model. Int. J. Energy Econ. Policy. 8(4), 42–54 (2018) 12. Villamar, D., et al.: Long-term deep decarbonisation pathways for Ecuador: insights from an integrated assessment model. Energ. Strat. Rev. 35, 100637 (2021). https://doi.org/10.1016/j. esr.2021.100637 13. Moglia, M., Podkalicka, A., McGregor, J.: An agent-based model of residential energy efficiency adoption. J. Artif. Soc. Soc. Simul. 21(3) (2018). https://doi.org/10.18564/jasss.3729 14. Ding, Z., Hu, T., Li, M., Xu, X., Zou, P.X.W.: Agent-based model for simulating building energy management in student residences. Energ. Buildings. 198, 11–27 (2019). https://doi.org/ 10.1016/j.enbuild.2019.05.053 15. de Wildt, T.E., Chappin, E.J.L., van de Kaa, G., Herder, P.M., van de Poel, I.R.: Conflicted by decarbonisation: five types of conflict at the nexus of capabilities and decentralised energy systems identified with an agent-based model. Energy Res. Soc. Sci. 64, 101451 (2020). https:// doi.org/10.1016/j.erss.2020.101451 16. Tian, S., Chang, S.: An agent-based model of household energy consumption. J. Clean. Prod. 242, 118378 (2020). https://doi.org/10.1016/j.jclepro.2019.118378 17. Lü, G., Batty, M., Strobl, J., Lin, H., Zhu, A.-X., Chen, M.: Reflections and speculations on the progress in geographic information systems (GIS): a geographic perspective. Int. J. Geogr. Inf. Sci. 33(2), 346–367 (2019). https://doi.org/10.1080/13658816.2018.1533136 18. Rising, J.: Decision-making and integrated assessment models of the water-energy-food nexus. Water Secur. 9, 100056 (2020). https://doi.org/10.1016/j.wasec.2019.100056 19. Huppmann, D., et al.: The MESSAGE integrated assessment model and the ix modeling platform (ixmp): an open framework for integrated and cross-cutting analysis of energy, climate, the environment, and sustainable development. Environ. Model Softw. 112, 143–156 (2019). https://doi.org/10.1016/j.envsoft.2018.11.012 20. Wiesmann, D., Lima Azevedo, I., Ferrão, P., Fernández, J.E.: Residential electricity consumption in Portugal: findings from top-down and bottom-up models. Energy Policy. 39(5), 2772–2779 (2011). https://doi.org/10.1016/j.enpol.2011.02.047 21. Oliveira Panão, M.J.N., Brito, M.C.: Modelling aggregate hourly electricity consumption based on bottom-up building stock. Energ. Buildings. 170, 170–182 (2018). https://doi.org/10.1016/j. enbuild.2018.04.010 22. Chevez, P.: Construcción de Escenarios Urbano-Energéticos a partir de la Implementación de Estrategias de Eficiencia Energética y Energías Renovables en el Sector Residencial, Universidad Nacional de Salta, Salta. https://ri.conicet.gov.ar/handle/11336/84424 (2017). Accessed 28 May 2023 23. Moya, D., Copara, D., Olivo, A., Castro, C., Giarola, S., Hawkes, A.: MUSE-RASA captures human dimension in climate-energy-economic models via global geoAI-ML agent datasets. Sci. Data. 10, 693 (2023). https://doi.org/10.1038/s41597-023-02529-w 24. Moya, D., Budinis, S., Giarola, S., Hawkes, A.: Agent-based scenarios comparison for assessing fuel-switching investment in long-term energy transitions of the India’s industry sector. Appl. Energy. 274, 115295 (2020). https://doi.org/10.1016/j.apenergy.2020.115295

40

D. Moya et al.

25. Sachs, J., Moya, D., Giarola, S., Hawkes, A.: Clustered spatially and temporally resolved global heat and cooling energy demand in the residential sector. Appl. Energy. 250, 48–62 (2019). https://doi.org/10.1016/j.apenergy.2019.05.011 26. Moya, D., Copara, D., Amores, J., Muñoz Espinoza, M., Pérez-Navarro, Á.: Caracterización de agentes de consumo energético en el sector residencial del Ecuador basada en una encuesta nacional y en los sistemas de información geográfica para modelamiento de sistemas energéticos. Enfoque UTE. 13(2), 68–97 (2022). https://doi.org/10.29019/enfoqueute.801 27. Moya, D., et al.: Geospatial and temporal estimation of climatic, end-use demands, and socioeconomic drivers of energy consumption in the residential sector in Ecuador. Energy Convers. Manag. 261, 115629 (2022). https://doi.org/10.1016/j.enconman.2022.115629 28. Bauer, N., et al.: Shared socio-economic pathways of the energy sector – quantifying the narratives. Glob. Environ. Chang. 42, 316–330 (2017). https://doi.org/10.1016/j.gloenvcha. 2016.07.006

Part II

Computational Modeling

Analysis of Incidence of Angle of Attack on Energy Efficiency of a Two-Dimensional Airfoil NACA 1412 Luis Gonzaga-Bermeo and Bristol E. Carriel

, Carlos A. Cuenca

, Jorge E. Game,

1 Introduction The process of carbon reduction of industrial sector in countries that signed the Paris agreement in 2008 has promoted the study and development of renewable energy sources as a replacement for fossil fuels. The EU has committed to make important efforts to achieve GHG emission reduction by at least 40% by 2030. Therefore, notables measurements have been adopted, such as integration of renewable energy, energy efficiency measures, sustainable manufacturing. In this research that is a second release from previous research [1], the asymmetric airfoil (NACA1412) is simulated when the effective angle of attack is added. According to literature, it will perform higher power extraction and efficiency under specific input parameters, which are going to be investigated. The environment where the airfoil is moving is water with density of 999 Kg/m3 and constant oncoming velocity Ux = 0.0045 m/s, laminar regime Re = 1100. This study aims to determine the optimal parameters regarding pitching angle, heaving amplitude, frequency of oscillation when the airfoils are involved in steady laminar flow (Re 1100). In addition, as per the recommendation shown in [2], to ensure that airfoils are operating in power-extraction regime, the feather parameter χ > 1 must be applied as necessary condition. This parameter qualifies the effect of the sinusoidal movement of airfoils on the flow regime. The taken range is 1 < χ < 2.5, which represents a pitching amplitude between 40° < Ө0 < 95°

L. Gonzaga-Bermeo (✉) Technische Universität Hamburg, Hamburg, Germany e-mail: [email protected] C. A. Cuenca · J. E. Game · B. E. Carriel Escuela Superior Politécnica del Litoral, ESPOL, Facultad de Ingeniería Mecánica y Ciencias de la Producción, Guayaquil, Ecuador © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_3

43

44

L. Gonzaga-Bermeo et al.

and maximum angles of attack between 1° < αt/4 ≤ 60°. Moreover, the effects on varying parameters such as nondimensional frequency 0.11 ≤ f * ≤ 0.20 and heaving amplitude 0.5
chord ≤ H0 ≤ 1.5
chord are investigated. A comparison of mapping of efficiency in the parametric space pitching amplitude vs non-dimensional frequency ( f *, Ө0) α is presented for heave amplitude H0 = 1
chord and pitching axis at 33% of chord length. In this research, the Eulerian frame of reference is the used method to know the response of the NACA profile when parameters vary. Furthermore, the use of a Dynamic Mesh allowed the deformation of the domain while the programming in C language of a User Defined Function (UDF) achieved the prescribed oscillatory movement for the airfoil. For previous 2D numerical simulations that have been carried out with an asymmetric airfoil, [2], the laminar regime of 1100, which is the best case, reached a maximum efficiency of 17.14%. However, when the airfoil rotates including effective angle of attack, it was found to have a growth efficiency. Finally, the validation of results is made when contrasting numerical simulation from Kinsey and Dumas, [3], with our study at same chord length of airfoil (240 mm), Reynolds number of 1100, non-dimensional frequency, f * = 0.14, and pitching amplitude between 60° < Ө0 < 80°.

2 Methodology The methodology applied for this study was to simulate the asymmetric airfoil at different parameters like non-dimensional frequency, amplitude of pitching angle from previous research, the addition of effective angles of attack in the equation of movement, and its derivation. After simulating with parameters described in Table 1, the force obtained was the vertical components for local lifting force and moment, where both results let us to compute the power efficiency and power extracted. In addition, the authors investigated the influence of the change of amplitude of heave over the efficiency and power extracted. A flowchart shows a general outline of procedure, see Fig. 6. To compare the efficiencies with the previous research, the same parameters for the pitching amplitude but also adding the effective angle of attack are presented in Sect. 3.

Table 1 Range of parameters to evaluate for asymmetric airfoil NACA 1412 when oncoming fluid flow is laminar (Re = 1100) at U1 = 0.0045 m/s f = 0.11 0.14 0.16 0.20

f
c U1

ω = 2πf 0.0130 0.0165 0.0188 0.0236

U1,m/s 0.0045

χ 1 < χ ≤ 2.5

θ0,deg θ0 > 35° θ0 > 41° θ0 > 45° θ0 > 51°

αmax,deg 5 ° < α(t) < 60° 1 ° < α(t) < 37° 1 ° < α(t) < 33° 0 ° < α(t) < 27°

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

2.1

45

Description of Movement and Equations

The motion of the asymmetric airfoil NACA 1412 is described as Heaving and Pitching movements executed simultaneously, with a fixed rotational axis Xp located at 33% chord from the leading edge, as shown in Fig. 1. The Eqs. 1 and 2 describe mathematically the Harmonic movement of the foil. The Eq. 1 stands for heaving with Amplitude H0, which is defined as one time the chord (1
chord = 240 mm). In addition, the angular frequency omega is defined as ω = 2π
f, where f is the frequency of oscillation and depends on the operating regime (propulsion or extraction of energy). Then, the phase angle Phi, Φ, is set at 90°. For the pitching movement Eq. 2 describes it, with amplitude Ө0 that varies between 50° < Ө0 < 80°, due to its maximum efficiency is reached in this range [1]. The velocities in heaving and pitching are shown in Eqs. 3 and 4.

2.2

hðt Þ = H 0 sinðωt þ ϕÞ

ð1Þ

θðt Þ = θ0 sinðωt Þ

ð2Þ

h_ ðt Þ = V y ðt Þ = H 0 ω cosðωt þ ϕÞ

ð3Þ

θ_ ðt Þ = θ0 ωcosðωt Þ

ð 4Þ

Effective Angle of Attack

The effective angle of attack is formed when locally at the airfoil the velocity vector deviates from incoming flow direction when is affected by the so-called induced velocity, which makes a reduction of the geometric angle of attack and forms the induced angle. The difference between them is named effective angle of attack.

Fig. 1 Sketch of movement of a two-dimensional airfoil NACA 1412 when heave and pitch are out of phase, and effective angle of attack is involved. The oncoming water comes from left to right side

46

L. Gonzaga-Bermeo et al.

The airfoil experiences an effective angle of attack, α, and upstream velocity, Veff, due to the imposed motion (heave and pitching) in a domain with upstream flow conditions. This angle is dynamic, and it is expressed in Eq. 5. The peak forces generated and the potential for dynamic-stall occurrence are heavily influenced by the highest values observed during a cycle. On the other hand, the reference system for pitching is positive when the airfoil is rotating counterclockwise, which causes a modification in the original formula from [3] (Fig. 2). αðt Þ = θðtÞ - arctan

V eff ðt Þ =

2.3

h_ ðt Þ U1

U 1 2 þ h_t

2

ð5Þ

ð6Þ

Operating Regime for Power Extraction and Feathering Parameter

The power extraction regime is reached under two necessary conditions, but not sufficient. This is when the “Feathering parameter,” χ, is positive and greater than 1, (χ > 1), and the second condition is when the effective angle of attack on every → quarter of period is negative αt/4 < 0. In addition, the local forces such as Lift, L , → ! and Drag, D , produce a resultant aerodynamic force R, which is decomposed into ! ! its components Rx and Ry , one can realize that the flow is producing a positive ! work over the airfoil when the angle between Ry and airfoil’s displacement forms

Fig. 2 Local effective angle of attack, and local lift force during 3/4 of a cycle

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

47

zero degrees. Furthermore, the maximum value for the effective angle of attack will be reached at odd quarter of period. That is, Tt = 14, 34 , 2n 4- 1 . . . n = 1, 2, . . . χ=

θ0 arctan

ð7Þ

H0 ω U1 →

The avoided case is when the aerodynamic force R is tangent to motion path and opposed to the airfoil’s displacement, which implies χ = 1 and effective angle of attack at odd quarter of a cycle, αt/4 equal to 0. The parameters to be evaluated are presented in Table 1. As explained before, one of the conditions for power extraction is when feathering parameter is greater than 1, for instance, when f * = 0.11, the pitching angle should be greater than Ө0 = 35°. Thus, the minimum angle Ө0 considered for testing at f * = 0.11 was 40°. Moreover, since the effective angle of attack varies in time, in Table 1 is presented the range for maximum effective angles, which are constant at quarter of a cycle.

2.4

Power and Efficiency Calculation

The power extracted per unit of depth is produced by vertical force, Ry(t), and velocity when heaving Vy(t) also by the Torque M(t), about the axis Xp and angular velocity α_ ðt Þ. P = Py ðt Þ þ Pα ðt Þ = Ry ðt Þ V y ðt Þ þ M ðt Þ α_ ðt Þ

ð8Þ

Then for calculation of efficiency, the total power available Pα from oncoming water flow through the extraction plane is calculated as follows: Pa =

1 3 ρU d 2 1

ð9Þ

Where, U1 is the velocity of flow, ρ is the water density (999 mkg3 ), and “d” is the overall extent of airfoil when moving in y direction. The efficiency is defined as the ratio of the cycle-average power extracted (P) to the total available power P1. η=

Py ðt Þ þ Pα ðt Þ 3 1 2 ρU 1 d

ð10Þ

48

2.5

L. Gonzaga-Bermeo et al.

Geometrical Model

The geometry of domain as well as the airfoil was modeled in Rhinoceros [4]. The airfoil has a chord length of 240 mm. The distance from airfoil’s leading edge to inlet is given in terms of chord length (10
chord), the distance from trailing edge to outlet is (16
chord). Finally, the distance from the top boundary domain to the starting position of the aerodynamic profile is (7
chord), while the distance from the bottom boundary domain to the lowest airfoil's profile surface is (9
chord) at t/ T = 0. The airfoil is moving vertically in the range 0 ≤ H0/chord ≤1. The domain dimension is specified to ensure that the motion of the foil and the control volume do not interfere with each other. It also ensures that the waterflow behind the airfoil is effectively captured. This was confirmed by examining the velocity vectors near the top and bottom boundaries since these vectors remained parallel.

2.6

Numerical Model (Mesh, Boundary Conditions, Dynamic Mesh, Solver, and Residuals

Meshing and Mesh Independence Analysis After modeling in Rhinoceros, the domains and airfoils were exported into Ansys Fluent [5]. For NACA 1412, three types of meshes were compared to guarantee independence of results. These meshes are coarse (21 k nodes), medium (61 k nodes), and fine (201 k nodes). To capture the boundary layer, and wake behavior, a refinement was done around the NACA profile. The “inflation” method was used to create at least 30 layers with quadrangular elements as illustrated in Fig. 5. Extreme care was taken to ensure that the aspect ratio of the generated elements does not exceed 3 to avoid errors caused by discontinuities. To achieve this, the height of the first layer was calculated using the Y+ theory, for laminar flows and a growth rate of 1.05. From the previous research [1], it was proved that the coarse mesh allows to obtain results under 10% of difference than the fine mesh. Therefore, the fluid domain will have 21 K elements. To reduce computational effort and time, it was proved that after the first period, the lifting and moment did not change, as shown in Fig. 3. As a consequence, only two cycles were simulated. On the other hand, the time step was determined by using the Courant– 1 Δt  1, where Δh is the size of the cell, Friedrichs–Lewy condition, CFL = UΔh measured horizontally around the airfoil profile, Δt is the time step, and velocity U1 = Ux = 0.0045 m/s, leading to time step, Δt, equal to 0.11 s. Boundary Conditions In this study, the fluid behaves as Laminar, Re = 1100, coming from left to right at constant velocity Ux = 0.0045. The Inlet BC is located at the left edge of the fluid domain, at 10
chord = 2400 mm. The Outlet BC is located

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

49

Fig. 3 Stability of periodic response for lift force y moment

Fig. 4 (a) Dimension of Domain; (b) NACA 1412 geometries with chord length (CL) 240 mm

at 16
chord = 3840 mm from trailing edge, at the right edge of the fluid domain. Its distance is far from the airfoil to not affecting the developed state of flow [6, 7]. The exit pressure was defined as “gauge pressure” equal to zero as seen in Fig. 4. Moreover, the wall BC is applied to the airfoil (solid wall), with no-slip condition (Ux = Uy = 0), this is based on viscosity of boundary layer theory. The last, but not the least important, the symmetry condition is applied to upper and bottom edge of → → fluid domain because there is no fluid flow crossing these boundaries v
n = 0 . Dynamic Mesh, and User-Defined Function The Dynamic Mesh selected method is Diffusion-based smoothing, which is recommended when there is large deformation of mesh and involves rotational movement. Although this method is computationally costly than spring-based smoothing, diffusion-based generates better quality meshes [8]. The formulation for the diffusion coefficient selected is Boundary distance: γ=

1 dξ

ð11Þ

50

L. Gonzaga-Bermeo et al.

(a)

(b)

(c)

Fig. 5 (a) Meshing around airfoil NACA 1412; (b) Meshing detail at leading edge; (c) Meshing detail at trailing edge

Where d is the normalized boundary (0.50 mm, seen in Fig. 5) and ξ, the diffusion parameter, is taken as 1.5, since high values (0–2) preserve the mesh close to the moving boundary. Thanks to the programming of a User-Defined Function (UDF), the airfoil NACA 1412 is following the equations of movement for heaving and pitching described in Eqs. 1 and 2. Solver and Residuals The model to solve the Navier–Stokes equations is “Laminar” Model, [5],while the numerical solution used “SIMPLE” scheme. The time stepping is second-order upwind scheme. On the other side, the convergence was evaluated when residuals were 10-3 and -8 10 and when number of iterations were 20 and 60. Nevertheless, the results were the same. Hence, the residual accepted was 10-3 that allowed the optimization of computational time of process [9] (Fig. 6).

3 Results 3.1

Forces and Moments Acting on NACA 1412

The response curves for Lift and Moment at each studied parameter were used to calculate the extracted energy, considering the effective maximum angle of attack.

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

51

Fig. 6 Flowchart process

This angle helps to determine the Ry force that contributes, along with the moment, to the energy that can be extracted. The responses corresponding to the value of Ө0 = 62.01° are shown in Fig. 7 for the simulated values of f . The efficiency values for each simulation, shown in Table 3, were calculated using the second period of the Ry and moment responses obtained from the parameters shown in Table 2. From these values, it is shown that for f * = 0.11, Ө0 = 62.01°, and αmax = 27.36°, the maximum efficiency value obtained is 26.30%, which corresponds to a 67% improvement compared to the efficiency value obtained without considering an alpha angle, which was 16.42%.

52

L. Gonzaga-Bermeo et al.

Fig. 7 Responses for asymmetric airfoil at 0.11≤ f ≤0.2, θ0 = 62.01°: (a) Ry; (b) Moment Table 2 Analyzed angles of attack, θ0, considering the maximum effective angle αmax for each f* value studied

3.2

,

, deg

40.00 43.32 49.00 51.98 62.01 72.34 77.97 83.65 94.65 5.35

8.67

14.35 17.33 27.36 37.69 43.32 49.00 60.00

-1.34

1.98

7.66

10.64 20.67 31.00 36.63 42.31 53.31

-5.15 -1.83

3.85

6.83

16.86 27.19 32.82 38.50 49.50

-8.17 -2.49 11.49

0.49

10.52 20.85 26.48 32.16 43.16

deg

f*; (T, s) 0.11; 34.65° (484.84) 0.14; 41.34° (380.95) 0.16; 45.15° (333.33) 0.20; 51.49° (266.67)

Extract of Energy and Efficiency on NACA 1412

The simulations were conducted for f * values ranging from 0.11 to 0.20 and for pitching amplitudes between 40° and 95°, including the effect of the induced alpha angle, αi, which allows to compute the effective angle of attack, with its maximum value, αmax, as shown in Table 2. These values allowed to determine the amount of harvest energy and their efficiencies. It was observed that for the f * = 0.11 and Ө0 = 62.01°, the maximum efficiency value obtained was 26.30%. This indicates that the inclusion of the effective alpha angle has a positive effect, as it increases the efficiency of this device when used as an energy extractor (Table 3). The effect of the Heave amplitude, H0, on energy generation and the obtained efficiency was also analyzed when values of Ho/c vary from 0.5 to 1.5. To compare them, the kinetic parameters, such as f * = 0.11, Ө0 = 62.01°, αmax = 27.36° were used. The reader can see the influence of effective angle of attack in Table 4.

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

53

Table 3 Summary of the efficiencies obtained for all the considered parameters

0.11 3.92% (6.46%) 12.34% (17.60%) 16.42% (26.30%) 9.27% (27.31%) 1.77% (23.14%)

43.32° 51.98° 62.01° 72.34° 77.97°

0.14 -4.80% (4.62%) 5.36% (10.52%) 13.66% (22.34%) 10.55% (29.39%) 2.93% (30.06%)

0.16 -13.35% (7.78%) -0.79% (9.68%) 9.63% (17.18%) 7.16% (26.62%) -1.43% (26.91%)

0.20 -38.66% (-) -17.78% (19.30%) -2.67% (23.45%) -1.17% (26.10%) -11.63% (21.87%)

Table 4 Efficiency and extracted power for NACA 1412 at different H 0=chord, f  = 0.11, θ0 = 62.01 ° , αmax = 27.36 ° , Re = 1100

/ ≠

Energy, kWh Energy, kWh

0.5 31.99% 2.60E-8 22.87% 1.86E-8

1.0 26.30% 4.28E-8 17.20% 2.80E-8

1.5 18.73% 4.56E-8 10.78% 2.63E-8

Considering the factors mentioned earlier, the highest achievable efficiency stands at 31.99% when Ho/c equals 0.5, representing a significant 22% improvement compared to H0/c set to 1.0. Nonetheless, the efficiency plunges (-29%) when heave amplitude increases. On the other hand, in Fig. 8, it is evident that the extracted power does not see any improvement with changes in the effective angle of attack for H0/c above 1.0. Consequently, it becomes comparable to cases where no effective angle of attack is applied. It is concluded that when heave amplitude increases, the efficiency will drop up to the pitching amplitude that would be greater than the induced angle αi, thus entering in the regime for power extraction (feathering parameter >1). The angle of attack delays this plunge of extracted energy. Moreover, when there is no incidence of effective Angle of Attack, α, the extracted power tends to reduce as efficiency does due to local force reduced when non angle of attack is included and faster leaves the power extraction regime. Enhancements associated with heave remains valid only for Ho/c ratios less than or equal to 1.

54

L. Gonzaga-Bermeo et al.

Fig. 8 (a) Efficiency obtained, and (b) Extracted power in kWh for different at different H 0=chord, f  = 0:11, θ 0 = 62:01 ° , αmax = 27:36 ° , Re = 1100

3.3

Mapping of Influence of Effective Angle of Attack

From the results obtained, it can be observed that the inclusion of the effective angle of attack the Lift and Moment responses compared to the analyses that do not include this angle. Additionally, it is noted that the addition of this parameter allows for an increase in the amount of energy extracted in the NACA 1412 and improves efficiency, as observed in Tables 3 and 4. The efficiencies results have been mapped in Fig. 9, which shows a comparison between the results obtained without the effective angle of attack (9a) and with the addition of this angle (9b). These graphs present the surface for peak efficiency. It is observed in (9b) that for f * equal to 0.14, the efficiencies are greater than 20% in comparison when no angle of attack is added, (9a). In addition, the effective angle of attack makes that zero or negative efficiencies move away from f * = 0.20 conversely when non angle of attack is added.

4 Discussion For the calculations performed, computers with the following specifications were used: Core i7 processor, 8 cores, and 64 GB of RAM, allowing each analysis to be completed in approximately 2 h. The numerical models were validated using a mesh independence analysis, considering three different types of meshes (21 k, 48 k, 250 k nodes). This ensured that the results did not vary when modifying the number of nodes. This provides reliability in the values obtained through the simulations in this study. On the other hand, it was shown the effect of the effective angle of attack on efficiency of extraction energy, which at wrong rotational directional setting may not work as harvester of energy but get into the propulsion regime.

Analysis of Incidence of Angle of Attack on Energy Efficiency. . .

55

Fig. 9 Efficiency at 0.11 ≤ f  ≤ 0.2 at θ0 for NACA 1412: (a) with effective AoA, α = 0, (b) without AoA, α ≠ 0

Finally, due to influence of effective angle of attack, the lifting force is not the component that produces work; instead a Resultant Force must be calculated and oriented at effective angle alpha.

5 Conclusions and Recommendations From the results obtained, the following can be observed: The effective angle of attack increases the resultant force, R, on the airfoil. Therefore, its component in Y direction, Ry, will be higher than lift force itself. The addition of the effective angle of attack affects the amount of extracted energy and the obtained efficiency for an NACA 1412 airfoil, reaching efficiencies, η, greater or equal to 30%, when f * = 0.14, pitching angle between Ө0 = 74° - 78°, and effective angle of attack αmax, between 33° and 37°. Thus, optimal parameters for peak efficiency are possible in this region. The extracted energy values are very small, so it is necessary to consider conducting an analysis that includes other parameters, such as modifying the geometry of the analyzed NACA profile or performing a 3D analysis that considers the depth dimension, to evaluate how these modifications influence the obtained values. Reducing the Heave amplitude from Ho/c = 1.0 to 0.5 is advantageous as it leads to an increase in efficiency, although it does not have the same effect on extracted power. Importantly, efficiency noticeable decreases for Ho/c ≥ 1.0 due to feathering parameters decreasing to less than 1, changing the regime. Finally, for extracted power when there is no incidence of effective AoA, it tends to decrease faster, basically due to induced angle being higher than pitching amplitude (X < 1), and the local force does not generate enough power.

56

L. Gonzaga-Bermeo et al.

Recommendations for the future work are the following: To analyze the airfoils considering their three dimensions. In addition, the variations on the H0/chord ratio from 0.20 to 0.50) at frequencies 0.11 < f * < 0.15 with Δf  = 0.01, pitching amplitudes 73° ≤ Ө0 ≤ 80° with Δθ = 1°, to determine the response of these changes and generate a more detailed space of efficiencies, η( f, θ0). To analyze the pressure coefficient along the chord and the effect of the leadingedge vortex shredding. Examine the behavior of the same parameters studied under different regimes of incident fluid flow and evaluate how this variation in the Reynolds number (Re) influences the amount of energy extracted and its corresponding efficiency.

References 1. Gonzaga-Bermeo, L., Cuenca, C.A.: Analysis of two-dimensional airfoil models as harvesters of energy. Green Energy Technol., 91–105 (2022). https://doi.org/10.1007/978-3-030-97862-4_7 2. Xiao, Q., Zhu, Q.: A review on flow energy harvesters based on flapping foils. J. Fluids Struct. 46, 174–191 (2014). https://doi.org/10.1016/j.jfluidstructs.2014.01.002 3. Kinsey, T., Dumas, G.: Parametric study of an oscillating airfoil in a power-extraction regime. AIAA J. 46(6), 1318–1330 (2008). https://doi.org/10.2514/1.26253 4. McNeel, R., et al.: Rhinoceros 3D, Version 6.0. Robert McNeel, Seattle (2010) 5. ANSYS Inc.: Ansys Fluent Theory Guide, 15th edn. ANSYS, Canonsburg (2013) 6. Versteeg, H.K., Malaladekera, W.: An Introduction to Computational Fluid Dynamics, 2nd edn. Pearson Prentices Hall, Edinburgh Gate (2007) www.pearsoned.co.uk/versteeg 7. Blazek, J.: Computational Fluid Dynamics: Principles and Applications, 3rd edn. Elsevier, London (2015). https://doi.org/10.1016/C2013-0-19038-1 8. Ansys Training: Ansys Fluent Dynamic Mesh Modeling. Virtual – WebEx. https://www.ansys. com/training-center/course-catalog/fluids/ansys-fluent-dynamic-mesh-modeling (2019) 9. ANSYS Inc.: Monitoring Residuals. ANSYS FLUENT 12.0 User’s Guide. https://www.afs. enea.it/project/neptunius/docs/fluent/html/ug/node812.htm (2012). Accessed 01 Sept 2021

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador: Simulation Tool Juan Carlos Solano , Valeria Herrera Miguel Caraballo, and Aníbal Lozano

, Ángel Ordóñez

,

1 Introduction 1.1

Background

Grid-connected photovoltaic (PV) systems are designed to operate in parallel with the conventional electrical grid. These systems utilize solar panels, among other electronic components, to supply all or part of a building’s electricity demand, a concept known as self-consumption. The electricity generated by these systems depends on solar irradiance, which means that it may not always provide the exact amount of energy required by a household. To ensure a continuous power supply, the PV system remains connected to the electrical grid. In instances where the system generates more electricity than is needed or consumed, the excess energy is fed back into the grid. These systems have the characteristic of providing some autonomy and potential economic savings in billing, depending on the tariff or incentive scheme applied [1]. Internationally, important policies have been implemented to incentivize and increase the use of energy from renewable sources [2], particularly in the use of PV in buildings [3–5]. The advantages that PV systems have over other renewable energy sources lie in the reduction of PV module prices (about 85% in recent years [6]) and the growth of research on increasing the efficiency of PV cells [7, 8]. At the J. C. Solano (✉) Facultad de la Energía, Universidad Nacional de Loja, Loja, Ecuador Universidad Nacional de Loja, Centro de Investigaciones Tecnológicas y Energéticas CITE, Loja, Ecuador e-mail: [email protected] V. Herrera · Á. Ordóñez · M. Caraballo · A. Lozano Facultad de la Energía, Universidad Nacional de Loja, Loja, Ecuador © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_4

57

58

J. C. Solano et al.

same time, according to the International Energy Agency [9], the benefits of this type of system installed directly in buildings and connected in self-consumption mode are very varied, among which those related to the reduction of electricity costs (on the part of the consumer), those related to the repowering of the distribution grid, and the improvement in the security of electricity supply stand out. Likewise, the social and environmental benefits should also be highlighted insofar as economic profitability is improved with the reformulation of electricity market prices and energy efficiency is increased with less use of traditional energy sources. In Ecuador, the existing legislation is relatively new. It officially started in 2018 through the regulation ARCONEL003/18, which establishes the technical, commercial, and legal conditions for users to implement PV generation facilities [10]. Three years later, an update was established with regulation ARCERNNR-001/2021 [11]. Despite the incentives presented in these regulations, the installation of gridconnected PV systems has a low number of registrations (around 100 installations across the country by 2021) [12]. This may be due to various factors, starting from administrative procedures, cost of the systems, and mode of compensation, among others. According to [13], several schemes incentivize PV self-consumption. One of them is the Feed-in Tariff (FiT) model, where distributors pay a specific amount for the electricity the PV system generates and injects into the grid. Another compensation scheme is called Net Metering or Net Energy Metering (NEM), in which the electricity distribution company charges only the net value resulting from subtracting the energy consumed from the energy generated for self-consumption. Another incentive scheme is Net Billing, which, unlike the previous scheme, uses a bidirectional meter to record the energy demanded by the customer and the excess energy that is injected into the grid by the PV system, valuing them separately and at different prices [14]. In Ecuador, according to the regulations [11, 12], a Net Metering scheme is implemented, providing benefits to the “prosumer” (a combination of producer and consumer) since the reduction in consumed kWh directly reflects on the electricity bill. This scheme allows for the accumulation of surplus energy that can be consumed in subsequent billing periods, potentially resulting in positive energy balances. In Ecuador, the greatest efforts have been made toward the rural electrification of remote areas without connection to the interconnected electricity system [15], which is why there are a greater number of publications describing experiences with off-grid systems compared to on-grid systems. Another interesting particularity regarding the connectivity of PV systems is the purpose given to this generation source in systems that interact with the electricity grid on the Ecuadorian mainland: mostly used to supply power supply stations for electric cars and electric passenger trams [16], lighting systems [17], a water desalination plant located on Floreana Island [18], and low-voltage grids in sensitive areas such as the Galapagos Islands [19, 20]. Similarly, small-scale PV applications (100 kW) have typically been restricted to the fields of telecommunications (powering repeaters, antennas) and rural electrification (mini-grids, tourist homes, etc.) [21]. On the other hand, since approximately

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

59

2014, research has been conducted on PV systems combined with control systems to efficiently manage the electricity demand of small- and medium-sized consumers [22, 23]. In addition to these experiences, an analysis of the research conducted in Ecuador regarding grid-connected PV systems concludes that this field of study is relatively new, with research conducted to determine their reliability, optimization, and cost-effectiveness. For example, Zambrano-Asanza et al. [24] conducted an estimation of PV potential in urban environments based on architectural conditions and building consumption. This study concludes that (in one case study), savings of up to 16 metric tons of liquefied petroleum gas can be achieved simply by using PV systems and proper demand-side management. On the other hand, specific applications of electricity consumption in buildings such as lighting and air-conditioning systems, which use electricity generation from PV systems, have been studied in Ecuador [17, 23, 25]. These studies conclude that this type of consumption can be supplied solely with solar PV energy available in the building itself, achieving savings in annual billing of up to 58%. In the same context, other studies [26–28] conclude that Ecuador has great potential for the installation of PV in buildings, mainly due to its geographical location with latitudes close to zero, which allows PV systems to easily adapt to the existing slope of buildings without affecting their production. Other studies have also focused on economic analyses and user incentives for the implementation of grid-connected PV systems [29]. For example, Benalcazar et al. [14] concluded that grid parity could not be achieved without some kind of incentive. Among the Ecuadorian universities that have pioneered the installation of PV systems in their buildings, The Salesian Polytechnic University (UPS) can be mentioned (Cuenca campus) [30], as well as the Particular Technical University of Loja (UTPL) [31]. The UPS installed a PV system applied to electromobility in 2018. The system has an installed capacity of 13.2 kWp in an area of 73.3 m2 and serves for the self-consumption of buildings on its university campus and for recharging two motorbikes and two electric cars. On the other hand, in 2019, the UTPL installed one PV system composed of 68 panels of 275 watts that are connected to the electricity grid, which has allowed it to produce about 20% of the energy consumed in a building with the consequent economic savings in billing. Note that a few days ago, the UTPL installed the second stage of this system with an extension of 55 kWp, which generated an average annual energy of 85 MWh. Finally, it is worth mentioning that Ecuador already has a regulation for microelectricity generation [11]. This new regulation promotes PV generation for small prosumers, which will allow buildings to cover their electricity demand and deliver the surplus to the National Interconnected System. However, the use of this regulation is still limited in our country, especially due to a lack of knowledge and experimental studies that demonstrate or refute the technical and economic viability of grid-connected PV systems.

60

1.2

J. C. Solano et al.

The Aim of the Investigation

Against this backdrop, the purpose of this chapter is to present the results obtained through a simulation tool created at the National University of Loja, which allows any user to determine the profitability of a PV system in their home by estimating potential savings on the electricity bill and the level of self-consumption. The tool utilizes databases from various sources [32–36] to assess the solar potential at a specific location. The Perez model is then used to calculate the diffuse irradiance for tilted surfaces. Additionally, the characteristics of the PV system are entered as input variables (PV capacity, tilt, orientation, installation costs, among others), as well as the specific characteristics of the household (electricity consumption, location). The tool is programmed to automatically calculate detailed billing costs and, based on the PV self-consumption level, provide estimates of potential savings and return on investment. The results also include a sensitivity analysis to visually depict the outcomes of hundreds of simulations. This simulation tool is based on a grid-connected PV system installed in a building at the Universidad Nacional de Loja (UNL) (see Fig. 1), and through measurements taken from this system, a decision support system has been developed to size and analyze the technical and economic feasibility of installing PV selfconsumption systems in Ecuador.

2 Methodology The methodology followed to assess the technical and economic viability of residential PV self-consumption in Ecuador is presented in the following flowchart (see Fig. 2).

Fig. 1 Left: Aerial photograph of the 5 kWp PV system at UNL. Right: Data diagram showing realtime monitoring

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

61

Fig. 2 The methodology used to calculate self-consumption, billing savings, and return on investment for grid-connected PV systems in Ecuador

2.1

Solar Potential

The solar potential in Ecuador has been published in the Solar Atlas of Ecuador [35] and more recently in the Solar Map of Ecuador [36]. In addition to this information, there are international sources with databases from which solar potential can be extracted for a specific location. For example, the National Aeronautics and Space Administration (NASA) [32], National Renewable Energy Laboratory (NREL) [33], and METEONORM [34]. From all these sources, the location (latitude) and solar potential (monthly average daily expressed in kWh/m2) have been obtained for each parish in each canton of the provinces belonging to the Southern Region of Ecuador, which includes El Oro, Loja, and Zamora Chinchipe. These data are part of the simulation tool, but it is also possible to manually input data for any location in Ecuador, including specific data from meteorological stations if available. All the data obtained from the sources above provide Global Horizontal Irradiance (GHI). However, to accurately assess how the PV system is influenced by panel tilt, the Perez model (A new simplified version of the Perez diffuse irradiance model for tilted surfaces) [37] has been utilized. This model, through extensive calculations, allows for the determination of monthly and annual irradiation on the tilted plane of the panels. An example of how monthly irradiation varies on the horizontal plane and the tilted plane is shown in Fig. 3. These variations play a crucial role and directly influence the performance of the PV system.

62

J. C. Solano et al.

Fig. 3 Monthly irradiation on the horizontal plane (blue) versus monthly irradiation on the tilted plane (red). Example of use of the Perez model for the city of Loja using data from NREL

2.2

PV System Characteristics

As a first step, it is necessary to gather the general details of the PV system that will be installed. The input variables for the tool are as follows: nominal power of the PV generator (in kWp), panel efficiency (in percentage), shading factor (ranging from 0 to 1, where 0 represents no shading and 1 represents complete shading), characteristic performance ratio (PR, with a value between 0.7 and 0.9. Losses correspond to DC-to-AC conversion, cell temperature, maximum power point tracking, etc.), and the cost of purchasing and installing the PV system, which will depend on the specific supplier and brand of the equipment (the approximate price in Ecuador ranges between 1000 and 1500 USD/ kWp) [38]. The values of global irradiance on an inclined surface allow obtaining the generator’s productivity (YR). With this value, the annual PV energy (EPV) can be calculated using the following formula: EPV = PNOM,G × Y R × ð1 - FSÞ × PR

ð1Þ

Where: PNOM, G is the nominal power of the PV generator (in kWp) YR is the annual PV energy produced by the PV system (in kWh/m2) FS is the shadow factor (0–1) PR is the system’s efficiency percentage (%).

2.3

Electricity Bill

To determine the value that would be paid to the electricity company if a PV system is installed (compared to the amount paid previously without PV), it is necessary to

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

63

Fig. 4 Normalization of monthly average electricity consumption in the city of Loja

know the monthly consumption of the household, the type of contract (residential, commercial, etc.), and the cost per consumed kWh. For this purpose, this study has extensively utilized the Tariff Schedule of the public electricity service in Ecuador [39]. By knowing how much a user pays for monthly consumption, the amount that would be paid if the energy produced by the PV system is deducted for that month (applying the Net Metering scheme) and can be subtracted, ultimately determining the monthly savings. Since the consumption varies each month, this tool allows for inferring the range of monthly energy consumption based on statistical data on electricity consumption (Fig. 4). These data are based on information provided to the research team of this chapter by the Empresa Eléctrica Regional del Sur (ERRSA) from over 60,000 residential meters in the city of Loja. However, users can also manually input homogeneous monthly consumption or, if available, the specific historical consumption data of the household for further analysis.

2.4

Sensitivity Analysis

Finally, by knowing the PV self-consumption and the savings in billing, it is possible to calculate the time it would take for the user to recover the investment in the purchase and installation of the PV system. From here, a sensitivity analysis has been conducted using the tool to determine how these parameters (self-consumption, savings in billing, and return on investment) vary within specific ranges. In the case presented in this chapter, a sensitivity analysis is shown by varying the installed PV power from 0.5 kWp to 10 kWp. At the same time, the monthly electricity consumption has varied from 0 kWh to 3000 kWh. In this way, the resulting graphs from the sensitivity analysis allow an understanding of the ranges in which the installation of these systems is economically profitable for the user.

64

J. C. Solano et al.

3 Results All formulations, models, input and output variables have been initially programmed in an Excel workbook (Fig. 5). The model is complex and occupies many sheets, but the main screen is responsible for displaying the main parameters. The screenshot shows a report with the key parameters that the installer must consider. It displays the annual solar PV energy produced by the system, the required panel area, the kWh/ kWp ratio, the annual savings, the self-consumption percentage, the installation cost, and finally, the time it would take to recover the investment. The specific example shown in Fig. 5 represents a particular case with fixed power and consumption. However, if the PV power increases or decreases and if household consumption increases or decreases, a sensitivity analysis graph (Fig. 6) is presented to determine immediately the variations. These graphs serve as a fingerprint of the specific installation considering the location, tilt angle, and solar potential of the site. The three graphs of self-consumption, savings in billing, and return on investment are shown as complements to each other. It is not recommended to decide based on a single graph. It is necessary to relate the three output parameters for a better analysis. For example, for monthly consumptions greater than 1000 kWh, the installation of any PV system would be profitable (return on investment in less than 10 years). However, not every PV system can cover the entire demand. The self-consumption graph shows that the PV generator power should be greater than 6 kWp to achieve self-consumption percentages close to 100%. The investment recovery time represents only the time it would take for the user to recoup the investment made in the system purchase but does not guarantee that the system will cover the entire demand. On the other hand, the annual savings in billing is a parameter of interest to the user, as it is not always about recovering the investment, but rather knowing how much can be saved month by month to invest in other things.

Fig. 5 Screenshot of the simulation tool developed in Excel

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

65

Fig. 6 Sensitivity analysis graphs for a residential PV system installed in the city of Loja

The efficiency and reliability of this simulation tool are based on its validation through the comparison and contrast with monitored PV systems installed in various cities across the country. As an example, three case studies can be mentioned (see Table 1) of different PV systems, whose analysis and results were presented in previous research [40]. Furthermore, these experimental results align with studies on the segmentation of residential customers for PV self-consumption purposes using profitability indices in Ecuador [41].

66

J. C. Solano et al.

Table 1 Characteristics of the installed and monitored PV systems for case studies

Location Manta Quito Quito

Avg. monthly consumption (kWh) 2153 1926 4962

PV power (kWp) 8.25 5.23 31.20

Actual Avg. monthly PV production 1012 640 3065

Actual Payback time (years) ≈4 ≈3 ≈2

Simulated Avg. monthly PV production 950 670 4000

Simulated Payback time (years) 3.80 3.45 1.78

4 Conclusions In Ecuador, the regulation of grid-connected PV systems for self-consumption is relatively new (around 5 years), and therefore, the implementation of these systems is not widely spread across the country, with only around 100 registered systems according to the latest report. Additionally, the low electricity prices make profitability a topic that users need to analyze beforehand, depending on their consumption and installation costs. The simulation tool developed for grid-connected PV systems in self-consumption mode was designed to allow any residential user in Ecuador to input their data (such as consumption, location, and desired installed capacity), providing them with the potential savings in their electricity bills and the return on investment for purchasing a PV system. The electricity billing structure in Ecuador is designed such that higher consumption results in higher bills, which directly affects profitability. Combined with the fact that Net Metering is applied in Ecuador, installations of PV systems for electricity consumption exceeding 1000 kWh/month can recover the investment in less than 10 years for almost any system size. For lower consumption levels, profitability depends on the installed power and the desired level of self-consumption. In this regard, the simulation tool provides additional value compared to other studies, as it allows users to input their data and perform multiple simulations to find the break-even point that helps them decide whether investing in a PV system is beneficial for them or not. Up until the time of writing this chapter, the model and formulations have been developed in an Excel workbook. However, efforts are underway to create a web-based tool to make it accessible to all users. Furthermore, the simulations conducted by the current tool are being experimentally validated through a 5 kWp PV system installed at the UNL, and the values of electricity consumption, billing, and return on investment have been compared with three residential PV systems installed in various locations in Ecuador, with the simulated values aligning with the experimental ones. Acknowledgments The authors acknowledge the support of the “Universidad Nacional de Loja” by means of the research project: 34-DI-FEIRNNR-2021 “Desarrollo de un sistema de soporte de decisiones para el autoconsumo fotovoltaico en el Ecuador.”

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

67

References 1. Bastida Molina, P., Saiz Jiménez, J.Á., Molina Palomares, M.P., Álvarez Valenzuela, B.: Instalaciones solares fotovoltaicas de autoconsumo para pequeñas instalaciones: aplicación a una nave industrial. 3C Tecnol. (Edición 21). 6(1), 1–14 (2017) 2. WBCSD: Low Carbon Technology Partnerships Initiative (2016) 3. Agdas, D., Barooah, P.: On the economics of rooftop solar PV adoption. Energy Policy. 178 (2023). https://doi.org/10.1016/j.enpol.2023.113611 4. Johanning, S., Abitz, D., Scheller, F., Bruckner, T.: The influence of financial benefits and peer effects on the adoption of residential rooftop photovoltaic systems. In: International Conference on the European Energy Market, EEM, vol. 2023-June. https://doi.org/10.1109/EEM58374. 2023.10161765 (2023) 5. Mansouri Kouhestani, F., Byrne, J., Johnson, D., Spencer, L., Hazendonk, P., Brown, B.: Evaluating solar energy technical and economic potential on rooftops in an urban setting: the city of Lethbridge, Canada. Int. J. Energy Environ. Eng. 10(1), 13–32 (2019). https://doi.org/10. 1007/s40095-018-0289-1 6. IEA, PVPS Annual Report (2022) 7. Yang, T., Athienitis, A.K.: A review of research and developments of building-integrated photovoltaic/thermal (BIPV/T) systems. Renew. Sust. Energ. Rev. 66, 886–912 (2016). https://doi.org/10.1016/j.rser.2016.07.011 8. Biyik, E., et al.: A key review of building integrated photovoltaic (BIPV) systems. Eng. Sci. Technol. Int. J. (2017). https://doi.org/10.1016/j.jestch.2017.01.009 9. IEA: World Energy Statistics 2017. IEA, Paris (2017). https://doi.org/10.1787/world_energy_ stats-2017-en 10. ARCONEL, “Resolución No ARCONEL-003/18 ─ Regulación para la Microgeneración fotovoltaica para autoabastecimiento de consumidores finales de energía eléctrica,” Quito (2018) 11. ARCERNNR, RESOLUCIÓN Nro. ARCERNNR -001/2021 (2021). https://www. controlrecursosyenergia.gob.ec/wpcontent/uploads/downloads/2021/03/ResolucionARCERNNR-001-2021.pdf 12. ARCERNNR, Registro nacional de autorizaciones para consumidores con SFV, Quito (2021) 13. Patricio Munoz-Vizhnay, J., Vinicio Rojas-Moncayo, M., Raul Barreto-Calle, C.: Incentive pertaining to energy the generation distributed in Ecuador. Ingenius-Revista Cienc. Y Tecnol. 19, 60–68 (2018). https://doi.org/10.17163/ings.n19.2018.06 14. Benalcazar, P., Lara, J., Samper, M.: Distributed photovoltaic generation in Ecuador: economic analysis and incentives mechanisms. IEEE Lat. Am. Trans. 18(3), 564–572 (2020). https://doi. org/10.1109/TLA.2020.9082728 15. Urquizo, J., Singh, P., Hidalgo-Leon, R., Villavicencio, V., Soriano, G.: Rehabilitation of solar home systems and sustainable development for an island community in Ecuador. In: 2019 IEEE global humanitarian technology conference (GHTC), pp. 164–171 (2019) 16. Arévalo, P., Cano, A., Jurado, F.: Comparative study of two new energy control systems based on PEMFC for a hybrid tramway in Ecuador. Int. J. Hydrog. Energy. 45(46), 25357–25377 (2020). https://doi.org/10.1016/j.ijhydene.2020.06.212 17. Hidalgo, A., Villacrés, L., Hechavarría, R., Moya, D.: Proposed integration of a photovoltaic solar energy system and energy efficient technologies in the lighting system of the UTA-Ecuador. Energy Procedia. 134, 296–305 (2017). https://doi.org/10.1016/j.egypro.2017. 09.529 18. Amoroso, F., et al.: Simulations of solar power systems to provide electricity to a model water desalination plant in Floreana Island, Ecuador. In: Proceedings of the ASME 2021 15th International Conference on Energy Sustainability, ES 2021. https://doi.org/10.1115/ ES2021-62841 (2021)

68

J. C. Solano et al.

19. Morales, D.X., Besanger, Y., Alvarez Bel, C., Medina, R.D.: Impact assessment of new services in the galapagos low voltage network. In: 2016 IEEE PES Transmission & Distribution Conference And Exposition-Latin America (PES T\&D-LA) (2016) 20. Morales, D.X., Besanger, Y., Sami, S., Alvarez Bel, C.: Assessment of the impact of intelligent DSM methods in the Galapagos Islands toward a smart grid. Electr. Power Syst. Res. 146, 308–320 (2017). https://doi.org/10.1016/j.epsr.2017.02.003 21. Feron, S., Cordero, R.R., Labbe, F.: Rural electrification efforts based on off-grid photovoltaic systems in the Andean Region: comparative assessment of their sustainability. Sustainability. (2017). https://doi.org/10.3390/su9101825 22. Naranjo-Mendoza, C., Gaona, G., Jesús, L.-V., Labus, J.: Performance analysis with future predictions of different solar cooling systems in Guayaquil, Ecuador. In: Proceedings of the ASME 2014 8th International Conference Energy Sustainable ES2014 (2014) 23. Solano, J., Olivieri, L., Caamano, E.: HVAC systems using PV technology: economic feasibility analysis in commercial buildings of Ecuador. IEEE Lat. Am. Trans. 14(2), 767–772 (2016). https://doi.org/10.1109/TLA.2016.7437221 24. Zambrano-Asanza, S., Zalamea-León, E.F., Barragán-Escandón, E.A., Parra-González, A.: Urban photovoltaic potential estimation based on architectural conditions, production-demand matching, storage and the incorporation of new eco-efficient loads. Renew. Energy. 142, 224–238 (2019). https://doi.org/10.1016/j.renene.2019.03.105 25. Peralta-Jaramillo, J., et al.: Design of a variable refrigerant flow air conditioning system based on solar energy. In: Proceedings of the LACCEI International Multi-Conference Engineering, Education Technology, vol. 2022-July, pp. 1–9. https://doi.org/10.18687/LACCEI2022.1.1. 502 (2022) 26. Serrano-Guerrero, X., Alvarez-Lozano, D., Romero, S.F.L.: Influence of local climate on the tilt and orientation angles in fixed flat surfaces to maximize the capture of solar irradiation: a case study in Cuenca-Ecuador. In: 2019 IEEE International Autumn Meeting on Power, Electronics and Computing, ROPEC 2019. https://doi.org/10.1109/ROPEC48299.2019.9057102 (2019) 27. Zalamea-Leon, E., Barragan-Escandon, A., Mendez-Santos, P.: Assessment of photovoltaic potential on sloped roofs on Ecuatorial-Andean housing typology. A hundred houses in Cuenca Ecuador as case study. In: 2018 IEEE ANDESCON (2018) 28. Zalamea-León, E., Mena-Campos, J., Barragán-Escandón, A., Parra-González, D., MéndezSantos, P.: Urban photovoltaic potential of inclined roofing for buildings in heritage centers in equatorial areas. J. Green Build. 13(3), 45–69 (2018). https://doi.org/10.3992/1943-4618.13. 3.45 29. Barzola, J., Rubini, L.: Análisis técnico y financiero de grid parity residencial con fuente de energía solar. Yachana Rev. Científica. 4, 11–18 (2015). https://doi.org/10.1234/ych.v4i1.27 30. PV Magazine: La Universidad Politécnica Salesiana de Cuenca, en Ecuador, inaugura su instalación fotovoltaica – pv magazine Latin America. https://www.pv-magazine-latam. com/2018/12/06/la-universidad-politecnica-salesiana-de-cuenca-en-ecuador-inaugura-suinstalacion-fotovoltaica/ (2018). Accessed 19 Sept 2020 31. UTPL: UTPL: pionera en autogenerar energía eléctrica en la Zona 7. https://noticias.utpl.edu.ec/ utpl-pionera-en-autogenerar-energia-electrica-en-la-zona-7 (2019). Accessed 10 Sept 2020 32. NASA Langley Research Center: Data Access Viewer. https://power.larc.nasa.gov/data-accessviewer/ (2023). Accessed 05 Feb 2023 33. NREL: National Solar Radiation Database. https://nsrdb.nrel.gov/data-viewer (2023). Accessed 01 Mar 2023 34. Meteotest: Meteonorm Version 8. https://meteonorm.com/en/meteonorm-timeseries (2023). Accessed 01 Mar 2023 35. CONELEC: Atlas solar del Ecuador con fines de generación eléctrica. http://biblioteca.olade. org/opac-tmpl/Documentos/cg00041.pdf (2008). Accessed 03 Feb 2020 36. Vaca-Revelo, D., Ordóñez, F.: Mapa solar del Ecuador 2019, 1st ed. Quito (2019)

Economic Analysis of Residential Photovoltaic Self-Consumption in Ecuador:. . .

69

37. Perez, R., Seals, R., Ineichen, P., Stewart, R., Menicucci, D.: A new simplified version of the perez diffuse irradiance model for tilted surfaces. Sol. Energy. 39(3), 221–231 (1987). https:// doi.org/10.1016/S0038-092X(87)80031-2 38. EL UNIVERSO: Interés en energía solar se duplica, la inversión puede ir de $ 1.000 a $ 40.000 para vivienda y en empresas sobrepasar $ 1 millón. https://www.eluniverso.com/noticias/ economia/interes-en-energia-solar-se-duplica-la-inversion-puede-ir-de-1000-a-40000-paravivienda-y-en-empresas-sobrepasar-1-millon-nota/ (2023). Accessed 12 June 2023 39. ARCERNNR: Pliego tarifario del servicio público de energía eléctrica. https://www. controlrecursosyenergia.gob.ec/wp-content/uploads/downloads/2022/05/Pliego-TarifarioServicio-Publico-de-Energia-Electrica_-Ano-2022.pdf (2022). Accessed 13 June 2023 40. Ordóñez, Á., Solano, J.C., Sánchez, E.: Techno-economic analysis of grid-connected photovoltaic systems for self-consumption in ecuador: experimental cases in the coastal and highland regions. In Congreso Internacional I+D+i Sostenibilidad Energética (2022) 41. Paul, J., Cárdenas, M., Heredia, J.: Segmentación de clientes residenciales con fines de autoabastecimiento fotovoltaico mediante índices de rentabilidad en Ecuador Segmentation of residential customers for photovoltaic self-supply purposes using profitability indexes in Ecuador. Rev. I+D Tecnológico. 18(2), 32–43 (2022)

The Role of Curved Buildings in Urban Wind Carlos Walter

and Jorge Lässig

1 Introduction Wind is a resource that exists worldwide and, under certain orographic conditions, can significantly increase its energy potential. The so-called concentrator effects can significantly improve the average speed of the wind flow. These effects do not occur only in nature but also in urban environments, which makes it recommendable to study as an energy resource. Integrating wind turbines into buildings is becoming a new possibility for energy efficiency. It has begun to be studied in university research centers on wind energy, such as the Technical University of Delft in the Netherlands. This university began to study the use of small-scale wind turbines in Dutch cities such as Amsterdam, The Hague, Tilburg, Twente, and the United Kingdom [1]. Placing a wind turbine above a building is a possibility to obtain electricity, taking advantage of the effect of wind acceleration on it. Wang et al. [2] discussed this possibility. On the other hand, Andrew Grant et al. [3] studied turbines located at vertical and horizontal axes inside ducts, which are installed on the edges of the terraces of high buildings, taking advantage of the suction that occurs on them. Lin Lu and Ka Yan Ip [4] performed simulation studies in Computational Fluid Dynamics (CFD) with different sets of buildings and obtained that the increase of wind speed between those buildings can be between 1.5 and 2 times the average speed of the wind, increasing the available wind power three to eight times. On the other hand, Nalanie Mithraratne [5] assessed the possibility of installing small wind

C. Walter · J. Lässig (✉) Present Address: College of Engineering, Universidad Nacional del Comahue, Neuquén, Argentina e-mail: [email protected]; [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_5

71

72

C. Walter and J. Lässig

turbines on roofs of houses in New Zealand and concluded that it could reduce carbon emissions in that country by 26–81%. Buildings shape can achieve that the wind speed accelerates above them up to 20% of the average in the urban environment; then this can be used as an energy resource. There are several studies that have quantified this effect in different buildings (Baines [6], Jeongsu Park et al. [7], Lassig [8] and [9], Sander Mertens [10], Vries [11]).

2 Circular Section Buildings The wind accelerations on the sides of buildings still need to be deeply studied. There are some conceptual examples where it is intended to place wind turbines on the sides of a cylindrical building (see Fig. 1), taking advantage of the acceleration of the wind at the first angles of attack. The flow around a circular cylinder is a very complicated phenomenon and has been the subject of interest in numerous studies in fluid mechanics, both numerical and experimental. It is dependent on the Reynolds number, but in the case of cylindrical buildings, this is very high, and therefore, the flow pattern corresponds to a separation of the boundary layer on the cylinder of turbulent type, as schematically indicated in Fig. 2. Fig. 1 Artist’s impression of a cylindrical building with vertical axis turbines [6]

The Role of Curved Buildings in Urban Wind

73

Fig. 2 Cylinder at high Reynolds numbers, whose boundary layer develops in turbulent flow (self-made image)

Fig. 3 Isotachs around a cylinder: the red zone is the highest speed

Conceptually, the acceleration of the wind when hitting a cylindrical obstacle can be seen in Fig. 3, where the colors represent isotachs, with red being the highest speed and blue the lowest speed. In potential flow, the acceleration of the wind on the sides at an angle θ from the point of impact of the wind on the cylinder can be estimated with the next equation: R uϴ = sin ϴ 1 þ u0 r

2

ð1Þ

This equation is represented in Fig. 4. The flow condition closest to the building wall is for r/R = 1, and at θ = 90°, the speed is doubled. As the power to be extracted from the wind is given Eq. (2), and the speed is cubed, then the power would be multiplied by 8 for a given speed. P= where:

1 :ρ:V 3 :CP :A 2

ð2Þ

74

C. Walter and J. Lässig

Fig. 4 Distribution of velocities on a cylindrical body according to the angle of attack ϴ (self-made image)

– – – – –

P: Power (Watt). ρ: Air density (kg/m3), V: Wind speed (m/s). CP: Coefficient of power of the wind turbine (-). A: Surface area of the wind turbine facing the wind (m2).

The aforementioned is for an ideal flow without friction or vortices; the buildings are in a turbulent environment, and their surfaces are rough. Therefore, giving values of wind speed increments on their sides less than those described above, but the example illustrates the acceleration of the wind near its wall and the increase in wind power possible to extract.

3 Curved Buildings Curved buildings accelerate the wind as it is done on the surfaces of airfoils. Many architects, when designing this type of building, use the shape of the circular arc airfoil, as illustrated in Figs. 5 and 6.

4 Methodology The present work concentrated on studying how and where the wind accelerates on a building with arc airfoil shapes, as indicated in Fig. 7.

The Role of Curved Buildings in Urban Wind

75

Fig. 5 Housing complex in Cipolletti, province of Río Negro, Argentina. Left (aerial view). Right (front view) Fig. 6 Circular Arc airfoil with a thickness of 32% of its chord

Fig. 7 Air flow lines in the wind direction at an angle of attack of θ = 0° (aerial view)

4.1

CDF Simulations

For the study, CFD trading software, based on the OpenFOAM® software package, was used in the first instance. It was simulated with a building 60 m high and 30 m front (or airfoil chord). The turbulent model used was the k-ε (see Fig. 7).

76

C. Walter and J. Lässig

Fig. 8 1:200 scale model tested in the wind tunnel

4.2

Wind Tunnel Simulations

To validate the software, scale tests were carried out in the wind tunnel belonging to the Environmental Fluid Dynamics Laboratory (LaDiFA) of the Faculty of Engineering of Universidad Nacional del Comahue. It is 7.40 m long with a test section of 0.90 m × 0.90 m, it is open (atmospheric boundary layer type) with a 6 CV electric motor with speed regulator/stabilizer (Fig. 8). The measurement of wind speed inside the tunnel was carried out with a hot film anemometer (CEM DT-8880), and to measure the turbulence in the wake of the building, a Pitot tube connected to piezoelectric pressure sensor was used, and this to a data logger that recorded the measurements at a frequency of 1000 Hz. The simulated boundary layer profile was for an exponent of Sutton’s law corresponding to an exposure between Class II (suburb) and Class III (city) of CIRSOC Standard 102 (2005), so the resulting velocity profile is shown in Fig. 9.

5 Results Wind velocities are obtained with the software (see Figs. 10, 11, 12 and 13), and we compared the percentage increase that is generated on the surface of the building at different heights, according to different wind directions (Table 1). Results presented in Table 1 showed that velocity increases depend on the relative direction of the wind and the chord of the building, and the highest ones are given at an angle of attack of 0° and are decreasing as the angle of attack grows. The increase in wind speed is close to 30% at an angle of attack of 0°.

The Role of Curved Buildings in Urban Wind

77

Fig. 9 Average speed/ Speed at 60 m vs Elevation in suburb, city, and tunnel environment

Average Speed 1.0000 0.9000 0.8000

Elevation (m)

0.7000 0.6000 0.5000 0.4000 0.3000 0.2000

1.20

1.00

0.80

0.60

0.40

0.20

0.0000

0.00

0.1000

Vm/Vm60 Vm tunnel

Vm city Vm suburb

Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Z Y

X

Fig. 10 Velocity field at an angle of attack of 0°: color red represents the highest velocity (slicer plane, right view)

78

C. Walter and J. Lässig Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Z X

Y

Fig. 11 Velocity field at an angle of attack of 0°: color red represents the highest velocity (slicer plane at a height of 45 m, aerial view)

Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Z Y

X

Fig. 12 Velocity field at an angle of attack of 30°: color red represents the highest velocity (slicer plane, right view)

The Role of Curved Buildings in Urban Wind

79 Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00

Z Y

X

Fig. 13 Velocity field at an angle of attack of 30°: color red represents the highest velocity (slicer plane at a height of 45 m, aerial view)

Table 1 Percentage increase (%) of wind speed on surface of building (Vs) relative to free wind (V0), according to the direction of the wind (angle of attack) at different heights Angle of attack (deg) 0 0 0 0 30 30 30 30 60 60 60 60 90 90 90 90

Height (m) 15 30 45 60 15 30 45 60 15 30 45 60 15 30 45 60

V0 (m/s) 14.73 16.06 17.16 18.13 14.73 16.06 17.16 18.13 14.73 16.06 17.16 18.13 14.73 16.06 17.16 18.13

Vs (m/s) 19.7 21.04 21.83 20.28 19.3 20.54 20.88 20.2 18.23 19.31 19.94 20.19 16.5 17.78 18.98 19.98

Percentage (%) 34 31 27 12 31 28 22 11 24 20 16 11 12 11 11 10

80

C. Walter and J. Lässig

6 A Case Study In the city of Cipolletti, province of Río Negro, in Patagonia, Argentina, there is a housing complex of three curved buildings based on the circular arc airfoil with some modifications (see Figs. 14 and 15). Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 Y Z

X

Fig. 14 Distribution of wind velocities over the housing complex when the wind is from the southwest (SW)

Y Z

X

Velocity [m/s] 22.04 20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00

Fig. 15 Distribution of wind velocities over the housing complex when the wind is from the west (W)

The Role of Curved Buildings in Urban Wind

81

Fig. 16 Wind rose in Cipolletti

The buildings are oriented between the west and southwest, which coincides with the maximum frequencies of the wind rose (see Fig. 16). It is possible to calculate the wind power on the surface of the building using Eq. (2). It shows that wind speed is the most impactful variable because it is cubed. This means that a 30% increase in speed leads to an increase close to 120% in wind power.

6.1

Weather Data

Our laboratory has an automatic weather station (Anemos™) installed in the Universidad Nacional del Comahue campus, which is only 10 km away from the Cipolletti housing complex, so the data obtained can be taken as representative of the wind at impact on buildings. Figure 17 shows registration data for 1 year (2010). The average annual wind power is defined as: Pm =

1 ρ 2

1

f ðvÞ  V 3 dv

ð3Þ

0

where f(v) is the wind distribution function for the measurement location. As a result, the potential measured at our weather station was 122 W/m2 for all wind directions, only about 105 W/m2 corresponding to the west direction. If we consider the potential 105 W/m2 and apply a speed increase on the side of the building, which is 30%, then the power obtained per square meter rises to 230 W/m2. With small vertical axis turbines, and assuming a CP of 0.30, over a year, about 606 kW/year per square meter will be produced.

82

C. Walter and J. Lässig

Fig. 17 Records of average wind speed on the Universidad Nacional del Comahue campus, during the year 2010

Observing Figs. 10 and 11, the red zone (maximum wind acceleration) represents about 400 m2 of its side, so the building could produce about 242 MW/year with small vertical axis turbines, not an insignificant value.

7 Conclusions If the building is built in a place where the wind rose presents a preferential direction, as in many places in Patagonia, then properly orienting the building it is possible to extract wind power from its sides. In the case of the building studied, it is estimated that the possible energy savings are between 10% and 16% of the annual energy consumption, where this range includes values found by other researchers mentioned above. Many buildings, with many small wind turbines on their sides in a city, could contribute to saving a significant percentage of the energy demand to the grid and also to savings in electrical infrastructure. In Argentina, particularly, the legislation in many provinces during the last 5 years allows domestic users to inject electricity into the electrical grid on lowand medium-voltage lines. The advantage of distributed generation is that it does not require any type of investment by the Government and allows investments in large power plants to reduce between 20% and 40% for local energy. The cost and the necessary investment are not analyzed in this work since this is the subject of other variables that escape wind engineering. Acknowledgments The work was supported through subsidies from the Secretariat of Science and Technology of Universidad Nacional del Comahue.

The Role of Curved Buildings in Urban Wind

83

References 1. Jadranka Cace, RenCom; Emil ter Horst, HoriSun; Katerina Syngellakis, IT Power; Maíte Niel, Axenne; Patrick Clement, Axenne; Renate Heppener, ARC; Eric Peirano, Ademe, Urban Wind Turbines – Guidelines for Small Wind Turbines in The Built Environment. http://www. urbanwind.net (2007). Accessed 10 Sept 2009 2. Wang, F., Baia, L., Fletcher, J., Whiteford, J., Cullen, D.: The methodology for aerodynamic study on a small domestic wind turbine with scoop. J. Wind Eng. Ind. Aerodyn. 96, 1–24 (2008) 3. Grant, A., Johnstone, C., Kell, N.: Urban wind energy conversion: the potential of ducted turbines. Renew. Energy. 33, 1157–1163 (2008) 4. Lu, L., Ip, K.Y.: Investigation on the feasibility and enhancement methods of wind power utilization in high-rise buildings of Hong Kong. Renew. Sust. Energ. Rev. 13, 450–461 (2009) 5. Mithraratne, N.: Roof-top wind turbines for microgeneration in urban houses in New Zealand. Energ. Buildings. 41, 1013–1018 (2009) 6. Baines, W.D: Effects of velocity distribution on wind loads and flow patterns on buildings. In: Wind Effects on Buildings and Structures, Procedures of Symposium No 16, pp. 198–225. National Physical Laboratory, Teddington (1965) 7. Park, J., Jung, H.-J., Lee, S.-W., Park, J.: A new building-integrated wind turbine system utilizing the building. Energies. 8, 11846–11870 (2015) 8. Lassig, J., Valle Sosa, J., Jara, U., Palese, C.: Determining the location of wind turbines on roof building with the help of wind tunnel. IOSR J. Mech. Civil Eng. 13, 08–13 (2016) 9. Lassig, J., Jara, U., Valle Sosa, J., Palese, C.: Urban environment: characterization of the wind in flat roofs. In: The Age of Wind Energy. Springer, Cham (2020) 10. Mertens, S.: Wind Energy in the Built Environment. Concentrator Effects of Buildings. MultiScience (2006) 11. Vries, E.: Conversation with Eize de Vries and Visit to a Small Wind Turbine on the Roof of a Mid-rise Building that Suffers from Frequent Yawing (2006)

Part III

Miscellaneous

Advances in H-Type Darrieus Turbines for Urban Environments in Colombian Territory Carlos V. M. Labriola, Andres F. Galindo Rojas, Elizabeth A. López, and Javier A. Rosero García

1 Introduction Eolic applications in the last 20 years have not all been successful, particularly those of large wind turbines integrated into buildings [1], given that there is an emphasis on the architectural aspects to make a great impression on the integration but not on the environmental elements, such as noise and vibrations of the turbines when rotating, that were transmitted to the building’s structure, making the project’s permanence unfeasible. Subsequently, those impacts were reduced with the consequent cost overruns over the original project thanks to the reengineering of the projects. In the case of small Horizontal Axis Wind Turbines (HAWTs) whose blades are made of very flexible plastic materials, their rotors hum in response to variations in wind speed produced by gusts. One notable example of these are Patagonian winds in Argentina [2]. Furthermore, these turbines must be oriented to the maximum incidence of the wind; for that reason, when there are sudden gusts of wind, they must change their orientation, which produces premature wear of their yaw system and the consequent friction, noise, and vibrations [3]. One way to solve that problem is to use a device that physically restrains the yaw system to limit the orientation of HAWT placed on the terraces’ edge. However, over time, winds can produce bothersome vibrations and noises within the physical limits of the yaw system, which eventually will cause an increase in installation and maintenance costs and, in the worst-case scenario, a reduction in the turbine life cycle.

C. V. M. Labriola Faculty of Engineering, Universidad Nacional del Comahue, Neuquén, Argentina A. F. Galindo Rojas (*) · E. A. López · J. A. Rosero García Universidad Nacional de Colombia, Bogotá, Colombia e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_6

87

88

C. V. M. Labriola et al.

On the other hand, small Vertical Axis Wind Turbines (VAWTs) can be omnidirectional; for this reason, they do not need an orientation system. The heavier parts are in the lower half of the turbine, closer to the ground; this gives the structure more stability and makes maintenance easier. In particular, Darrieus vertical axis turbines are the only turbines that, provided that their rotor works by lift force with a minimum drag force, have an efficiency similar to that of horizontal axis turbines. However, one disadvantage of Darrieus turbines is their inability to self-start (usually due to symmetrical profiles) [4]. In this context, this work details the design of a non-symmetrical, cambered airfoil profile that allows starting an H-rotor Darrieus turbine in light winds under environmental conditions in Bogotá and San Andrés, Colombia. Given that the profile is provided with a variable geometry that would allow controlling the power output once its nominal speed has been reached, this helps the final brake once the speed has reached its permanent maximum for a given time, avoiding its quick deterioration.

1.1

Low Reynolds Number Profiles

To modify the profiles of the blades of the Darrieus Type H wind turbine and not have to use an airflow concentrator, the aerodynamic profile with the best aerodynamic characteristics must be selected. This implies finding the profile that presents the best output torque and power coefficients for the turbulent flow conditions of the area where the wind turbine would be located. CFD (computational fluid dynamics) techniques and different turbulent flow models are used to solve the RANS (Reynolds-Averaged Navier–Stokes) equations. Such is the case of [5], which, based on simulations with finite volumes in the Ansys-Fluent software, which had a simulation domain of 25 times the diameter of the turbine, found that the asymmetric profile NASA LS(1)-0413 had a power coefficient (cp) improvement of up to 16% over the NACA 0021 profile. Likewise, the fact that asymmetric profiles have been studied to operate in small H-Type Darrieus vertical axis wind turbines stands out as a promising solution for the self-starting problem [6] because they produce higher power coefficients and better airflow adherence, especially for low wind speeds, such as those that can occur in urban environments [7]. For this reason, starting from asymmetric profiles, it is possible to make changes to the thickness or the deflection angle of the trailing edge of the profile to improve its performance: • Ibañez, E 2019: From modified NREL s813 profiles with a deflection angle of 2
to 15% of the chord, it was possible to increase the ratio between lift force and drag (L/D) by 78.4% [8]. • Martinez et al.: They managed to increase cp by up to 15% by reducing the thickness of NREL s815 profiles [9].

Advances in H-Type Darrieus Turbines for Urban Environments. . .

1.2

89

Profile Design Development

The development of new airfoil profiles started based on NREL s813 and NASA LS (1)-0413 profiles, first performing an analysis with the Java Foil program to determine those with the best L/D ratio for Reynolds Number (Re) between 10,000 and 90,000 (as was made in previous studies [10]). Once the L/D performance of the profiles was defined, their geometry began to be varied, looking for an improvement in L/D and finding the best two airfoil profiles with changes in their thickness and with the addition of a deflection angle in the trailing edge (Fig. 1), achieving two profiles whose L/D ratio increased between 13.44% and 20.6% for the original starting profile. That was necessary because the H-Type Darrieus turbines work mainly by lift force, so achieving the highest lift force with the lowest drag force is desirable. After obtaining the best results for the original profile, the two best airfoil profiles were compared through Ansys-Fluent, with a CFD domain of 25 times the turbine diameter, the turbulence model SST K-ω, and applying the sliding mesh method with a fixed inflow of 5 m/s, concluding that the GLM4 (see Fig. 2) airfoil profile achieved the best performance with a maximum cp of 0.26 at a TSR (tip–speed ratio) of 2.5 (see Fig. 3). In addition, it was studied Re over the airfoil profile as a result of changing their shape to produce stopping in the wind turbine’s operation (as was reviewed based on the results of previous studies [13]). The profile was bent between 35
and 70
, at one point or at two points along the chord line (c) to achieve the total bent. Of all the variants studied (Fig. 4), three variations of the geometry of the GLM4 profile were chosen according to the lowest Re (Fig. 5) obtained in each simulation, corresponding to the greatest braking. Airfoil profile

Re =10000 Re =90000 L/D max L/D max

LS(1)-0413 LS(1)-0413-Mod-4 S813

7.5 9.0 8.7

21.8 24.7 23.6

10.9

30.7

Increase % 20.6

13.44 s813-Mod-5

Fig. 1 L/D ratio comparison of airfoil profiles Fig. 2 GLM4 profile designed based on LS(1)-0413 profile

Description NASA Low wind speed serie [11] 4 °of deflection angle on 25% of the chordline. NREL for HAWT [12] Reduction of the maximum thickness to 10%. 2° of deflection angle on 25% of the chordline.

90

C. V. M. Labriola et al.

Fig. 3 Power coefficient GLM4 and s813 modified profile

Fig. 4 Simulated cases and chosen cases according to Reynolds number

It can be seen that the best case is with the 70
variation, followed by option 4 of 35 and then option 2 of 35
. The selected cases are being tested in a wind tunnel in a two-blade H-Type Darrieus model (swept area 48 cm  32.8 cm) from previous studies [11], with increasing tunnel speed, to evaluate the braking power concerning the H-Type Darrieus rotor. Once the tests in the wind tunnel have been completed, the results of the simulations with CFD and those carried out in the tunnel are compared.


Advances in H-Type Darrieus Turbines for Urban Environments. . .

91

Fig. 5 Geometry variations in GLM4 profile selected by minor Re

Fig. 6 Wind profile in Bogotá and San Andrés locations

2 Prototype in Colombia 2.1

Study of the Resource

The wind resource was studied for two possible places to install a wind turbine: Bogotá and San Andrés Island. For Bogotá (with an air density of 0.9 kg/m3), only in time slots around noon and early afternoon, the winds were 3 m/s or higher. On the other hand, in San Andrés, the wind was constant almost all day, with a mean wind speed of 7 m/s and a density of 1.2 kg/m3, concluding that the San Andrés conditions turned out to be better for the installation of wind turbines for urban applications. However, Bogotá’s location provides the minimum requirement for first-time wind turbine assessments. To sum it up, both locations’ wind profiles are shown in Fig. 6.

92

2.2

C. V. M. Labriola et al.

Turbine Prototype

A prototype turbine with an H-Type Darrieus rotor with a vertical axis was designed in Bogotá, with dimensions of 1 m blade length (H ) and 1 m diameter (D) to develop a nominal power of 300 W at 7 m/s, which can be a larger and more powerful scalable model for San Andrés. The wind turbine’s main characteristics are shown in Table 1. The design process that has been followed is shown in Fig. 7. It highlights four design stages:

Table 1 H-Type Darrieus Turbine prototype characteristics Turbine characteristics Total height [m] Turbine support base [m  m] Chord length (c) [m] Number of blades (n) Solidity (σ) H/r (r ¼ radius) Optimal TSR Connection point Supports per blade Total mass [kg]

Unit/ description 2.3 1.5  1.5

Equation-reference – –

0.12 3 0.36 2 2.2

– – σ ¼ nD∙ c 2 < H=r < 3 [14] TSRopt ¼ 2.693 ∙ σ 0.329  1.605 [15] [16] [17]

50% over chord length 2 at 20.07% of the blade tip 84

Fig. 7 Design process flowchart

–

Advances in H-Type Darrieus Turbines for Urban Environments. . .

93

• Mechanical structure design: The worst-case loads were analyzed for a wind speed of 18 m/s, equivalent to the onset of a tropical storm. A safety factor of more than 2 was achieved for each part. • Power system design: A 300 W generator at 500 RPM was selected and connected to a Mppt(Maximum Power Point Tracking) controller, which was connected to a battery charging system for powering the sensor and communication systems. • Sensor selection: For monitoring the turbine’s operation, sensors were selected to measure the wind speed and the current and voltage generated. • Communication system design: For signal capture, an Arduino UNO was connected to a Raspberry Pi 4 with an internet connection thanks to a router with a 3G /4G connection. The data are sent to the OSIsoft Pi System platform for visualization and storage.

3 Verifications and Test After the installation of the prototype, it was possible to verify the self-starting with a 3 m/s breeze without mechanical or electrical impulse. The position of the supporting arms relative to the chord line of the blades was optimized, and the protrusions of screws and nuts on the blades and flanges of the central axis were eliminated to improve rotor efficiency. The wind turbine has a multipolar Nd-Fe-Bo magnet generator so as not to use a multiplier, which made it heavier and required a greater moment of inertia for starting, reducing the overall efficiency. The generated power goes through a power inverter, a voltage regulator that connects to the load and batteries with a 220 V AC inverter. A communications system supervises the turbine to measure the generated voltage and current, which can be used to calculate rotor speed. In addition, a bicycle brake can stop the turbine when a local anemometer reports wind speeds higher than 18 m/s (the wind speeds of a tropical storm). The prototype was installed on the roof of the Faculty of Economics of the Universidad Nacional of Colombia (UNAL), Bogota, as is shown in Fig. 8 (left). Additionally, a nearby solar-powered weather station provides data on wind speed, wind direction, temperature, pressure, humidity, and solar radiation (Fig. 8 right). Moreover, a pyramidal lower base has been designed to support the communication and electronic systems; the whole structure is attached to the roof by the weight of sandbags to provide an easy installation without the need for deadbolts to anchor the prototype to the building roof. Also, it was made of standardized angle iron, whose structure was conceived to be separated into two parts for easy handling during installation and transport. Finally, the weather station’s data are sent through a Raspberry Pi 4, which receives the values via MODBUS RS-485 serial while simultaneously capturing the other signals of the wind generator provided by an Arduino UNO connected to

94

C. V. M. Labriola et al.

Fig. 8 H-Type Darrieus prototype designed at UNAL Bogotá (left), weather station (right) Fig. 9 Conceptual design for H-Type Darrieus prototype for the Botanical Garden of San Andrés Island

another Raspberry Pi. After being captured and saved on the Raspberry Pi, these data are sent to the Pi System platform through a modem with a SIM card.

4 Planned Applications A survey of possible places for a second prototype has been made on the island of San Andrés, Colombia. It was defined as the place of installation of the wind turbine on the roof of the viewpoint of the Botanical Garden in charge of the UNAL Caribe headquarters (Fig. 9). Another expected application is using the GLM4 profile in hydrokinetic turbines, Fig. 10, to obtain electrical energy for marine and fluvial applications using a horizontal axis turbine in rivers, mighty channels, and marine applications [18].

Advances in H-Type Darrieus Turbines for Urban Environments. . .

95

Fig. 10 Arrangement on shore for in situ hydrokinetic turbine rotor test in Río Li-may, Neuquén, Patagonia, Argentina

5 Conclusions The bibliographical review clearly showed the advantages of the small VAWT applications over the HAWT ones in an urban context, which is why the focus of this chapter is on the application of the new profile to an H-Type Darrieus turbine of quick and simple construction. The asymmetric profiles can increase self-starting in H-Type Darrieus turbines, especially those modified by lower Reynolds number airfoil families like NASA LS (1)-04XX and NREL S8XX with augmentations in the L/D ratio between 13.44% and 20.6%. CFD simulation shows that GLM4 achieves a cp of 26% for a TSR of 2.5 at a low wind speed of 5 m/s. Furthermore, the prototype shows self-starting at a low wind speed of 3 m/s. To achieve good performance in small turbines, emphasis should be placed on eliminating all the slight imperfections in the rotor and optimizing the positions of the stem and blades. All the efficiency improvements and cost reductions in these wind turbines are essential to compete with other renewable energy equipment on the market and to allow integration among all of them so that their installation and maintenance can also be carried out by the generating users connected to the grid.

6 Recommendations It is worth mentioning that the wind should be studied in higher areas of Bogotá as mountain ranges around the city, which may have higher wind speeds at the cost of a slight reduction in air density. In the case of San Andrés, a coral island with a high cordon of about 30–50 m along its length, the wind will be studied at similar heights in buildings to see the wind potential of those locations.

96

C. V. M. Labriola et al.

The mechanical resistance of the blades can be improved by using composite materials (fiberglass or carbon with plastic resins). Still, this would only be justified for large-scale turbine production due to their cost. The results of the simulations should be compared to be trusted; consequently, tests must also be carried out in wind tunnels or in situ. Acknowledgments This research was supported by Electrical Machines & Drives research Group (EM&D) from Universidad Nacional for supporting this research and project Think Green on the island of San Andres, BPIN 2016000100002 EEDAS ESP.

References 1. Bahrain World Trade Center Building, Architect W.S. Atkins. https://es.wikiarquitectura.com/ edificio/bahrain-world-trade-center/. Accessed 12 May 2023 2. Labriola, C.: Work Shop, Construcción de aspas de pequeñas turbinas eólicas. Univesidad Nacional de la Patagonia Austral, sede Caleta Olivia, Santa Cruz (2016) 3. Palese, C.: Análisis del Recurso Eólico, curso de la Especialización de Energía Eólica. Facultad de Ingeniería, Universidad Nacional del Comahue (2021) 4. Riquelme, J.: Integración y Análisis de pequeñas turbinas eólicas en entornos urbanos Pequeñas turbinas eólicas. Editorial de la Universidad de Nueva León, México (2023) 5. Mohamed, M., Dessoky, A., Alqurashi, F.: Blade shape effect on the behavior of the H-rotor Darrieus wind turbine: performance investigation and force analysis. Energy. 179, 1217–1123 (2019) 6. Sengupta, A., Biswas, A., Gupta, R.: Comparison of low wind speed aerodynamics of unsymmetrical blade H-Darrieus rotors-blade camber and curvature signatures for performance improvement. Renew. Energy. 139, 1412–1142 (2019) 7. Mazarbhuiya, H., Biswas, A., Sharma, K.: Low wind speed aerodynamics of asymmetric blade H-Darrieus wind turbine-its desired blade pitch for performance improvement in the built environment. J. Braz. Soc. Mech. Sci. Eng. 42 (2020) 8. Ibáñez, E.: Simulación de perfiles asimétricos para Turbinas Eólicas de Eje Vertical. Universidad Nacional del Comahue (2019) 9. Martinez, R., Urquiza, G., Castro, L., Garcia, J., Rodriguez, A., Tenango-Pirin, O., Dávalos, J., Caldiño, U.: Shape effect of thickness of the NREL S815 profile on the performance of the H-rotor Darrieus turbine. J. Renew. Sustain. Energy. 13, 01330 (2021) 10. Galindo, A.F., Rosero, J.A., Labriola, C.V.: Estudio de potencial de generación eólica y diseño de perfil asimétrico para un aerogenerador Darrieus tipo H. In: 2022 IEEE Biennial Congress of Argentina (ARGENCON), pp. 1–8 (2022). https://doi.org/10.1109/ARGENCON55245.2022. 9939833 11. McGhee, R.J., Beasley, W.D., Whitcomb, R.: NASA low and medium -speed airfoil development. NASA Tech. Memo. 78709(974), 22 (1979) https://ntrs.nasa.gov/citations/19800012809 12. Tangler, J.L., Somers, D.M.: NREL airfoil families for HAWTs. https://doi.org/10.2172/ 10106095 (1995) 13. Labriola, C.: Thesis: Aerodynamic Brakes for Vertical Axis Wind Turbine for Patagonia Argentina. University of Reading, London (2000) 14. Hand, B.P., Kelly, G., Cashman, A.: Aerodynamic design and performance parameters of a lifttype vertical axis wind turbine: a comprehensive review. Renew. Sust. Energ. Rev. 139 (2021). https://doi.org/10.1016/j.rser.2020.110699

Advances in H-Type Darrieus Turbines for Urban Environments. . .

97

15. Rezaeiha, A., Montazeri, H., Blocken, B.: Towards optimal aerodynamic design of vertical axis wind turbines: impact of solidity and number of blades. Energy. 165, 1129–1148 (2018). https:// doi.org/10.1016/j.energy.2018.09.192 16. Bianchini, A., Balduzzi, F., Ferrara, G., Ferrari, L.: Influence of the Blade-Spoke Connection Point on the Aerodynamic Performance of Darrieus Wind Turbines, V009T46A012. https://doi. org/10.1115/GT2016-57667 (2016) 17. Ahmadi-Baloutaki, M., Carriveau, R., Ting, D.: Straight-bladed vertical axis wind turbine rotor design guide based on aerodynamic performance and loading analysis. Proc. Inst. Mech. Eng. Part A J. Power Energy. 228, 742–759 (2014). https://doi.org/10.1177/0957650914538631 18. Labriola, C.: FAIN 4/242, Proyecto de investigación Estudio de sistemas Fluidodinámicos para la Patagonia Argentina. Facultad de Ingeniería de la Universidad Nacional del Comahue (2022)

Assessment of Green Hydrogen Production from Hydropower in Ecuador Ángel Recalde , Víctor Acosta, Giordy Ortiz, Ricardo Cajo and Carolina Godoy

,

1 Introduction Due to its favorable geographical location, Ecuador power system counts on a largely hydro-based generation system. During the rainy season, spillage in main dams is almost inevitable, causing the release of important water flow that could be used in other activities. The rainy season in general is between December and May; however, it is split into two main sub-seasons, that is, from February to May and from October to November. This amount of water could deliver other advantages in complementary processes or industries including the production of hydrogen, which has been attractive since the deployment of renewable energies at medium and large scales [1]. Ecuador has made significant progress in terms of energy access, with high coverage of electricity services (>97.29%) by 2021 [2]. The country has pursued an energy policy called “change in the energy matrix” aimed at achieving a high renewable energy share in its power generation matrix, primarily focusing on the development of large-scale hydropower infrastructure under the central government’s guidance [3, 4]. As a result, by March 2023, hydropower has reached a substantial 73% share of electricity generation, while other renewable energy sources remain limited at 1.55% [5]. According to the current Master Electricity Plan, Ecuador is on track to meet its renewable energy targets for the overall power

Á. Recalde (✉) · V. Acosta · G. Ortiz · R. Cajo Escuela Superior Politécnica del Litoral, ESPOL, Facultad de Ingeniería en Electricidad y Computación, Guayaquil, Ecuador e-mail: [email protected] C. Godoy Technical University of Munich – Renewable and Sustainable Energy Systems, München, Germany © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_7

99

100

Á. Recalde et al.

matrix, and future planning includes expanding hydropower capacity by an additional 3.6 GW by 2027. It is important to mention that most of the future projects are conceived as run-of-river power plants, i.e., without reservoirs [3]. Under this scenario of hydropower growth, Ecuador has a lot of potential in producing green hydrogen as an energy vector and chemical raw material for various industrial processes [3]. Presently, Ecuador faces a deficiency in national policies pertaining to hydrogen production regulations and the establishment of a domestic market that promotes competitive pricing and enhanced efficiency. However, preliminary studies have emerged, exploring the prospects of integrating green hydrogen production within the Ecuadorian context. For instance, [6] highlights the importance of renewable energy sources, particularly hydropower, in facilitating sustainable hydrogen production, and emphasizes the potential role of hydrogen in achieving Ecuador’s decarbonization goals. [7] examines the feasibility and implications of producing hydrogen through hydropower, considering the country’s abundant hydro resources. It analyzes the benefits and impacts of integrating hydrogen-based transportation systems, including reduced emissions, energy efficiency, and enhanced sustainability. Green hydrogen is defined as the hydrogen obtained through renewable energy sources such as wind, solar, biomass, and other green sources [8]. Until recently, the significance of green hydrogen requirements in the region had been largely overlooked. However, the COVID-19 pandemic and subsequent challenges in gas energy trading in Europe, coupled with the Russian–Ukrainian conflict, exposed the vulnerabilities of the existing energy trade systems. In response, European countries have set new targets to diversify their energy markets and achieve sustainable goals. The abundance of hydroelectric capacity in South America, including countries like Brazil, Paraguay, Colombia, and Ecuador, has garnered significant attention from governments worldwide. Countries like Germany, with ambitious renewable energy targets and a desire to reduce dependence on fossil fuels, recognize the potential of South America’s hydroelectric capacity. The stability and reliability of hydropower as a renewable energy source make it an attractive option for countries seeking to transition to a low-carbon economy. Governments are actively exploring partnerships and collaborations to leverage South America’s hydroelectric potential and facilitate the import of clean energy, such as green hydrogen, generated through hydropower. This presents opportunities for both South American countries with abundant hydro resources and countries like Germany seeking to diversify their energy sources and reduce carbon emissions. In line with the Ecuadorian plans for expanding the generation system, which includes the development of hydropower, Ecuador has positioned itself as a potential supplier of green hydrogen for Germany and other markets [9]. In this context, the primary objective of this study is to analyze and validate the feasibility of hydrogen production using hydropower as the main energy source. The key contributions of our research encompass the following:

Assessment of Green Hydrogen Production from Hydropower in Ecuador

101

• Estimating the potential of green hydrogen production in Ecuador by assessing the availability of surplus hydropower and water resources. • Analyzing the trading opportunities and economic aspects associated with green hydrogen production, particularly considering the utilization of excess water from hydro dams. • Assess the influence of hydraulic generation costs on the economic viability of green hydrogen production through electrolysis. The investigation was conducted utilizing data obtained from official Ecuadorian institutions, including the Transmission Operator, the National Dispatcher, and information published in the National Planning Portfolio for the electrical system generation, transmission, and distribution systems [3]. This work is divided as follows. In Sect. 2, hydrogen technology and green hydrogen state of the art are explored. Then, Sect. 3 expands on the current state of the Ecuadorian electrical system and its potential based on the hydropower resources. Later, Sect. 4 presents the complete feasibility analysis for hydrogen production considering available data and information from government institutions. Then, Sect. 5 provides results and analysis. Finally, conclusions and recommendations are presented in Sect. 6.

2 State of the Art Hydrogen gains attention as a clean energy source, useful as fuel or energy carrier. It is pivotal for decarbonizing transport, industry, and power sectors [10, 11]. Hydrogen, essential in the energy sector, comes from diverse sources. Global demand rebounded to 94 Mt in 2021, surpassing pre-pandemic levels. Conventional and new sectors contributed, with progress seen in steel, hydrogen applications, and pilot shipping projects. Power projects could reach 3.5 GW by 2030. However, policies suggest 115 Mt hydrogen demand by 2030, short of climate goals (130 Mt) and net zero (200 Mt) by 2050 [12]. According to the 2022 report of the IEA [12], the primary pathway for hydrogen production was through natural gas utilization, which amounts to approximately 94 million tons per year, constituting 62% of the total. Additionally, about 18% of hydrogen emerged as a secondary product during naphtha reforming within refineries. This incidental hydrogen was subsequently harnessed for various other refinery operations like hydrocracking and desulphurization. Another notable source was coal, contributing 19% to the overall hydrogen output in 2021, with a significant presence in China. Minimal quantities of hydrogen (less than 1%) were derived from oil resources. The predominant utilization of fossil fuels for primary hydrogen production significantly contributes to noteworthy emissions of carbon dioxide (CO2), leading to the release of over 900 million metric tons of CO2 annually. Presently, the average carbon emissions intensity linked to worldwide hydrogen production stands at approximately 12–13 kg CO2-equivalent per kilogram of hydrogen (kg CO2-eq/kg H2). This spectrum in emissions intensity encompasses diverse allocation

102

Á. Recalde et al.

approaches employed for the by-product hydrogen generated within refineries [13]. As an alternative, green hydrogen has emerged as a key element to achieve net-zero emissions from heavy industry and transport and help to boost renewables in the energy mix and decarbonize energy-intensive industries [14]. Green hydrogen is technically produced via electrolysis, which employs an electrolyzer with an anode and cathode separated by an electrolyte. Electricity splits water molecules, yielding hydrogen at the cathode and oxygen at the anode [15]. Proton Exchange Membrane (PEM) electrolysis, using a solid polymer membrane, is efficient at lower temperatures [16]. Alkaline electrolysis, with an alkaline solution, suits large-scale production due to higher temperatures [17]. Solid Oxide Electrolysis Cells (SOECs), using ceramic oxides, offer efficiency at elevated temperatures [18]. Renewable sources like solar/wind power are crucial for carbon-free hydrogen. Ongoing research and development aim to enhance efficiency, cost, and scalability of these technologies for wider adoption. Focusing now on production costs, as most of the hydrogen production relies on natural gas, the production cost is around 1.40 US$/kg. When factoring in CO2 costs of approximately 50 US$ per tonne, the cost increases to 1.80 US$/kg. Including additional expenses for carbon capture and storage to reduce 75% of the CO2 emissions would raise the cost of fossil-based hydrogen to approximately 2.20 US $/kg. In contrast, green hydrogen costs range from (3.40–6.60) US$/kg, resulting in an average cost gap of approximately 3 US$/kg between fossil-based and renewable hydrogen [19]. Green hydrogen has the potential to facilitate the attainment of net-zero carbon dioxide (CO2) emissions in energy-intensive sectors such as steel, chemicals, long-haul transport, shipping, and aviation, which are challenging to decarbonize. However, to make green hydrogen economically viable on a global scale, production costs need to be reduced. Cutting green hydrogen production costs is vital for wider adoption. While low electricity costs are important, they’re not enough. Decreasing investment expenses for electrolysis is key. IRENA targets a 40% short-term and up to 80% long-term reduction. This involves optimizing electrolyzer design, scaling, materials, operations, applications, and learning. Reduced electricity costs, lower electrolyzer capex, efficiency gains, and optimized operation could lead to an 85% long-term cost decrease [20].

3 Ecuadorian Current Situation Electricity demand in Ecuador has been steadily increasing, so generation system has been progressively upgraded to accommodate new hydroelectric power plants due to Ecuador’s strategic location and its water resources, which start in the Andean mountains and flow down to the Amazon region. By 2023, the Ecuadorian installed capacity reached 8886.06 MW, from which 58.43% corresponds to hydropower plants, 2.83% corresponds to other renewables technologies (wind, photovoltaic, biomass, and biogas), and 38.73% from nonrenewable plants [5].

Assessment of Green Hydrogen Production from Hydropower in Ecuador

103

Table 1 Paute Integral hydroelectric generation system [3]

Mazar Molino Sopladora Cardenilloa a

Power [MW] 170 1100 487 596

Average annual production [GWh] 800 4900 2770 2356

Reservoir level difference [m] 146 668 363 370

Design flow [m3/s] 140 170 150 180

Turbines 2 Francis 10 Pelton 3 Francis 6 Pelton

To be implemented in the future

Ecuador has been actively pursuing the development of its renewable energy sector since the last century, with a particular focus on hydroelectric power. One notable hydroelectric facility in the country is Paute Integral, which harnesses the water flow from the Paute River, averaging around 110 m3/s of water flow. Situated in the provinces of Azuay, Cañar, and Morona Santiago, the Paute Integral facility comprises three existing power plants: Mazar, Molino, and Sopladora. Additionally, there are plans for the construction of a fourth power plant, Cardenillo, in the future. The configuration and components of the Paute Integral Hydroelectric facility are detailed in Table 1. The existing hydropower plants in the Paute Integral, namely Mazar, Molino, and Sopladora, are interconnected in a series configuration, where the water flow from one plant is utilized as the inflow for the next. The process begins with the Paute River supplying water to the Mazar Dam, and the resulting turbined flow, along with spill water and other inflows, becomes the total inflow for the Amaluza Dam, which supplies the Molino power plant. The turbined water from the Molino plant is then directed to an interconnection chamber, where it is divided into spill water and a flow that serves as the inflow for the Sopladora power plant. The spilled water from the Amaluza dam, interconnection chamber, and the turbined water from Sopladora are eventually merged back into the Paute River downstream. Other renewable energy potential different than hydroelectric has been studied and implemented as well, however, this potential is still relatively small compared to hydroelectric. According to a study in [21], the suitable land for photovoltaics is around 9.3% of the Ecuadorian continental land area, and the promising area for photovoltaics hardly reaches the 0.32%. This means that only 820 km2 can be used at 100% for continuous power generation with a total average energy production of 6482 GWh/year. On the other hand, wind power could account for 884.2 MW of feasible capacity with a total average production of 1517.2 GWh/year. In this study, these findings are considered the most conservative. Alternate research, like [22], suggests a feasible PV installation area of 3312 km2, i.e., a greater area, but it is a less conservative value. The deployment of such infrastructure is far from being attractive in the short term, while hydro is promoted as having greater technoeconomic advantages.

104

Á. Recalde et al.

4 Green Hydrogen Production Feasibility Green hydrogen production requires an intensive and progressive investment portfolio, so that the projects are feasible not only in the year of installation but also during the following years [13]. The feasibility of green hydrogen involves assessing electrolysis plant electricity demand, load forecast, generation expansion, and water availability. By comparing generation and load demand, future energy surplus can be identified, while water availability guides production estimates. Power system planning examines generation expansion and load demand, ensuring security and reliability. Green hydrogen introduces a unique demand: sufficient energy during electrolysis and ample water resources. It relies on both factors, setting it apart in its dual dependence on electricity and water.

4.1

Power Demand Forecast

Power demand shows a monotonically increasing trend in every power system due to population increase and industrial development. Accurate forecasting plays a crucial role in the efficient operation and planning of electric utility companies, as well as in the decision-making processes of energy suppliers, generators, and market participants. Load forecasting can be categorized into short-term (up to 1 week), medium-term (1 week to 1 year), and long-term (over 1 year) forecasts and can be determined by using several analytic methods based on statistics and advanced techniques such as artificial intelligence [23]. One of the most common statistical methods is the econometric method, which uses regression to fit historical data to understand what could be expected in the future. In the Electrification Master Plan [3], the Ecuadorian demand has been projected to 2027 considering three different scenarios (low, average, and high) based on the consumers’ differentiation and the probability of occurrence of consumers growth. The demand variation has been modeled using the Gross Domestic Product (GDP) as the key parameter for the demand projection. Based on the average case, and using the projection of the Electrification Master Plan, the demand has been further projected to 2032 using extended simple exponential smoothing, an econometricbased method [3]. 2032 has been chosen to account for a horizon period of 10 years starting in 2022. The Holt’s extended simple exponential smoothing allows the forecasting of demand with a linear trend [24]. This method involves a forecast equation and two smoothing equations, one for level and the other for trend. The forecast equation is F tþhjt = St þ nGt The level equation can be written as

ð1Þ

Assessment of Green Hydrogen Production from Hydropower in Ecuador

105

Fig. 1 Power demand 10-year forecast, from 2022 to 2032

St = αF t þ ð1 - αÞðSt - 1 þ Gt - 1 Þ

ð2Þ

Finally, the trend equation is Gt = βðSt - St - 1 Þ þ ð1 - βÞGt - 1

ð3Þ

Where St denotes the estimate of the level of the series at time t, Gt is an estimated of the trend (slope), α is the smoothing parameter for the level, 0 ≤ α ≤ 1, and β is the smoothing parameter for the trend, 0 ≤ β ≤ 1. By following Holt’s method, the electrical demand was projected to 2032 as shown in Fig. 1. Starting in 2027, a noticeable decrease in the slope becomes apparent, diverging from the preceding years’ upward trajectory. This shift represents a more cautious outlook, mirroring the prevailing global landscape marked by potential economic contractions among major world powers due to the ongoing conflict between Ukraine and Russia. This conservative projection acknowledges the uncertainty stemming from geopolitical tensions and their potential impact on economies worldwide. An average of 3.13% of annual increase is expected for load demand growth.

4.2

Power Generation expansion

The Electrification Master Plan 2018–2027 [3] presents a power generation expansion plan (portfolio) based on the demand growth and other parameters including industrialization, economic targets, energy adequacy, sustainable targets, among others. It lists each generation project that enters the power system every year with its estimated capacity. The total power capacity up to 2032 is depicted in Fig. 2.

106

Á. Recalde et al.

Fig. 2 Total power generation capacity up to 2032

Total power generation capacity must be balanced against power demand and operational resources such as ancillary services and security reserve. The addition of special industrial demand could have been part of the power demand forecast as a step increment in the year of implementation. However, power for green hydrogen production will come from the spill obtained in hydro dams, i.e., the power produced if the spill was used as normal inflow to supply the additional power for electrolysis plants. Even when any amount of spill could exist, the power generation capacity allows the determination of green hydrogen production upper limit. For instance, spills could be generous at some point, but generation capacity is a constraint that limits the amount of spill that could be turbined before the generation system gets to its maximum. Besides, the feasibility analysis considers the spare power left once the power balance has been performed including ancillary services and reserve. In this case, it has been considered that ancillary services and reserve account for 20% of the installed generation capacity for each year, that is, 20% is allocated and cannot be used in the feasibility analysis. Any spare power left could be potentially used for green hydrogen production, given that spill is available. Although Fig. 2 depicts all generation including nonrenewables, it can be assumed that contract arrangements between electrolysis plants and hydroelectric plants could be done such that only renewable hydro energy is utilized. After considering the power balance between generation and demand including allocation of 20% of generation capacity for ancillary services and reserve, the average amount of power that could be available in the next decade is around 1330 MW as can be seen in Fig. 3 (hatched section in gray color). This available power can be considered as excess power, and it is continuously increasing every year, and this feature offers an optimistic landscape for green hydrogen production. The total excess power might not be used completely on green hydrogen production because it could not be economically feasible to increase the electrolysis plant

Assessment of Green Hydrogen Production from Hydropower in Ecuador

107

Fig. 3 Power balance 2022–2032 in Ecuadorian power system

capacity every year. A baseline production for the decade could be adopted in this case of industrial production of hydrogen.

4.3

Spilled Turbinable Energy

Hydroelectric power plants equipped with dams offer control over water inflow, spillage, and turbined water. During the rainy season, there is a significant increase in spillage due to the dams reaching their maximum capacity. As a result, the power plants are unable to operate at full capacity, and the energy potential of the water is discharged as spillage without useful utilization. This spill is said to have an excess energy, which is commonly known as Spilled Turbinable Energy, and it is a crucial factor in evaluating the feasibility of green hydrogen production in Ecuador as previously mentioned. Mazar, Molino, and Sopladora are the main hydroelectric power plants with water dams Mazar and Amaluza, from the Amazon Basin whose main inflow comes from the Paute river. Although there are other minor dams such as Daule Peripa (Marcel Laniado hydro power plant), Pisayambo (Pucara hydro), and Baba, Mazar, and Amaluza are significantly bigger compared to the rest of dams. The average spills in these dams during the years 2016–2021 are detailed in Table 2. The energy that can be obtained from spill water can be calculated using the following expression: E spill = ρwater gVh

ð4Þ

108

Á. Recalde et al.

Table 2 Average spill in Paute Integral dams during the years 2016–2021 [Hm3] [25]

January February March April May June July August September October November December

Table 3 Maximum and minimum levels of Paute Integral dams [m.a.s.la]

Maximum level Minimum level a

Mazar 6.48 0 31.28 12 202.89 145.11 93.40 17.16 0 7.02 9.90 7.34

Mazar 2153 2007

Amaluza 3.64 4.78 51.04 61.78 262.90 300.64 215.45 65.36 1.68 21.22 12.94 7.06

Amaluza 1991 1323

meters above sea level

Where g is the gravity constant, V is the monthly volume of the spill, h is the level difference in the dam, i.e., the difference between the maximum level just before spill starts and the minimum level. The minimum level can be assumed to be the river level downstream of the power plant. The maximum and minimum levels are listed in Table 3, according to CENACE. Having detailed the average spill volume and dam levels in Table 2 and Table 3, the energy that could be obtained from the spill can be calculated using (4). The average results for a year can be depicted in Fig. 4. The Paute Integral dams could potentially provide an estimate of 1.48 × 108kg H2/year considering the monthly average results presented in Fig. 4. Mazar and Amaluza dams are hydraulically connected in series, and both are within the two most important dams in the Amazon basin, reason for which these are the most promising candidates for green hydrogen production. The monthly energy potential, i.e., spill availability of Fig. 4 presents large variation throughout a year, so that a strategy to produce green hydrogen must be presented. These strategies are presented as scenarios in Sect. 5.

4.4

Cost Breakdown

To estimate the economic feasibility of a green hydrogen project, the power plant costs, the energy costs, and the production cost must be determined. The electrolysis plant considered in the cost estimation is the SIEMENS SILYZER 300 [26]. It contains a total of 24 modules with PEM-WE technology

Assessment of Green Hydrogen Production from Hydropower in Ecuador

109

Fig. 4 Monthly average spilled turbinable energy for Paute Integral (Mazar and Amaluza dams)

and a power capacity of 17.5 MW. Its nominal hydrogen production is 335 kg H2/hour. Its efficiency is always greater than 76.5%, which allows a relatively economic implementation. The calculation of production cost involves electrical energy cost Celect and water cost CH2 O , investment cost Cinv, and operation and maintenance costs COM. All these cost components serve to obtain a unitary projection cost, which is defined as

cunitary =

total projected cost projected quantity of hydrogen produced

ð5Þ

Let the cost of electric energy and water used in the electrolysis be defined as C elect = qH2 ρEU cgen

ð6Þ

C H2 O = qH2 O ρH2 O cH2 O

ð7Þ

Where qH2 is the amount of hydrogen to be produced, and qH2 O is the volume of water used to produce hydrogen. ρEU is the amount of energy consumed by the SIEMENS SILYZER 300 to produce 1 kg of hydrogen, in this case kWh ρEU = 53:53 kg H2 ; cgen is the generation cost involving the following costs: 0.020 US$/kWh is the energy cost in Paute Integral, 0.011 US$/kWh is the excess generation cost (generation what is scheduled by the power system operator), and 0.0068 US$/kW is the associated transmission cost, so cgen = 0:036USkg H2 . ρH2 O is the amount of water required to produce 1 cubic meter of hydrogen, 3 ρH2 O = 0:001 mm3HH2 O2 , cH2 O is a referential cost of water in Azuay province according to water distribution company ETAPA [27] cH2 O = 4:20USm3 H2 O. This cost is

110

Á. Recalde et al.

slightly lower than in [27] assuming a preferential water price based on a government-driven or private agreement contract. Now, the cost of investment can be written as Cinv = Ep cCEL AF

ð8Þ

cCEL is the cost of hydrogen production that depends on the amount of hydrogen to be produced. Based on available information from manufacturers presented in [28, 29], a relation between unitary cost of electrolysis per production capacity can be obtained, such that an approximation can be determined by the following equation: cCEL = 2424:9γ - 0:1062

ð9Þ

Where γ is the flow of water in m3/hour. Considering 3750 m3/hour (which is equivalent to 1.041 m3/s or 1041 l/s), cCEL = 1012:7USkW. Ep is defined as the ratio between the amount of electrical energy required to produce a certain quantity of hydrogen and the number of hours of availability of the electrolysis plant AE, i.e., q 2 ρEU E p = HAE . Investments’ values are typically depreciated with time; devaluation happens in the facility life cycle. To account for the depreciation, the investment cost is annualized, and it uses AF as an annuity factor, AF =

ð1 þ iÞn i ð1 þ i Þn - 1

ð10Þ

Where i is the interest rate, and n is the depreciation period in years; for the industrial sector in Ecuador i = 0.05 and n = 10 as a reasonable assumption. Finally, the operation and maintenance will be assumed to be 7% of the total investment cost, so that COM = 0.07Cinv. Now cunitary can be written as cunitary =

C elect þ C H2 O þ C inv þ C OM qH 2

ð11Þ

By utilizing the cost definitions provided earlier, it is possible to calculate the cost of hydrogen production. In the future, as the green hydrogen project is implemented, the currently wasted water will be harnessed by turbines to generate electricity for hydrogen production. To determine the available electrical energy for the electrolysis process, the power balance between the generation expansion and the demand forecast is assessed. Additionally, the amount of water that will be used for electrolysis is considered as spilled turbinable energy to estimate the amount of hydrogen that can be produced in a SIEMENS SILYZER 300. The spilled turbinable energy is now assessed in different scenarios to determine the annual schedule for hydrogen production.

Assessment of Green Hydrogen Production from Hydropower in Ecuador

111

5 Results and Analysis Once the demand has been projected and the Spilled Turbinable Energy (STE) has been identified, the feasible scenarios for green hydrogen production considering the cost breakdown detailed in Sect. 4 can be proposed. Two operation scenarios are proposed to manage peak electrical demand. In the first, the plant operates 3904 h annually. It runs 12 h daily in June and July and 8 h otherwise. Operation begins at midnight to avoid peak power demand coincidence. In the second scenario, the electrolysis plant is working 8410 h in a year; thus, the electrolysis plant works most of the day every day. In both scenarios, STE water is turbined to produce additional energy to supply electricity for the electrolyzer plant, that is, electricity from the grid is used in both cases such that hydropower is used to produce green hydrogen through the STE utilization. Projected at 3.13% annual growth, power demand contrasts with the expansion plan. Identifying 1333 MW of underutilized capacity, potential for green hydrogen emerges. This power could theoretically yield 1.48 × 108 kg hydrogen yearly, yet practical limitations apply. Spill-to-power balance prevents all water from converting to energy, reducing real production compared to theory. The contrast between underutilized generation capacity and the amount of spilled turbinable water drives an accurate estimation for hydrogen production. Besides, the spilled turbinable energy depicted in Fig. 4 shows significant variations throughout the year. Considering the generation capacity and spilled water contrast, only 4.45 × 105 kg of hydrogen per year could be obtained, which is equivalent to an annual average of 24.6 GWh of spilled turbinable energy per year. Paute Integral is the ideal candidate due to water access and proximity to the power grid. From 2016 to 2021, Mazar and Molino contributed 21.87% of Ecuador’s energy with a 54.20% utilization [30]. This suits green hydrogen production at the Sinincay Substation, 8 km from Cuenca’s industrial zone. The substation gathers 230 kV energy from Mazar and Molino, housing a 167 MVA transformer at 230/69/ 13.8 kV. Connected at 69 kV, it surpasses the 10 MW medium-high voltage threshold. Water from Cuenca’s dam serves the plant, with Yanuncay river’s 4 m3/s flow supporting hydrogen amidst irrigation and supply needs. With location, connectivity, and 24.6 GWh Spilled Turbinable Energy established, the candidate electrolysis plant’s production can be determined. Due to spill fluctuations, two scenarios emerge, based on: • The number of hours per year that the electrolysis plant is producing hydrogen, having a daily schedule avoiding peak hours. As mentioned in Sect. 4, the SIEMENS SILYZER 300 has been chosen due to its relatively economic cost of production at a high electrolysis efficiency. The scenarios have the following schedules: • Under Scenario 1, around 79.9 × 103 kg H2/month is feasible, rising in June and July, with a yearly total of 1000 tons hydrogen. Paute’s energy use becomes 0.58%, or 30.7 GWh/year, with an 8-h daily schedule.

112 Table 4 Production costs for Scenarios 1 and 2

Table 5 Hydrogen sale figures for Scenarios 1 and 2

Á. Recalde et al. Type Tons of H2 produced Production cost US$/kg H2

Type Tons of H2 produced Hydrogen sale price US$/kg H2

Scenario 1 1000 5.34

Scenario 1 1000 8.34

Scenario 2 2720.7 3.34

Scenario 2 2720.7 6.34

• Under the second scenario, annual hydrogen production can reach up to 2720.7 ton H2/year or 205.3 × 103 kg H2/month equivalently. The total yearly energy utilized from Paute Integral is around 2.37% of its generation or 125.6 GWh/year. The daily schedule is 24 h per day. The production costs for both scenarios are depicted in Table 4. For both scenarios, a single electrolysis plant of 17.5 MW has been deployed. The electricity (electrical energy) cost is 0.020 US$/kWh (Paute Integral) as mentioned in Sect. 4.4. Other production costs have been reported in [28, 29]; however, these figures could remotely serve as a comparison because of the significant differences in national policies, import costs, green energy policies, and the existence of a new hydrogen market. Alternatively, if two 35 MW electrolysis plants are considered, total Paute energy usage increases to 5.37%, equaling 284.5 GWh/year and 5591 tons of hydrogen yearly at 96% production. While feasible, pilot plants often precede new technology deployment for secure learning, considering social acceptance. In countries like Ecuador, where hydro conflicts are common, new tech introduction should involve social participation in decision-making. Though not this study’s focus, a multicriterion, multi-actor analysis is recommended in project planning. Once the production costs have been calculated, the distribution costs should be added to obtain a sale price; the price of hydrogen must be competitive when introduced to an energy market. Storage and transport costs can be considered in the product distribution. Hydrogen transport cost is in the range of 2.3–2.6 US$/kg, while the storage cost falls between 0.3 and 0.4 US$/kg [31]. Assuming the international values of 2.3 US$/kg for transport and 0.3 US$/kg for storage, an additional 2.60 US$/kg of distribution is added to the production costs in Table 4. Assuming a profit of 0.40 US$/kg, the sale prices are as shown in Table 5. With respect to hydrogen storage, it is assumed that hydrogen is kept as a gas, i.e., no liquefaction; otherwise, significantly higher costs could compromise the feasibility of the project. Sensitivity analysis indicates key factors influencing hydrogen price: (1) electrolysis plant cost, (2) electricity cost, (3) interest rate, and (4) plant availability. Future’s cheaper renewable energies enhance green hydrogen tech investment and mitigate risks.

Assessment of Green Hydrogen Production from Hydropower in Ecuador Table 6 Total annual cost [US dollars]

Type Operation and maintenance Equipment and materials Storage Transport Business (payroll, legal) Total

Table 7 Economic performance indicators [US dollars]

Type Net present flow Initial investment NPV Payback period (years) The IRR (%) Cost–benefit ratio

Scenario 1 209,630 2,994,720 30,000 2,300,000 21,402 5,825,753

Scenario 1 28,823,023 18,241,300 3,752,317 6.33 8.85% 1.58

113 Scenario 2 215,629 5,799,392 816,234 6,027,785 42,804 12,901,844

Scenario 2 42,575,430 18,241,300 14,246,190 4.28 18.23% 2.33

In the deployment of an electrolysis plant SIEMENS SILYZER 300, there are costs associated with the investment in construction and start-up. Recalling the cost of investment Cinv in Sect. 4, it can be rewritten as follows:

Cinv =

qH 2

kWh 53:53 kg H2

AE

1012:7US kWÞ

ð1þiÞn i ð1þiÞn - 1

ð12Þ

It can be observed that the investment is based on the amount of hydrogen produced qH2 and the number of available hours of the electrolysis plant AE. Cinv has been useful for estimating hydrogen cost; however, construction and start-up require additional finance analysis. Investors need to study the cash flow, and two main parameters to determine the economic feasibility of the project, the Net Present Value (NPV), and the Internal Rate of Return (IRR). IRR is of special interest as it is used to estimate the profitability of potential investments. The total annual cost in both scenarios previously proposed is tabulated in Table 6. Where the interest rate i = 5% and the period n is 10 years. Considering US$ 8.34 and US$ 6.34 as the sale prices for scenarios 1 and 2, respectively, and the total annual cost of Table 6, economic performance with NPV and IRR can be obtained, as shown in Table 7. The payback period and the cost–benefit ratio have been included in Table 7 as key performance indicators in addition to NPV and IRR. The payback period is the number of years required to recover the initial investment, while the cost–benefit ratio indicates the relationship between the relative costs and benefits of the project, expressed in monetary or qualitative terms.

114

5.1

Á. Recalde et al.

National Grid Impact on Production

Green hydrogen’s heavy reliance on hydropower in Ecuador necessitates assessing its impact on economic dispatch, vital for the country’s electricity generation. This analysis helps optimize strategies for sustainable energy production and understand implications on the existing system. Hydrological changes, affecting water availability and energy prices, influence generation costs. Dry seasons lead to water scarcity in both Pacific and Amazon basins, prompting thermoelectric plants to increase production and increase generation costs. Dry seasons typically span from September to March, as shown in Fig. 5. Water inflow variations lead to significant generation cost fluctuations. Estimating costs involves calculating average daily generation expenses on a typical day. Using data from CENACE and the Master Electrification Plan [3], power plants are categorized: hydro, thermal, and other renewables. With total generation and scheduled power known, the average daily generation cost is found by dividing average total daily energy by total income at nominal energy price. This cost insight aids in evaluating the addition of green hydrogen electrolysis plants to the grid, as depicted in Fig. 6. Ecuadorian rivers and the water availability can vary based on multiple factors, including the pace of climate change, local geographic and climatic conditions, and the specific impacts being considered. However, some effects of climate change on Ecuadorian rivers are already being observed, and the impacts are expected to intensify over the coming decades. Projections suggest that the impacts of climate change on Ecuadorian rivers are likely to become more pronounced in the coming decades, particularly as global temperatures continue to rise. Intensified droughts, more intense rainfall events, altered river flow patterns, and changes in water quality could become more evident in the latter half of the twenty-first century [32].

Fig. 5 Total cumulative hydrological profile in Ecuador in the Amazon basin

Assessment of Green Hydrogen Production from Hydropower in Ecuador

115

Fig. 6 Relation between demand and average generation cost

This study’s scope extends until 2040, thus its primary focus does not encompass water availability. However, for analyses extending beyond 2050, it becomes imperative to include considerations for water availability. As time progresses, the impact of changing climatic conditions on water resources will become more pronounced, making it essential to address this aspect in comprehensive assessments. In Fig. 6, Scenarios 1 and 2 are depicted at the two left bars (in the 17.5 MW hydrogen plant capacity installed). Additional bars have been added to represent the hypothetical increment to two plants (35 MW), 3 plants, 6 plants, and 9 plants. The blue lines represent the trend as if there are two cases where the plants are working 8410 h (24 h per day) or 3904 h per year (8 h per day), respectively. It can be observed from Fig. 6 that the average generation cost increment is between 0.03% and 0.14% per electrolysis plant deployed in the power system. Each electrolysis plant has a capacity of 17.5 MW and is assumed to be operational 8 h per day (scenario 1) or 24 h per day (scenario 2), which corresponds to 36% and 96% of availability, respectively. The maximum addition on electrolysis capacity is assumed to be 158 MW (equivalent to nine plants) considering that the cost of generation of other renewables different than hydro could decrease in time. For example, considering a time horizon to 2040, the levelized cost of hydrogen (LCOH) from solar and wind is expected to be lower than the hydro case as depicted in Fig. 7. Figure 7 has considered the cost reduction in renewable energies expected in 2040, i.e., an estimated 41% reduction for wind energy, and 60% for solar energy. For hydraulic energy, the variation cost will experience a very small positive change (increase of 0.11%) during the analyzed period. In addition to the advantages of producing green hydrogen for industrial consumption or export, there is the sustainable side. There are significant contributions in reducing emissions.

116

Á. Recalde et al.

Fig. 7 Levelized cost of hydrogen to 2040

Fig. 8 CO2 emissions reduced for Scenario 1

5.2

Hydrogen Value Chain

Hydrogen is used in refineries such as Esmeraldas, Libertad, and Shushufindi in hydrocracking and hydrotreating processes mainly for reducing the amount of sulfur in oil. The amount of CO2 that is reduced can reach up to 5595 tons H2/year. For instance, emission reduction in percentage for Scenario 1 is depicted in Fig. 8. Another industry where hydrogen could contribute to reducing emissions is the transport sector and thermal generation. Transportation has the largest oil consumption in Ecuador. Progressive replacement of oil technology to electric or hydrogendriven vehicles is possible. In the case of thermal power plants, reductions in emission reach the figures in Fig. 9 if Scenario 2 is used, for example.

Assessment of Green Hydrogen Production from Hydropower in Ecuador

117

Fig. 9 Quantity of CO2 reduction in Scenario 2

Recent exploration of export opportunities highlights China and India as major contributors to hydrogen demand by 2050, China accounting for 25% and India growing significantly. China’s demand is industrial, while India’s steel production fuels its needs. The United States ranks third due to transport sector demand. The top 10 countries might represent two-thirds of global hydrogen consumption in the Paris Agreement’s 1.5 °C scenario. Global hydrogen trade is evident through project announcements and collaborations. Germany, Japan, and the Netherlands are active importers [9].

6 Conclusions In conclusion, the feasibility analysis of green hydrogen production in Ecuador utilizing hydropower resources has demonstrated promising potential. The utilization of Spilled Turbinable Energy (STE) for an electrolysis plant in the Paute River region has been identified as a viable option. The economic evaluation conducted in Sect. 5 indicates positive profitability for a green hydrogen plant. Two scenarios have been proposed, considering the power balance, the generation capacity, the availability of spilled water, and the seasonality of the rainy season. In the first scenario, the plant operates at full capacity during the rainy season and one-third of a day per day for the rest of the year, yielding a production of up to 1000 tons of hydrogen annually. In the second scenario, the plant operates continuously throughout the year, producing 2720.7 tons per year. The return of investment for both scenarios is projected to be achieved in 6.3 years and 4.2 years, respectively, with a profit margin of US $0.58 and US $1.33. Future research will focus on exploring the conditions required for establishing a green hydrogen market. Given that transportation is a key consumer in the market, it can play a significant role in setting local prices. Additionally, the study will

118

Á. Recalde et al.

investigate hydrogen exports, taking into account storage logistics and efficient power generation using stored hydrogen as a raw material. This comprehensive analysis will provide valuable insights into the potential development of a green hydrogen industry and its integration into the energy landscape.

References 1. Fúnez Guerra, C., Reyes-Bozo, L.: El hidrógeno como vector energético (2019) 2. ARCERNNR: Estadística Anual y Multianual del Sector Eléctrico Ecuatoriano 2022. https:// www.controlrecursosyenergia.gob.ec/estadisticas-del-sector-electrico-ecuatoriano-buscar/ (2022) 3. ARCONEL: Plan Maestro de Electrificación 2018-2027 (2018) 4. MICSE: Agenda Nacional de Energía 2016-2040. https://biblioteca.olade.org/opac-tmpl/ Documentos/cg00362.pdf (2016) 5. ARCERNNR: Balance Nacional de Energía Eléctrica 2023. https://www. controlrecursosyenergia.gob.ec/balance-nacional-de-energia-electrica/ (2023) 6. Pelaez-Samaniego, M.R., Riveros-Godoy, G., Torres-Contreras, S., Garcia-Perez, T., Albornoz-Vintimilla, E.: Production and use of electrolytic hydrogen in Ecuador towards a low carbon economy. Energy. 64, 626–631 (2014). https://doi.org/10.1016/j.energy.2013. 11.012 7. Posso, F., Espinoza, J.L., Sánchez, J., Zalamea, J.: Hydrogen from hydropower in Ecuador: Use and impacts in the transport sector. Int. J. Hydrog. Energy. 40(45), 15432–15447 (2015). https:// doi.org/10.1016/j.ijhydene.2015.08.109 8. Zhou, Y., Li, R., Lv, Z., Liu, J., Zhou, H., Xu, C.: Green hydrogen: A promising way to the carbon-free society. Chin. J. Chem. Eng. 43, 2–13 (2022). https://doi.org/10.1016/j.cjche.2022. 02.001 9. IRENA: Global hydrogen trade to meet the 1.5 C climate goal part I: Trade outlook for 2050 and way forward. https://m.h2fcp.org/content/global-hydrogen-trade-meet-15°c-climate-goal-part1-trade-outlook-2050-and-way-forward (2022) 10. Oliveira, A.M., Beswick, R.R., Yan, Y.: A green hydrogen economy for a renewable energy society. Curr. Opin. Chem. Eng. 33, 100701 (2021). https://doi.org/10.1016/j.coche.2021. 100701 11. Seck, G.S., et al.: Hydrogen and the decarbonization of the energy system in europe in 2050: A detailed model-based analysis. Renew. Sust. Energ. Rev. 167, 112779 (2022). https://doi.org/ 10.1016/j.rser.2022.112779 12. IEA: Global hydrogen review 2022 (2022) 13. IEA: Towards Hydrogen definitions based on their emissions intensity. https://iea.blob.core. windows.net/assets/acc7a642-e42b-4972-8893-2f03bf0bfa03/ Towardshydrogendefinitionsbasedontheiremissionsintensity.pdf (2023) 14. IRENA: Green hydrogen: A guide to policy making. https://www.irena.org/-/media/Files/ IRENA/Agency/Publication/2020/Nov/IRENA_Green_hydrogen_policy_2020.pdf?rev=c0 cf115d8c724e4381343cc93e03e9e0 (2020) 15. Shiva Kumar, S., Lim, H.: An overview of water electrolysis technologies for green hydrogen production. Energy Rep. 8, 13793–13813 (2022). https://doi.org/10.1016/j.egyr.2022.10.127 16. Maric, R., Yu, H.: Proton exchange membrane water electrolysis as a promising technology for hydrogen production and energy storage. In: Nanostructures in Energy Generation, Transmission and Storage. IntechOpen (2019) 17. Brauns, J., Turek, T.: Alkaline water electrolysis powered by renewable energy: A review. Processes. 8(2), 248 (2020). https://doi.org/10.3390/pr8020248

Assessment of Green Hydrogen Production from Hydropower in Ecuador

119

18. Kim, J., et al.: Hybrid-solid oxide electrolysis cell: A new strategy for efficient hydrogen production. Nano Energy. 44, 121–126 (2018). https://doi.org/10.1016/j.nanoen.2017.11.074 19. IRENA: Making green hydrogen a cost-competitive climate solution. https://www.irena.org/ News/pressreleases/2020/Dec/Making-Green-Hydrogen-a-Cost-Competitive-Climate-Solution (2020). Last accessed 13 June 2023 20. IRENA: Green hydrogen cost reduction, scaling up electrolysers to meet the 1.5 C climate goal. Abu Dhabi. https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2020/Dec/ IRENA_Green_hydrogen_cost_2020.pdf (2020) 21. Jara Alvear, J.E.: Solar Photovoltaic Potential to Complement Hydropower in Ecuador: A GIS-Based Framework of Analysis. Lund University (2018) 22. Villacreses, G., Martínez-Gómez, J., Jijón, D., Cordovez, M.: Geolocation of photovoltaic farms using Geographic Information Systems (GIS) with Multiple-criteria decision-making (MCDM) methods: Case of the Ecuadorian energy regulation. Energy Rep. 8, 3526–3548 (2022). https://doi.org/10.1016/j.egyr.2022.02.152 23. Singh, A.K., Khatoon, I., Muazzam, M.: An overview of electricity demand forecasting techniques. Netw. Complex Syst. 3(2), 38–48 (2013) 24. Hyndman, R.J.: Forecasting: Principles and Practice, 2nd edn. OTexts (2018) 25. CENACE: Informe Anual 2021. https://www.cenace.gob.ec/informe-anual-2021/ (2022) 26. Siemens: PEM electrolyser technology: Flexible, efficient and scalable. Siemens Energy. https://www.energyforum.in/fileadmin/user_upload/india/media_elements/Presenta tions/20210714_h2_large/Siemens_Energy.pdf (2021) 27. ETAPA: Tarifario del servicio de Agua Potable y Saneamiento Vigente año 2023. https://www. etapa.net.ec/principal/agua-potable/operacion-y-mantenimiento/tarifario-agua-potable-2023 (2023). Last accessed 8 Aug 2023 28. Zamora Garrido, A.: Diseño y dimensionamiento de una planta de obtención de hidrógeno mediante electrólisis de agua de 124MW alimentada con energía solar fotovoltaica en la provincia de Huelva y análisis de viabilidad. Universitat Politecnica de Valencia (2022) 29. Troncoso, F.: Claves del Hidrogeno Verde. Rev. Induambiente. 171, 30–33 (2021) 30. ARCERNNR: Estadística Anual y Multianual del Sector Eléctrico Ecuatoriano 2021. https:// www.controlrecursosyenergia.gob.ec/estadisticas-del-sector-electrico-ecuatoriano-buscar/ (2021) 31. MT, Acuerdo Ministerial No. MDT-2022-234. Ministerio del Trabajo, pp. 1–4 (2022) 32. Armenta Porras, G.E., Villa Cedeño, J.L.: PROYECCIONES CLIMÁTICAS DE PRECIPITACIÓN Y TEMPERATURA PARA ECUADOR, BAJO DISTINTOS ESCENARIOS DE CAMBIO CLIMÁTICO. https://info.undp.org/docs/pdc/Documents/ ECU/14 Proyecciones de Clima Futuro para Ecuador en base a IPCC-AR5.pdf (2016)

Feasibility of Shallow Geothermal Installations for Cooling Purposes in Tropical Climate Mariana Villafán-Sierra, Daniela Blessent, Jacqueline Lopez-Sanchez, Carlos Ernesto Arrieta-Gonzalez, and Mauricio Gonzalez-Palacio

1 Introduction Although geothermal energy is usually associated with volcanism and electrical production from high-enthalpy fluids, shallow or very low enthalpy resources, with temperatures below 30
C, can be carried on with geothermal heat pumps (GHPs) and used for many other purposes that require cooling or heating. Shallow geothermal resources are baseload, renewable, and clean and can be tapped almost everywhere for heating or cooling purposes with GHPs [1–3]. The growing awareness and popularity of GHPs are demonstrated by the continuous increase in their demand for both residential and industrial purposes in recent years [3], as well as their use in 58 countries worldwide in 2021 [4]. The ground is a suitable source/sink of heat, and the installation of heat exchangers connected to a GHP can follow a horizontal or vertical design called GCHP (ground-coupled heat pumps) or GSHP (ground source heat pumps). At shallow depths (0–100 m approximately), the ground can be considered a geothermal reservoir [5] at constant temperature. Although there is daily/seasonal variation in the first few meters depending on the climatic conditions, the temperature is almost constant up to 100 m depth, and then it increases due to the geothermal gradient [6]. Other heat source/sink options are surface water bodies, with surface

M. Villafán-Sierra · D. Blessent (*) · J. Lopez-Sanchez Programa de Ingeniería Ambiental, Universidad de Medellín, Medellín, Colombia e-mail: [email protected] C. E. Arrieta-Gonzalez Programa de Ingeniería en Energía, Universidad de Medellín, Medellín, Colombia M. Gonzalez-Palacio Programa de Ingeniería en Telecomunicaciones, Universidad de Medellín, Medellín, Colombia © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_8

121

122

M. Villafán-Sierra et al.

water source heat pumps (SWSHPs) [3, 7–10], and groundwater heat pumps (GWHP), which use water pumped out of aquifers [11]. During the summer, GHPs use the ground as a heat sink, while in the winter, they use it as a heat source. They offer higher energy efficiency for air conditioners compared to traditional air source heat pumps [2, 12]. The implementation of GHPs has demonstrated cost savings for heating, ventilation, and air-conditioning (HVAC) applications and adaptation to the required needs (cooling, heating, or both) and to the size of the space. The five leading countries for GHPs in terms of both installed capacity and annual energy use are China, USA, Sweden, Germany, and Finland [4]. In Latin America, interest in geothermal resources is increasing, due to the established sustainability goals [13]. Geothermal installations are becoming a valuable resource in the region [14], since their use can help reduce greenhouse emissions, mitigate climate change, improve energy security, and diversify the energy market. Therefore, it is important to encourage their development, as Latin America and the Caribbean are very favorable regions for the development of this industry. In addition, it is essential to analyze the performance of GHPs in the tropics, as most applications are reversible at higher latitudes, providing both heating and cooling depending on the season. In the Chilean project “Invernadero geotérmico para la reinserción social en Aysén,” a horizontal GCHP system was installed to heat at 18
C in a 420 m2 greenhouse during the winter, using a GHP of 22 kW [15]. No use of GHPs is reported in Argentina before 2020 by Lund and Toth [4]. A recent study investigated the air conditioning of a single-family house in the province of Córdoba, using a ground-air heat exchanger system in a ring configuration, due to the use of a single duct, at a depth of 2–3 m [16]. Similarly, in Perú, [17] designed the geothermal probes and selected the heat pump for a geothermal installation for the supply of heating, cooling, and domestic hot water in the district of la Molina in Lima. In Uruguay, the first geothermal energy system in the country was installed and continues to be used in Los Pinos Educational Centre, located in the Casavalle neighborhood; here, a 30 m deep well extracts groundwater with a submersible pump and sends it to the GWHP, where the heat transfer occurs before the water is reinjected into the aquifer with a second well [18]. Moreover, a call for a low-enthalpy geothermal energy project was launched in the country at the end of 2021 [19]. In Colombia, a pilot project was operating for a cold room with a volume of 90 m3 and a temperature of approximately 2
C and a temperature of at the Gran Sabana Industrial Park located in Tocancipá, close to Bogotá, using a vertical GCHP installation with three boreholes drilled until an average depth of 80 m [20]. Nevertheless, there have been no further GHP installations in the country. Therefore, this study aims to evaluate the possibility of using a geothermal installation to cool a preservation room at the flower farm in La Ceja municipality. This work provides insights into low geothermal installations in Colombia, as well as in any other region with similar climatic conditions and where GHPs are still not commonly installed. This chapter is structured as follows: after describing the materials and methods employed (site description, temperature monitoring, cooling load calculation, and comparison of traditional HVAC, SWSHP, and GCHP cooling systems), the Results

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

123

and Discussion sections present the main insights of this work, and finally, in the Conclusions, the whole study is summarized.

2 Materials and Methods 2.1

Site Characterization

La Virginia S.A.S flower farm is located in the La Ceja municipality, Antioquia, Colombia (Fig. 1a), and is part of the Falcon Farms holding company, which operates in Colombia, Ecuador, and United States. La Ceja is located at 2200 m altitude, and its mean annual temperature and precipitation are 18
C and 2099 mm, respectively. According to Poveda [21], the yearly cycle of hydro-climatology in Colombia consists of two rainy seasons (April–May and October–November) and two dry seasons (December–February and July–August). This bimodal cycle can be altered by the influence of El Niño/Southern Oscillation (ENSO) or La Niña phenomena [21, 22]. The flower crop operates under a long-term sustainable agricultural model and is certified by the Rainforest Alliance [23], which emphasizes a commitment to continuous improvement, sustainability training, and clear benefits for [24]. The farm has 214,000 m2 of greenhouse area, and flowers are exported and distributed in the United States [25]. Four rainwater harvesting, artificial ponds are used for irrigation purposes and hereafter called simply “reservoirs” (Fig. 1a). Reservoirs 1 and 4 (Fig. 1b) are the closest ones to the cold room (between 200 and 300 m away), so they were selected for temperature monitoring, to determine their potential as heat sinks. Since they are artificial surface water bodies, they are not a relevant ecological system, reducing the environmental impacts potentially associated with their use. The production process of exportation flowers consists of several stages, such as germination, transplanting to the greenhouse, growing period, harvesting, packaging, preservation, distribution, and marketing. This study focuses only on the preservation stage, as this is when the highest energy consumption occurs, requiring about 30% of the total energy consumption of the flower farm (currently about $2000 USD per month). The preservation of flowers at 3
C for approximately 3 days is essential to ensure their freshness for the subsequent distribution and exportation. The proposed energy alternative can contribute to reduce production costs and improve environmental standards. Also, the farm could become a sustainable environmental benchmark, serving as a model for other agricultural activities in the region.

124

M. Villafán-Sierra et al.

Fig. 1 (a) Location map of La Virginia S.A.S flower farm (b) Zoom map of Reservoirs 1 and 4

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

2.2

125

Remote Sensing and Temperature monitoring

The temperature-sensing device built is equipped with two K-type thermocouples with a resolution of 0.5
C, a 120 W solar panel, a 12 V DC battery, a WLAN router, and a printed circuit board that captures and stores temperature data for a given period. The device (Fig. 2) is capable of capturing data every 15 min (for each of the two thermocouples) and sending them to a cloud platform, where it is possible to track and manage the data in real time. One thermocouple recorded the ambient temperature, while the other the water temperature. The device was installed in Reservoir 1 on March 2021 (Fig. 3), at 2.2 m from the surface considering that the reservoir bathymetry indicates that the maximum depth is 2.8 m. It is important to mention that a deeper monitoring point might lead to future problems if a heat exchanger (HE) is installed in a later stage of the project, since it cannot be installed at the bottom of the reservoir because of the risk that the plants and sludge would affect the operation. The tool collected integrated and representative temperature data for 5 months to analyze the temperature behavior in both seasons, March–April–May representing the rainy season, and June–July for the dry season.

Fig. 2 Measure and transmission device: (1) 12 V DC battery, (2) solar panel controller, (3) port to connect the sensors, (4) microcontroller, (5) WLAN router, and (6) cellular modem

126

M. Villafán-Sierra et al.

Fig. 3 Monitoring temperature tool with solar panel installed at the one bank of rainwater harvesting Reservoir 1, with sensor submerged 2 m deep

2.3

Cooling Load Calculation

The cooling load calculation estimates the flows and heat sources, which may vary during the day due to external loads, such as cold air leaks, heat from fans, and air renewals [26]. The primary internal heat generation sources are transmission load, products, internal equipment, and infiltration load (Qi). The cold room analyzed in this study consists of two spaces: the first of these has an area of 160 m2 and height of 3.0 m, while the second has an area of 234 m2 and height of 3.6 m (Fig. 4). The roof and walls are isolated with polyurethane of 80 mm thickness, and the floor is constructed using a 20 cm width heavy concrete block. The cooling loads are calculated using simplified equations based on ASHRAE [27]. The calculation steps to obtain the heat gain (Q Total) that needs to be removed by the refrigeration system are shown in the following equations. The conduction load Q1 (kWh/day) consists of the heat transfer through the building walls, roof, doors, and floor etc [26]. It is calculated as indicated by Eq. 1 for each surface and then summed up: Q1 ¼ U  A  ðT out  T in Þ  24=1000

ð1Þ

where U is the value of insulation (W/m2K), A is the surface area (m2), Tout is the ambient temperature (K), Tin is the internal temperature in Kelvin degrees (K), 24 is the conversion from hours to days, and 1000 is the conversion from W to kWh. It is important to note that the cold room is divided into two spaces separated by a folding door (Fig. 4) as it adapts to the storage demand of the flower crop. In this case, a storage capacity of 100% is assumed. Internal cooling loads Q2 (kWh/day) include

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

127

Fig. 4 Cold room sketch

the heat transfer due to the product, workers, equipment, and lighting [26]. Equation 2 calculates the heat produced by the flowers when they enter the room:

Q2 ¼ m  C p  T enter  T storage  1=3600

ð2Þ

where m is the mass of the stored product (kg), Cp is the specific heat capacity (kJ/
Ckg), Tenter is the entering temperature in Celsius degrees, Tstorage is the storage temperature in degrees Celsius, and 3600 is the conversion from kJ to kWh. Equation 3 calculates the heat gains due to the respiration process of cut flowers Q3 (kWh/day): Q3 ¼ m  Resp  1=3600

ð3Þ

where m is the mass of the stored product (kg/day), Resp is the Respiration heat (kJ/kg), and 3600 is the conversion from kJ to kWh. Equation 4 estimates the heat transfer of the workers inside the room Q4 (kWh/day):

128

M. Villafán-Sierra et al.

Q4 ¼ People  Time  Heat  1=1000

ð4Þ

where People is the number of persons working inside the cold room every day, Time refers to the working hours (h/day), heat is the average heat loss of a person (W), and 1000 is the conversion from W to kWh. Equation 5 determines the heat contributed by the lamps Q5 (kWh/day): Q5 ¼ Lamps  Time  Wattage  1=1000

ð5Þ

where Lamps are the number of lamps, Time is the hours of use per day (h/day), Wattage is the energy consumed by each light (W), and 1000 is the conversion from W to kWh. Equation 6 calculates the heat released by the fans inside the cold room Q6 (kWh/day): Q6 ¼ Fans  Time  Wattage  1=1000

ð6Þ

where Fans is the number of fans, Time refers to the hours of use per day (h/day), Wattage is the energy consumed by each fan (W), and 1000 is the conversion from W to kWh. Equation 7 gives the result of the infiltration air load Q7 (kWh/day): Q7 ¼ R  V  EðT out  T in Þ1=3600

ð7Þ

where R is the cold store volume air changes per day, V refers to the cold room volume (m3), E is the energy per cubic meter of air (kJ/m3
C), Tout is the ambient temperature in Celsius degrees, Tin is the internal temperature in degrees Celsius, and 3600 is the conversion from kJ to kWh. Equation 8 is the total cooling load QTotal (kWh/day), which is the sum of all the previous operation: QTotal ¼

Qi

With i ¼ 1, 2, . . . , 7

ð8Þ

Finally, the result above is multiplied by a Safety Factor (SF), generally adding 10–30% onto the previous result, as follows (Eq. 9): QTotal ¼

Qi  SF

ð9Þ

The data considered for the cooling load estimation are summarized in Table 1, with a SF ¼ 1.2. The storage cold room capacity value (25,000 kg) is provided by the farm, while thermal properties of flowers are assumed from typical known ranges. In addition, an alternative cooling load calculation was performed using the software Danfoss coolselector2 [28]. This software provides experimental air-conditioner performance curves to calculate the required cooling capacity for a

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

129

Table 1 Values used for the calculation of the cooling load Parameter Roof and walls heat transfer coefficient U (W/m2K) Floor heat transfer coefficient U (W/m2K) Total mass of stored flowers (kg) Flowers entered per day (kg/day) Flower specific heat capacity (kJ/kg
C) Flower respiration heat (kJ/kgday) People working inside the cold room per day Total number of lamps Total number of fans Cold room internal temperature (
C)

Value 0.28 0.75 25000 7200 3.85 2.092 3 15 12 3

cold room. Here, the results obtained using the simplified equations proposed by [27] are compared with the result obtained using the experimental database provided by the software Danfoss coolselector2. The cooling load was also estimated to size the SWSHP system’s capacity and determine the practical heat load Heat LoadCL (kWh/month) (Eq. 10), so that it could be compared to the heat load found from the efficiency (Eqs. 11 and 12). It is worth mentioning that according to the information given by the farm, the current heat load per month is 18,000 kWh, so the values calculated before were compared to the last value (assuming that the latest is the actual heat load value): Heat LoadCL ¼ Cooling load ðKWh=dayÞ  ð30 dayÞ=ð1 monthÞ

ð10Þ

where Heat LoadCL refers to the theoretical heat load found from the cooling load. Cooling load is the amount of energy that removes heat and maintains the cold room at the required temperature.

2.4

Traditional HVAC vs SWSHP System

When assessing the efficiencies of heating and cooling systems, the Energy Efficiency Ratio (EER) stands out as a pivotal metric. By harnessing the EER, we can establish a standardized foundation for contrasting the performances of various systems. Both the Conventional HVAC and the SWSHP systems focus on transferring heat from one location to another. The distinction lies in their heat management methodologies. The Conventional HVAC concentrates the heat from an environment by compressing a refrigerant. This accumulated heat is then released to the surrounding atmosphere via a condenser. In contrast, while the SWSHP also extracts heat, it diverges in its dispersion method. Instead of expelling heat into the air, the SWSHP relocates this energy to water bodies like seas, rivers, or lakes. The use of natural

130

M. Villafán-Sierra et al.

water reservoirs means the system can leverage a heat sink with exceptional thermal capacity and inertia. A notable limitation, however, is the SWSHP system’s augmented water pumping demands. For a precise comparison of the conventional and the proposed geothermal refrigeration methods, we employed empirical formulations rooted in the theoretical performance delineated by Carnot. Equation 11 represents the highest efficiency for a thermodynamic cycle [7, 28]: Emax ¼ θ1 =ðθ1  θ2 Þ

ð11Þ

Here, Emax denotes the apex efficiency achievable, with θ1 and θ2 being the required temperature in Kelvin and the source temperature in Kelvin, respectively. From Eq. 11, it can be appreciated that the highest efficiency can be calculated depending on the temperature of the heat sink and source, using an idealized Carnot cycle [29]. However, in practice, the Emax or EER (Energy Efficiency Ratio) will be much lower than the ideal [7]. For this reason, as shown by [30], the actual efficiency is assumed to be 30% of the ideal efficiency: E max

real

¼ Heat loadE ¼/ E max

ð12Þ

where Emax real or Heat loadE is the heat load found from the efficiency, α is a performance constant, which is 0.30, and Emax is the best performance of a machine proposed by Carnot.

2.5

Ground-Coupled Geothermal Horizontal System

The design of a horizontal ground-coupled system typically requires installation at depths of 2–3 m [5]. For this configuration, knowledge of the soil temperature is essential. To determine this, measurements were taken at a 2-m depth continuously from September 2021 to March 2022, covering both the rainy and dry seasons. The recorded data consistently showed a soil temperature of approximately 19
C. To develop a preliminary design of the horizontal ground-coupled system, we utilized an Excel-based tool provided by [31]. This tool incorporates the methodology proposed by IGSHPA [32]. The total length L of the ground-coupled heat exchanger in a closed-loop horizontal configuration was calculated using the formula by [32]: L ¼ Qcond  Rt = ðELT þ LLTÞ=2  T g

ð13Þ

where Qcond (in W) denotes the heat pump’s condenser power in cooling mode, which depends on its EER and the required cooling load QTotal, Rt (m∙
C/W) is the total thermal resistance of the pipes and the ground. ELT and LLT (in
C) are the

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

131

Entering Liquid Temperature and Leaving Liquid Temperature, respectively, within the heat pump system. Tg (in
C) indicates the measured soil temperature at the chosen installation depth. Further details and associated calculations can be referenced in [32].

3 Results Figure 5 illustrates the graph derived from the analysis of 20,000 temperature data points, collected by the remote sensing device for both water and ambient air across rainy and dry seasons. The variations in water temperature are found to be less pronounced compared to those in air temperature. During the rainy season, the water temperature consistently remains slightly below the air temperature from 9:00 to 13:00. This phenomenon may play a crucial role in determining the efficiency of the SWSHP. As the SWSHP relies on heat exchange with the water, a cooler water temperature can provide more efficient heat transfer. This behavior results in a higher EER during this period as compared to the traditional systems. This is evident from the efficiency data, where the SWSHP’s EER is highest between 9:00 and 13:00, as shown in Fig. 6. However, preceding the 9:00–13:00 interval, the water temperature is found to be approximately 5
C warmer than the ambient air temperature. It reaches its peak value after 13:00 and then begins to cool down but remains consistently higher than the air temperature throughout the day. The warmer water temperatures before and after the 9:00–13:00 interval can lead to reduced heat transfer efficiency for the SWSHP, causing it to have a lower EER during these times.

Fig. 5 Water and air temperature changes during rainy and dry seasons in Reservoir 1

132

M. Villafán-Sierra et al.

Fig. 6 EER behavior for different refrigeration systems

In contrast, during the dry season, the water temperature consistently shows a reading 5
C higher than the air temperature until 8:00. Subsequently, from 8:00 to 17:00, the water temperature is observed to be cooler than the air temperature. Such temperature variations influence the efficiency of the SWSHP, as the system struggles to maintain optimal heat transfer when the water is warmer than the air. The efficiency data reveal that the EER of the SWSHP is indeed lower when the water temperature is warmer than the air. A significant drop in air temperature to 14
C is recorded after 17:00, whereas the water temperature remains steady, fluctuating between 18 and 20
C. In such cases, the traditionally system can potentially operate more efficiently. The theoretical cooling load calculated for the preservation room was 486 (kWh/day), whereas the software prediction was 543 (kWh/day). Despite the discrepancy, both values lie in the same order of magnitude, underscoring the reliability of the measurements and calculations. It is possible to conclude then that there is a clear correlation between the efficiencies of the SWSHP and the temperature differences between the air and water. The optimal operational period for the SWSHP, in terms of efficiency, is between 9:00 and 13:00, during which the water temperature is slightly cooler than the air. It is essential to factor in these temperature variations when considering the feasibility and efficiency of geothermal systems in specific environmental conditions. The current cooling system, having been in operation for several years, understandably does not exhibit the same performance and efficiency it once did upon initial installation. As such, it provides the foundational baseline for our comparative

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

133

Fig. 7 Average soil and air ambient temperature monitoring for 7 months

study. When examining the efficiencies of the three cooling systems, the SWSHP emerges as an especially strong contender during the mid-morning to early afternoon window. Between 09:00 and 13:00, even with its variable performance, the SWSHP consistently demonstrates a commendable EER. This superiority in efficiency, when extrapolated over extended usage, signifies potential substantial energy savings and consequent reductions in operational costs. Contrarily, the existing traditional refrigeration system, bearing the marks of age and continual wear, lags significantly in its efficiency. It falls behind both the SWSHP and a fresh traditional system, often by margins exceeding 50%. Such a pronounced disparity accentuates the criticality of timely system maintenance and occasional upgrades. Considering these observations, stakeholders could lean toward adopting SWSHP systems, not solely for their immediate efficiency benefits but also for their promising adaptability to fluctuating environmental temperatures. Nevertheless, it is crucial to acknowledge that, in this context, a brand-new traditional system did surpass the SWSHP in terms of efficiency. Although the complete design of the installation is beyond the scope of this work, the soil temperature was recorded from September 2021 to March 2022, and the data collected are shown in Fig. 7. The results of the monitoring show that the soil maintains a nearly constant temperature during the day without being influenced by air temperature changes. The slinky heat exchanger proposed here requires approximately 4150 m length calculated with Eq. 13, and with a distribution arrangement characterized by three or six trenches, covering an area of 2660 m2 or 3040 m2, respectively.

134

M. Villafán-Sierra et al.

4 Discussion According to [7], a minimum depth of 3–4 m for SWSHP systems is required. However, this arrangement has been used in water bodies with shallower depths, with mixed results [31]. Hence, it is possible to exceed these recommendations if the detailed analysis is performed. Reservoir 1 is not as deep as [7] recommended. A large surface is another essential characteristic of SWSHP. This means that the water body has a long-term heat replenishment and an adequate heat exchange area, guaranteeing the system’s efficiency [7]. However, the water drop during a severe dry season could affect the system’s efficiency and complicate the installation of a SWSHP.

4.1

Temperature monitoring

The water temperature registered in the rainwater reservoir is higher than expected (Fig. 5), which could reduce the performance of the geothermal system. In addition, water temperature varied with outdoor conditions due to its shallowness. The temperature difference between that of the reservoir and that required by the cold room (3
C) is significant (ΔT ¼ 15
C), indicating that a lot of work is required on the part of the GHP to supply the cooling demand, generating high-energy consumption and operation costs. During cooling mode, the dumping of heat raises the reservoir’s temperature. Banks [7] stated that this frequently happens in many shallow water bodies and non-stratified lakes, such as Reservoir 1, where the main natural modes of heat loss are evaporative and back-radiation losses. If a considerable heat load is imposed on a shallow water body, the increased evaporation rate may result in an undesirable decline in water level and an elevated temperature [7, 33], which will affect the efficiency of the SWSHP system. This variation may affect the reservoirs’ hydrological dynamics and thermal equilibrium, which may or may not lead to changes in its ecological system, even if it is an artificial water reservoir and does not hold strategic ecosystems, like in this case. Therefore, it is recommended to assess the heat budget and thermal balance of the water reservoir, which was beyond the research aims of this work. To obtain the total cooling with a SWSHP system, a water temperature of 10
C or below is necessary [34]. However, the water temperature in Reservoir 1 remains at approximately 19
C for both seasons, meaning that it is not feasible to use a single SWSHP system to refrigerate the cold room at 3
C. Nevertheless, other options can be considered, such as horizontal/vertical GCHP or the SWSHP as air pretreatment, before cooling by the conventional method. All these options could reduce the costs associated with flower production, and the initial investment could be recovered in the medium term, although a detailed economic analysis is required and will be the object of further studies.

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

4.2

135

Cooling Load and Refrigeration Options

The temperature difference between that of the source and that required in the cold room is directly related to the amount of energy consumed by the HP: the greater the ΔT, the greater the energy demand by the GHP. Banks [7] and IGSHPA [32] suggested that a shallow water body with a peak cooling load under 17.4 W/m2 or in constant replenishment could be considered acceptable as a cooling heat sink. However, in this investigation, the cooling load per area is 57 W/m2, meaning that the cooling space is too big to be cooled by the surface water body. The SWSHP is a very interesting and viable proposition compared to the traditional refrigeration system, since it facilitates a reduction in the operating costs of the system, considering that the GHP only uses 30% of the electrical energy used by a conventional refrigeration system. Nevertheless, a new traditional refrigeration system is still a better option than a SWSHP. However, the useful life of the systems is also a relevant factor. A conventional cooling system can operate for 20 years, while a geothermal system can operate for 30–35 years [20]. Considering all the aspects evaluated above, such as water temperature, reservoir dimensions, and climatic conditions, a SWSHP is not an attractive alternative since many of the values suggested in the literature concerning reservoir depth and temperature do not apply in this case, which is why other options should be considered.

4.3

Ground-Coupled Geothermal Horizontal System

A horizontal GCHP system installed at a depth of 2.0 m could be a suitable alternative. It is preferred to vertical borehole heat exchangers as it helps avoid high drilling costs, although a sufficient area is required for its installation. The horizontal slinky heat exchanger requires a surface of approximately 3000 m2, as mentioned previously. This is considered acceptable and feasible, considering that the installation is for a farm with large land availability. However, the excavation of the first 2–3 m of soil should be conducted in an area where flowers are not planted. As a first approach to the financial feasibility study of the project, a SWOT matrix (Table 2) is proposed to identify the Strengths, Weaknesses, Opportunities, and Threats (SWOT) that characterize the potential installation of the proposed geothermal system.

136

M. Villafán-Sierra et al.

Table 2 SWOT matrix for the proposed GHP installation S Strengths

W Weaknesses

O Opportunities

T Threats

GHP consumes 30% of electrical energy in comparison to a traditional refrigeration system GHP is a renewable, clean, and cost-effective energy alternative A single SWSHP is not enough The heat sink is hotter and shallower than expected High initial investment cost This pilot project can demonstrate the potential of geothermal energy in the sector It would be the first geothermal project developed in the Antioquia region and the second in Colombia The production cost will be reduced Poor knowledge about GHP operation in the tropics High initial investment cost Currently uncompetitive in the market

5 Conclusions The possibility of implementing a SWSHP system to cool at 3
C a flower preservation room in La Virginia S.A.S in La Ceja, Antioquia (Colombia), was evaluated, considering the depth and size of the different water reservoirs available in the farm. Rainwater harvesting pond identified as Reservoir 1 was considered the most suitable heat sink, since it was the deepest and closest to the cold room. However, during the severe dry season, its water level can drop, complicating the operation of the SWSHP. Subsequently, the water temperature was monitored for 5 months (including rainy and dry seasons) with a sensing device programmed to record temperature data every 15 min with two thermocouples: the first measuring the water temperature at 2 m depth, and the second capturing ambient air temperature. The data were collected on an online platform. In addition, the cooling loads were calculated using a simplified methodology taken from the ASHRAE Association [27], and a comparison between traditional refrigeration systems and the SWSHP was presented using an empirical equation based on the maximum efficiency proposed by Carnot. It was found that Reservoir 1 is shallower and with higher water temperature than that recommended by the literature for SWSHP application. It also showed a temperature variation similar to the ambient temperature, although in a smaller range. Regarding the EER comparison between both refrigeration systems, a SWSHP is better than the current traditional refrigeration system; however, a new conventional refrigeration system can be more efficient than a SWSHP. Nonetheless, it is important to consider that the useful life of a GHP is longer than that of a traditional refrigeration system. Therefore, considering all the above aspects, it is not feasible to use a single SWSHP system to refrigerate the cold room. It is important to know how the heat discharged to the reservoir water can affect the ecology of the surface water body or its utility value to other users. This analysis was beyond the goals of this study, but a heat balance of the lake is recommended to perform a detailed risk assessment of the impact of a cooling scheme on a lake. A horizontal GCHP system

Feasibility of Shallow Geothermal Installations for Cooling Purposes. . .

137

was also proposed as an alternative geothermal installation. Soil temperature was monitored for a month, and the area required by the heat exchangers was estimated. Further investigation should focus on monitoring soil temperature in other areas of the country and on the measurement of soil thermal properties for different lithologies to generate a database available for further studies. This work helps foster the analysis of the viability of shallow geothermal systems in tropical climate in Latin America, where there is growing interest in direct applications of geothermal energy [13], but few studies have been published.

References 1. Naili, N., Hazami, M., Attar, I., Farhat, A.: Assessment of surface geothermal energy for air conditioning in northern Tunisia: Direct test and deployment of ground source heat pump system. Energy Build. 111, 207–217 (2016) 2. Rivas-Cruz, F., Onofre-Hilario, F., Fuentes-Torres, R., Torres-Luna, V., Gonzales-Reyes, I.: Ground heat exchanger software: State of art. XXV Congreso Anual de La Asociación Geotérmica Mexicana, pp. 18–20 (2018) 3. Wang, G., Wang, W., Luo, J., Zhang, Y.: Assessment of three types of shallow geothermal resources and ground-source heat-pump applications in provincial capitals in the Yangtze River Basin, China. Renew. Sust. Energ. Rev. 111, 392–421 (2019) 4. Lund, J.W., Toth, A.N.: Direct utilization of geothermal energy 2020 worldwide review. Proceeding World Geothermal Congress 2020+1, pp. 1–39 (2021) 5. Perovic, P., Trucco, C., Tálamo, A., Quiroga, V., Ramallo, D., Lacci, A., Baungardner, A., Mohr, F.: Guía técnica de diseño de bomba de calor geotérmica (Fondo editorial del IDAE. Ahorro y Eficiencia Energética en Climatización (2010) 6. US department of Energy: How Geothermal Heat Pumps Work. Energy Efficiency and Renewable Energy (2011). https://energy.gov/sites/prod/files/guide_to_geothermal_heat_pumps.pdf 7. Banks, D.: An Introduction to Thermogeology: Ground Source Heating and Cooling, vol. 2, pp. 1–539. Wiley-Blackwell (2012) 8. Mitchell, M.S., Spitler, J.D.: Open-loop direct surface water cooling and surface water heat pump systems – A review. HVAC&R Res. 19(2), 125–140 (2013) 9. CIBSE Heat Pump Association – HPA and GSHP Association: Surface Water Source Heat Pump: Code of Practice (2015) 10. Luo, J., Luo, Z., Xie, J., Xia, D., Huang, W., Shao, H., Xiang, W., Rohn, J.: Investigation of shallow geothermal potentials for different types of ground source heat pump systems (GSHP) of Wuhan city in China. Renew. Energy. 118, 230–244 (2017) 11. Rafferty, K: An information survival kit for the prospective geothermal heat pump owner. Geo-Heat Center (2001, March) 12. Yang, H., Cui, P., Fang, Z.: Vertical-borehole ground-coupled heat pumps: A review of models and systems. Appl. Energy. 87(1), 16–27 (2010) 13. Gischler, C., Perks, M., González, C., Correa, C., Aragón, R., Haratsu, M., García Fernandez, J., Siroit, G.: Harnessing geothermal potential in Latin America and The Caribbean: A perspective on the road ahead. Inter-American Development Bank (IDB) (2020) 14. Ministerio Federal de Cooperación Económica y Desarrollo-BMZ: Fomento de la geotermia en Centroamérica. Deutsche Gesellschaft Für Internationale Zusammenarbeit (GIZ) GmbH (2021) 15. CEGA – Centro de Excelencia en Geotermia de los Andes, Homepage (2017). http://www. cega-uchile.cl/investigacion/#lineas-de-investigacion. Accessed 12 July 2023 16. Rodríguez Leguizamon, A.G.: Energía Geotérmica Para Viviendas En Córdoba. Universidad Nacional de Córdoba, Trabajo Final de Especialización en Tecnología Arquitectónica (2021)

138

M. Villafán-Sierra et al.

17. Corbacho Morales, J.A.: Diseño y climatización de una casa residencial mediante una bomba de calor geotérmica de muy baja temperatura en la ciudad de Lima-La Molina. Universidad Nacional Mayor de San Marcos (2018) 18. Radiomundo En Perspectiva Homepage.: https://enperspectiva.uy/en-perspectiva-programa/ entrevistas/centro-educativo-de-casavalle-realiza-la-primera-experiencia-con-energiageotermica-en-uruguay/. Accessed 12 July 2023 19. Ministerio de Ambiente Homepage.: https://www.gub.uy/ministerio-ambiente/comunicacion/ convocatorias/convocatoria-proyectos-para-desarrollo-proyecto-piloto-geotermia-baja. Accessed 12 July 2023 20. Ortiz, A: Avances y experiencias prácticas en el diseño y construcción de sistemas geotérmicos de baja entalpía. Comité Regional de La CIER Para Centroamérica y El Caribe (2020). https:// www.youtube.com/watch?v¼P5mcg1XfwmE 21. Poveda, G.: La hidroclimatología de Colombia: Una síntesis desde la escala inter-decadal hasta la escala diurna. Acad. Colomb. Cienc. 28(107), 201–222 (2004) 22. Ramírez-Builes, V.H., Jaramillo-Robledo, Á.: Relación entre el índice oceánico de El Niño y la lluvia, en la región Andina central de Colombia. Cenicafé 60(2), 172 (2009). https://biblioteca. cenicafe.org/bitstream/10778/228/1/arc060%2802%29161-172.pdf 23. Rainforest Alliance Homepage.: https://www.rainforest-alliance.org/es/perspectivas/quesignifica-rainforest-alliance-certified/. Accessed 12 July 2023 24. Falcon Farms Homepage.: https://falconfarmsonline.com/. Accessed 12 July 2023 25. Falcon Farms Homepage.: https://falconfarmsonline.com/la-virginia-farm/. Accessed 12 July 2023 26. Bhatia, A.: Cooling load calculations and principles. Energy Policy, 4081–4092 (2006) 27. ASHRAE: American Society of Heating, Refrigerating and Air-Conditioning Engineers: Manual-Fundamentals Guidelines (2017) 28. Danfoss Homepage.: https://www.danfoss.com/en-us/service-and-support/downloads/dcs/ coolselector-2/. Accessed 12 July 2023 29. Shang, R., Zhang, Y., Shi, W., Wang, X., Zhang, Y.: Fresh look and understanding on Carnot cycle. Energy Procedia. 61, 2898–2901 (2014) 30. Aravena, D., Garcia, K., Muñoz, M., Maripangui, R.: Propiedades termales de sedimentos en Coyhaique. Centro de Excelencia En Geotermia de Los Andes (CEGA). Departamento de Geología, Universidad de Chile (Santiago) (2015) 31. Gutierrez-Soleibe, F.: Propuesta de instalación geotérmica somera a partir de bombas de calor para la climatización de espacios en la Universidad EIA. Universidad Escuela de Ingeniería de Antioquia (EIA) (2021) 32. IGSHPA: Ground Source Heat Pump Residential and Light Commercial. Design and Instalation Guide. Oklahoma State University (2009) 33. Kavanaugh, S: Pond loops, lake loops, or surface water heat pumps. Outside the loop a newsletter for geothermal heat pump designers and installers, pp. 1–8 (1999) 34. Kavanaugh, S., Rafferty, K.: Geothermal heating and cooling: Design of ground-source heat pump systems. ASHRAE. ISBN: 9781936504855 (2014)

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines in Developing Countries with a Focus on Colombia’s Context Brandon Martínez , Carlos Arrieta , Ainhoa Rubio , Mario Luna Hernando Yepes , Edwin Chica , Laura Velásquez , and Juan Pablo Gómez Montoya

,

1 Introduction The growing recognition of the damage caused by human activities on the environment, including the increase in CO2 emissions, global warming, and pollution, has driven the development of clean and sustainable renewable energy technologies. Currently, renewable energies make up nearly 25% of the world’s energy demand, with 6321 TWh of electricity generated [1]. As society becomes more focused on preserving the environment, the role of renewable resources in electricity generation becomes more important. These technologies, in addition to ensuring environmental preservation, promote socioeconomic growth [2]. One popular option is hydropower, which captures the kinetic energy of water currents to produce electrical energy. In Colombia, 70% of electricity comes from hydroelectric sources, due to the country’s abundant water resources [3]. However, hydropower has drawbacks, such B. Martínez · C. Arrieta (✉) · M. Luna Facultad de Ingeniería, Grupo de Investigación en Ingeniería en Energía, Universidad de Medellín, Medellín, Colombia e-mail: [email protected] A. Rubio · H. Yepes Grupo de Investigación en Nuevas Tecnologías, Sostenibilidad e Innovación (GINSTI), Departamento de Ingeniería Mecánica, Universidad Francisco de Paula Santander, Ocaña, Colombia E. Chica · L. Velásquez Grupo de Investigación Energía Alternativa, Facultad de Ingeniería, Universidad de Antioquia, Medellín, Colombia J. P. Gómez Montoya Docente de tiempo completo en Investigación, Universidad Tecnológica del Perú, Lima, Peru © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_9

139

140

B. Martínez et al.

as flooding of agricultural lands, disruption of natural habitats, sedimentation problems, changes in water availability, and methane gas emissions [4]. To mitigate these issues, the use of hydrokinetic turbines (HKTs) has been proposed for power generation in Colombia. Preferred for their easy-to-use designs, affordability, and accessibility, these turbines are often touted as having minimal environmental impact—as cited by multiple studies. [1, 5]. Using HKTs together with wind turbines has been seen as a potential solution. A horizontal-axis HKT captures energy from water flow and turns it into rotational energy to produce electricity [6]. As noted by Gaurav Saini & R.P. Saini [7], a turbine’s efficiency is closely linked to the behavior and conditions of the fluid. In 2015, Argentina introduced a plan to replace diesel engines with hydrokinetic energy. This initiative aims to increase understanding of the technology and incorporate RE into the local power grid [8]. The projected capacity of this system exceeds 1 kW, potentially powering 20,000 urban homes [9]. Furthermore, HKTs offer a new way to harness RE from both freshwater and marine sources [10]. They are less weather-dependent than wind or solar energy sources [11], marking them as a reliable renewable energy option. The benefits of HKTs for electricity generation are widely recognized, but it is also known that these technologies can have adverse environmental, social, and economic impacts. These impacts can manifest themselves through changes in water flow, local sediment movement, and stress on aquatic species, among others [11]. Despite these concerns, available research on the specific impact of these technologies is limited. To fill this knowledge gap, it is necessary to conduct Environmental Impact Assessments (EIAs) on hydrokinetic energy projects. EIA is an element of the “rational model” of planning and decision-making to predict impacts and evaluate alternatives [12].

2 Guidelines for EIA in Colombia EIA is a vital process for evaluating the potential adverse effects of projects on the environment and implementing strategies to alleviate them. Several authors, including [6, 13, 14], underscore the significance of EIA guidelines. These guidelines ensure an in-depth and meticulous assessment, both physically and socially. They equip project proponents with both technical and procedural insights for a comprehensive EIA. However, when addressing development projects in a specific sector, it’s essential to have precise EIA guidelines that align with both international and local legal stipulations. Without dedicated guidelines and due to contextual disparities, using regulations from other jurisdictions for analogous projects can be problematic. Thus, having guidelines tailored to assess the environmental impacts of projects within a specific jurisdictional framework is pivotal. Such guidelines resonate more closely with the legal, standard, and procedural nuances of that nation. Creating localized EIA guidelines can enhance inter-departmental cooperation and awareness.

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

141

It’s paramount that these guidelines are molded to fit the context where they’re applied. Yet, a deficiency in the provision of such guidelines is evident in numerous developing nations [15]. Colombia, being a developing country, resonates with this observed trend. Colombia lacks specific EIA guidelines for assessing the environmental impacts of projects that use emerging technologies like HKTs, leading to challenges and uncertainties during evaluations. Even though the government has shown an inclination towards energy recovery from renewable sources, there are no dedicated EIA guidelines tailored for the energy sector or for HKT projects [16]. Even though terms for environmental impact reporting were established in Colombia in 2014, they don’t cater specifically to RE projects or any kind of project [17]. The Colombian government has adopted some guidelines from other countries, but these tend to be broad and don’t cater to the unique social, political, and institutional nuances of each jurisdiction. Thus, this chapter aims to craft guidelines specifically for evaluating the environmental impacts of HKT-based power generation projects within the Colombian milieu. The insights offered in this document could also serve other similar contexts outside of Colombia. This research intends to identify and analyze environmental factors that could result in adverse effects during the deployment of HKTs. We aim to provide a comprehensive understanding of the impacts HKTs may have on the Colombian population. To achieve this, we have collected and analyzed data related to the process and identification involved in the EIA for this technology’s implementation. Our focus encompasses environmental, social, and economic consequences. Furthermore, based on these findings, we propose contingency strategies designed to reduce the project’s environmental impact.

3 Materials and Methodology 3.1

Case Study

Colombia possesses significant hydrokinetic potential due to its geographical and hydrographical location, presenting a broad range of opportunities for utilizing the country’s water resources [18]. This potential is particularly important for communities located in non-interconnected zones (NIZ) with limited or no access to energy resources, which encompasses approximately 17 departments ranging from the Caribbean to the Pacific regions, as well as the Orinoco and Amazon areas [19]. Figure 1 highlights the hydrological potential of the region by analyzing both the spatial and temporal distribution of water flows in the country and the geographical locations of the non-interconnected zones. Most of the NIZs are situated in remote areas characterized by limited infrastructure and low quality of life. The Basic Needs Index in these regions does not exceed 40% [20]. Factors such as low population density, marginal location, inadequate infrastructure, and resource scarcity contribute to high costs for residents to access

142

B. Martínez et al.

Fig. 1 Hydroenergy map of Colombia and Non-interconnected zones in Colombia

Legend Hydrokinetic_potential Departments without NIZ Departments with NIZ

electricity. Consequently, they often resort to alternative sources, including firewood, coal, and fossil fuels, to meet their basic needs [21]. Table 1 displays the number of municipalities in each department of the country affected by a lack of electricity. Approximately 49 of them have issues in the NIZ, including deficiencies in energy, water, and sewage services, as well as difficulties accessing education, health, drinking water, and communication [22]. To enhance the living conditions and increase the welfare and quality of life in all municipalities, particularly those that are isolated or not interconnected, the implementation of HKT for power generation is proposed. This solution is both accessible and affordable for these communities, considering the behavior and conditions of fluids, as well as the geographical proximity to hydroenergetic zones [23]. Moreover, the regulatory system ensures that the specific characteristics of each municipality are considered during the implementation of this technology [24].

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

143

Table 1 List of non-interconnected zones Department Amazonas Antioquia Caquetá Casanare Cauca Choco

Guainía Guaviare Meta Nariño Putumayo Vaupés Vichada

3.2

Municipality Puerto Nariño, Leticia Vigía de fuerte, Murindó Cartagena de Chaira, Solita, Solano Orocué López de Micay, Timbiquí, Guapi, Piamonte Acandí, Unguía, Juradó, Riosucio, Carmen del Darién, Bahía Solano, Bojayá, Alto Baudó, Medio Atrato, Nuquí, Certequí, Bajo Baudó, Sipí, Litoral de San Juan Inírida El retorno, Calamar, Miraflores La Uribe, Puerto Concordia, Puerto Rico, Mapiripán, La macarena Santa Bárbara de Iscuandé, El Charco, La Tola, Olaya Herrera, Mosquera, Francisco Pizarro Puerto Leguízamo Mitú, Carurú, Taraira Puerto Carreño, Cumaribo, Primavera, Santa Rosalía

Population 55.844 10.535 61.535 8.102 67.148 213.47

31.514 41.816 89.297 176.355 20.045 35.392 68.776

Design and Implementation of Hydrokinetic Mini-Turbine Systems

In this study, a hydrokinetic mini-turbine system was analyzed by creating a model based on the design of a 4.5 kW HKT prototype developed within the framework of a Cooperation Agreement between the National University of Cuyo and INVAP Ingeniería S.A. (I.I.S.A.) [25]. The use of horizontal axial flow turbines was preferred since they offer high efficiency, reliability, and environmental friendliness and represent the design with the highest percentage of applicability among HKT technologies [26]. Figure 2 illustrates the HKT design developed within the framework of the cooperation agreement between the National University of Cuyo and INVAP Ingeniería S.A. [27].

3.3

Life Cycle Assessment of a Hydrokinetic Turbine

Life Cycle Assessment (LCA) evaluates the environmental impact of an electric power system from resource extraction to end of life, ensuring that all stages are considered. It assesses the impact of hydrokinetic systems, including raw material extraction, construction, operation, maintenance, and decommissioning, as shown in Fig. 3. Traditional evaluations focus on direct costs, while LCA assesses emissions and resource use associated with engineering activities and turbine construction and operation. LCA provides a comprehensive understanding of the environmental impact of the hydrokinetic system [28].

144

B. Martínez et al.

r flow

Wate

Fig. 2 System design for the National University of Cuyo and INVAP Ingeniería S.A.

Fig. 3 Hydrokinetic energy production system

Figure 4 depicts the system boundaries of the hydrokinetic turbine life cycle. It shows the inputs of services and emissions outputs, including atmospheric and residual emissions, at each stage of the project life cycle. Higher emissions are expected during manufacturing and assembly, which involve complex activities and energy extraction from the water body. The operation stage is expected to have a direct impact on fauna and flora [28].

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

145

Fig. 4 Project life cycle system boundaries

3.4

EIA Methodology

An EIA process must adhere to the regulatory requirements of the applicable jurisdiction. For instance, in Colombia, an EIA for the construction and operation of hydroelectric power-generating plants must comply with the Terms of Reference set out in Resolution 1519 of July 26, 2017 [29]. Generally, an EIA study should describe the assessment methodologies used to enable an integrated, comprehensive, systematic, and multidisciplinary analysis. This includes discussions on causal relationships in the environmental impact assessment. The stages of the EIA process include: (i) understanding the project documents and conducting a literature review, (ii) scoping, (iii) data collection, (iv) modeling and methods, (v) analysis of alternatives, (vi) impact assessment, (vii) development of an environmental management plan, and (viii) development of a report. These stages are described below. Understanding the Project Documents and Literature Review At this stage, it is crucial to conduct several key activities to gather information about the project. These include identifying the design and technology to be used, the project components, its purpose, the stakeholders involved, and the general activities during construction, operation, and decommissioning. Additionally, it is essential to review relevant government documents to obtain information on the legal requirements for conducting an EIA and the current environmental conditions in and around the project site. Reviewing relevant literature is also advisable to understand the potential impacts of the project, the mitigation measures taken to address them, and the necessary monitoring mechanisms. Lastly, it is important to describe the international legal context and local legislative obligations related to the EIA and the environmental governance framework. Typically, HKT technology consists of an electrical power generation facility structured with system supports and a turbine. The hydrokinetic turbine design comprises a series of blades arranged in a unit, mounted underwater at equidistant distances from the water body’s surface and the ground. Water body currents are converted into oscillatory movements, which mechanical gears then transform into rotational movements. Finally, a generator connected to the gearing produces electrical energy [6]. The structure of an HKT includes an overall structural system, including the blades, stator, rotor, multiplier, and generator.

146

B. Martínez et al.

Scoping The assessment scope encompasses a preliminary analysis covering site selection, stakeholder identification, funding sources, project schedule definition, data requirements, and relevant data sources. The process involves consultations with the project proponent, local community, and regulatory agencies. Meetings with municipal authorities may address issues such as material transportation, energy supply, environmental impacts, and site selection. The scoping process facilitates the identification of all project impacts, potential mitigations, and community roles in assessing and monitoring impacts. Through comprehensive dialogue, all stakeholder concerns and suggestions are included, ensuring a sustainable and responsible project. Moreover, this process identifies opportunities to enhance efficiency and reduce environmental impact [30]. Data Collection Given the nature of the technology that will be used for energy generation from HKTs, baseline data are needed for energy generation from HKTs. Relevant information sources include government agencies, reports, literature, research organizations, and field studies. Data to be collected include weather conditions, water, air, ecosystems, and land use activities. Site-specific investigations help improve and validate available data. Habitat studies can identify ecological status near the site, and properties of soil, water, and air should be studied, including pH, nutrients, heavy metals, PM10/2.5, COx/NOx/SOx, and noise levels [31]. In addition to environmental data, collecting demographic profiles, cultural information, and socioeconomic data via surveys and secondary sources is necessary. Demographic analysis helps in understanding household occupations, indigenous community presence, and vulnerable individuals. On the other hand, it is necessary to consider heritage and industry near the site [30]. Modeling and Methods Modeling the interaction of the turbine with the environment may be necessary, depending on the data available on the design of the HKT technology. For example, water quality modeling and ecological modeling can help understand the extent of impacts on water, flora, and fauna. When modeling, parameters such as sedimentation patterns, species interaction, pollutants, system size, turbine height in the water body, and media interaction should be considered, depending on the technology type, its location, the local environment, and the capability of the HKT technology. Modeling should be conducted in a justified manner using appropriate techniques. In addition to modeling, methods such as network analysis, overlay methods, and checklist methods are used to identify and assess a project’s impacts [32]. Analysis of Alternatives The analysis of alternatives is an essential component of the EIA process for projects. It is necessary to conduct this analysis at various stages of the process to avoid or reduce physical and social impacts [33]. The project site selection phase should include evaluating the waste-to-energy technology, identifying impacts, mitigation measures, costs, and monitoring. Community acceptance, overall design, and energy production are important factors in deciding whether to

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

147

proceed. Site selection should consider topography, current land use, community acceptance, nearby habitats, and physical infrastructure. Mitigation measures should be analyzed to reduce impacts, and alternative analysis can identify better options. Methods such as multi-criteria analysis (MCA), expert opinions, cost–benefit analysis, and community involvement can be used to identify alternatives [34]. It is essential to emphasize that the analysis of alternatives must be rigorous and well documented to ensure transparency and effective participation of all stakeholders. Impact Assessment To assess the potential impacts of the HKT project, it is necessary to evaluate the construction, operation, and decommissioning phases. Predictive techniques can be used to understand the extent of predicted effects, and local communities and affected people should be involved in the assessment. The significance of the impacts must be determined based on existing environmental conditions and their nature, and the threshold level of each parameter established by regulations should be considered [17]. Development of the Environmental Management Plan Developing an environmental management plan is essential to address all potential significant environmental impacts of the project and to adopt the necessary mitigation measures to reduce them to an acceptable level [35]. The development of an environmental management plan should address impacts during all project stages and include mitigation strategies for each impact, based on their importance. It’s important to analyze different mitigation options and select the most cost-effective and least impactful one. Development of a Report After assessing the impacts of a project, it is important to prepare a report that complies with general mandates and guidelines. The EIA report should include information on the project’s background, legal requirements, findings of the assessment, potential environmental and social consequences, analysis of alternatives, and remedial measures to be taken during different project phases. It should also contain a non-technical summary for the community to understand the potential impacts [36]. It is important to provide a version of the EIA report in the local language to ensure that the information is accessible to the local community.

4 Results and Discussions 4.1

Impact of the HKTs Project

The EIA process allows for a flexible approach to impact assessment, with various methodologies applicable to different types of plans. The choice of technique and methodology depends on the specific circumstances of the case, primarily relying on a qualitative impact assessment method that incorporates a level of subjectivity in decision-making [37]. The selection of the appropriate assessment technique should be based on a review of accumulated experience from previous plans and

148

B. Martínez et al.

applications [38]. As a result, EIA is well suited for implementing the precautionary and preventive principles in spatial planning, avoiding potential environmental and social conflicts. This study examined the life cycle of the HKT, from the project implementation process to its useful life. Variations in environmental impacts at each stage of the project were presented and are described below: Project in Execution The construction of the HKT application project is estimated to take 1–4 months, depending on the level of interaction between the system and its environment, as well as the maintenance of the HKT devices [39]. Activities such as land clearing, foundation installation, and logistics are included in the project. The main expected impacts during construction are employment generation, alteration of soil and water components, air pollution from construction noise, and social and visual impacts. Table 2 presents a model for the assessment of potential impacts in the implementation phase of an HKT project, in the form of a matrix. In this table, the letters C, P, D, E, and M represent “Class,” “Presence,” “Duration,” “Evolution,” and “Magnitude,” respectively. Employment Generation Implementing a logistics management system for the HKT project can create employment opportunities for professionals with expertise in construction, installation, maintenance, and safety. This can have a significant impact on the local community by creating jobs for both internal and external personnel involved in supplying goods and services to the project [6]. Alteration in the Soil and Water Component Erosion of soil layers can increase due to activities related to excavations, construction of embankments, and land development, which cause the removal of vegetation and alterations of large volumes of soil. Areas where the land is stripped of soil protection and/or vegetation are susceptible to instability due to erosion processes. Erosion affects the soil to the point of decreasing its effectiveness as a functional element within the project solutions in the long term [40]. This process uses both nonrenewable and renewable resources, such as water for domestic and industrial uses, and electrical energy for industrial uses in the installation of equipment and verification of operation. During the construction process, which includes the installation and joining of machinery and the removal of earth layers to adapt the land, various impacts may occur. These include alterations such as the generation of construction waste, which, even when considered non-hazardous inert products, can impact the landscape environment and are also susceptible to contaminating soils and aquifers [41]. In the case of spills, there is the production of concrete mixtures at the construction site, which occurs during the stages of land preparation, creation of the end supports, and paving. Spills of concrete mixtures contribute to liquid and solid discharges containing grease, oils, sediments, and construction waste, affecting the water network in the ecosystem [42, 43]. Social Impact From a social and environmental perspective, the acquisition of property and the water body concession permit can impact the cultural heritage of

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

149

Table 2 Model for assessing potential impacts during the execution phase of an HKT project Project in execution Impact area Description of impacts Air quality Significant deterioration in air quality may occur in the study area due to the emission of particulate matter during the excavation process, waste transportation, and project design and installation. Soil loosening during excavation can contribute to this effect Noise and Health hazards and disturbance of nearby wildlife vibration Water quality Construction of infrastructure near rivers or water bodies can lead to water quality deterioration. Excavation activities can result in sand and soil draining into the water, while backfilling can also negatively impact water quality. Dust generation from construction materials and wind-borne transport can increase water turbidity Soil quality Excavation activities and soil transportation can cause the topsoil to loosen and become compacted, leading to changes in soil properties that can significantly affect the natural soil cycle and dependent ecosystems Terrestrial Land use conversion for the project and consequent loss habitat of biodiversity can have a significant impact on native vegetation and wildlife. Clearing land for construction can result in habitat loss and negatively affect local biodiversity. Construction activities, including earthwork, can disturb terrestrial habitats. Increased noise levels and frequent traffic movements associated with the project can disrupt local wildlife Landscape and Installation of a turbine in the area can have significant aesthetic view landscape impacts, including vegetation removal and structure placement. These changes may affect the aesthetic elements of the environment, including natural and cultural heritage Human health Emission of fumes, SOx, NOx, and other pollutants leads to the deterioration of human well-being

Impact evaluation C P D E M

the area as well as the natural dynamic interaction of the environment [44]. These factors interact in a similar spatiotemporal moment, generating a conflict between the project and the communities. Therefore, the dynamics and transformation processes are contextual to each society and its development. Alteration of the Air Component The generation of noise and vibrations directly affects human and environmental comfort and is considered an important environmental pollutant. It also impacts the psychological and psychosocial aspects of people, as well as the behavior of the flora and fauna in the environment [45]. Increased sound pressure in the environment has a detrimental effect on fauna and flora, creating disturbances that can stress the natural behavior of wildlife,

150

B. Martínez et al.

as well as impair the well-being of people in the vicinity of the project. This has a detrimental effect on the ecosystem, compromising the health and equilibrium of the environment. On the other hand, the transportation of materials and workers can be considered one of the main sources of environmental impact due to the production of particulate matter, which is based almost entirely on petroleum derivatives (gasoline, diesel), a limited natural resource. This PM causes various impacts on the environment and society, including a decrease in visibility in the atmosphere and pollution, which leads to an increase in respiratory diseases. This consumption is the main source of atmospheric emissions in the sector and has other negative effects on the environment, including contributing to the production of greenhouse gases [46]. Visual Impact The impact of HKTs on visual aesthetics is a topic of debate. While some may argue that these turbines have a negative impact on the visual environment, others may contend that they can be incorporated into the landscape in a way that is visually appealing. The placement and design of turbines can play a crucial role in determining their visual impact, with factors such as distance from shore, blade design, and color scheme all contributing to their overall appearance. Additionally, the perception of visual impact can vary depending on the individual and their relationship to the environment in question. Some may see hydrokinetic turbines as an innovative and sustainable energy solution, while others may view them as an eyesore. Ultimately, whether hydrokinetic turbines have a significant impact on the visual environment is subjective and depends on a variety of factors. Project in Operation The useful life of a turbine can be extended up to 25 years [47], depending on the technology employed, the maintenance of the equipment, and the interaction of the turbine with the environment. When operating a turbine system, a variety of environmental effects can be generated, including air, soil, and water pollution, as well as impacts on human health and ecology. Therefore, it is critical to conduct an adequate assessment of the processes and implement proper internal logistics to minimize these effects. Table 3 presents a model for assessing potential impacts during the operation phase of an HKT project. Emission of Greenhouse Gases and Other Pollutants to Air Hydrokinetic turbines are often lauded for their potential to generate clean, renewable energy without producing greenhouse gas emissions. However, it is important to consider the full life cycle of these technologies, including their impacts on air quality and greenhouse gas emissions. One key area of concern is the transportation of hydrokinetic turbines to their installation sites. This process can involve the use of fossil-fuelpowered vehicles, such as trucks and boats, which can contribute to air pollution and greenhouse gas emissions. Additionally, the transportation of personnel to and from the installation site can impact local air quality and emissions. Noise and Vibration The installation and operation of hydrokinetic turbines have the potential to significantly impact the aquatic environment in which they are placed. The vibrations emitted by these turbines can alter the behavior of fish and

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

151

Table 3 Model for assessing potential impacts during the operation phase of an HKT project Project in operation Impact area Description of impacts Water and The turbine’s operation can change sediment transport, aquatic habitat leading to water quality deterioration in nearby lakes or rivers. The turbine’s operation can also affect aquatic habitats, causing species loss and behavioral changes due to enclosure and water quality alterations Terrestrial The project’s spatial limitations can alter natural ecohabitat system interactions, potentially leading to biodiversity loss. It is essential to consider the project’s potential impact on the surrounding environment and take measures to minimize any adverse effects Air quality Significant air quality deterioration may occur in the study area due to particulate matter emissions from material degradation during operation. Soil loosening caused by system vibrations may contribute to this effect Noise and The project’s operation can generate vibrations that vibration may affect the behavior of fauna interacting with the turbine. These vibrations can cause stress due to environmental changes Interruption of The application of this technology can result in biotic system changes, potentially affecting the local econcommunity life omy in the industrial and fishing sectors. These changes may also have social implications, especially for downstream communities Soil quality Changes in water flow and sediment transport caused by a turbine can lead to erosion, potentially resulting in land loss in the study area and impacting the local ecosystem

Impact evaluation C P D E M

other organisms living in the water, leading to changes in their feeding, mating, and migration patterns. These changes, in turn, can have severe repercussions on the reproduction and survival of these species, which is a significant problem caused by this phenomenon. Water Quality and Aquatic Habitat The design and operation of the turbine greatly influence the potential mortality of aquatic ecosystem species. Although small turbines tend to avoid impacts on their structure, the rotational speeds of the rotor increase the probability of damage to the system. These can present risks of collision, showing signs of maceration, laceration, abrasion, and contusion in the fauna. The level of injury depends on the design of the blade, the animal, and the circumstantial conditions. The orientation of the body during impact also has a significant influence on the level of injury to both the animal and the turbine structure [48]. Turbines also affect the distribution patterns of sediments, as the amount of kinetic energy in water may be modified, altering the turbulent boundary. Additionally, turbines can affect water quality by causing an increase in turbidity, reducing light penetration into the

152

B. Martínez et al.

Table 4 Model for evaluating potential impacts during the dismantling phase of an HKT project Dismantling of the project Impact area Description of impacts Landscape and The project’s dismantling may have negative effects on aesthetic view the local population by harming the cultural or environmental heritage of the area. It could also cause displeasure by altering the landscape and environment Air quality Dismantling the project can lead to a decrease in air quality due to dust and emissions from machinery and vehicles. This may adversely affect the health of the local population and the surrounding environment Noise and The use of dismantling machinery that produces high vibration sound pressures must be considered during the dismantling process. This could lead to disturbances in the area’s native fauna, as well as issues related to the wellbeing of workers and users Waste disposal The accumulation of construction waste may cause soil and water pollution if solid and liquid waste is not properly disposed of or adequately treated. Proper waste management is essential to prevent environmental contamination

Impact evaluation C P D E M

water, and affecting the photic zone boundary, as well as temperature and dissolved oxygen levels, resulting in the modification of aquatic habitat. Moreover, metallic contaminants from the turbine structure could be trapped in the sediment bed and reintroduced to the water column by bed erosion [49]. In general, the installation of hydrokinetic turbines can alter the natural flow of water, causing changes in water temperature, velocity, and sediment transport. These changes can lead to alterations in aquatic ecosystems, including changes in habitat availability, fish migration patterns, and food webs. Social and Natural Impact The social impacts of hydrokinetic turbines can be significant and far-reaching, particularly for communities that rely on local waterways for their livelihoods or cultural heritage. Changes in water flow and sediment distribution can affect local fishing and recreational activities, while the installation and operation of turbines can disrupt cultural heritage sites and traditions. Moreover, changes in property values due to the presence of turbines can also have economic impacts on local communities. Dismantling of the Project At the end of an HKT’s useful life, all the facilities and infrastructure created will be dismantled to return the land to its pre-installation condition. The disposal of materials will be carried out in accordance with existing waste legislation. To begin with, the different dismantling and restoration operations must be identified; afterward, the individual tasks of each area must be specified, justified, and evaluated economically; subsequently, the Restoration and Revegetation Plan must be drawn up, with an economic appraisal; and finally, the waste

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

153

generated from the dismantling works must be quantified and appraised [50]. Table 4 presents a model in matrix form for assessing potential impacts during the dismantling phase of an HKT project. Impact on the Environment (Water, Air, and Soil) Construction waste can be generated not only during the construction phase but also during the termination of the useful life of each stage and the subsequent deconstruction and demolition of equipment. This last stage is important in terms of potential waste volume as it may contain persistent or biodegradable substances that can become pollutants in various forms, such as gaseous, leachates, or sediments [51]. These are present in the preliminary stages of construction of stockpiles and containment barriers, as well as in the disassembling of each piece of machinery used in the system for the application of electric power generation from the propeller-type hydrokinetic turbine. On the other hand, the transportation of construction materials is one of the main sources of environmental impact because it generates particulate matter in the transfer of construction waste to the adaptation sites. This is based on the consumption of energy products derived from petroleum (gasoline, diesel), a limited natural resource [46]. The consumption of oil-based resources is the primary source of the sector’s atmospheric emissions, which negatively affect the environment [52], as well as contributing to the production of GHGs. It should be noted that the dismantling of machinery also generates greenhouse gases and particulate matter, albeit on a smaller scale. The generation of noise and vibrations in the dismantling process of each of the equipment used in the project can also be cited as another effect caused by the decoupling activity of the stages. This can affect the health of workers and alter the local fauna due to increased stress resulting from auditory disturbances [45].

4.2

Evaluation of Impacts

Impact assessment is a key tool for evaluating and assessing the positive or negative effects that environmental aspects have on the environment, society, and the ecosystem [33]. It is necessary to evaluate the significance of potential impacts according to strict criteria, considering the social, territorial, and environmental context of a country. This should include factors such as whether the impacts comply with current regulations and standards, any impact on endangered species, protected species, and protected areas, the positive or negative consequences of the environmental change produced, the likelihood that the impact will occur, the duration of the impact’s active existence, the speed of the impact’s development from the time it begins until it manifests with all its consequences, the magnitude or size of the environmental change produced by an activity, the proximity to cultural and natural heritage, and the perception of the impacts by the local community. In Table 5, a classification of the impacts can be established, and a range of numerical values can

154

B. Martínez et al.

Table 5 Categorization of environmental significance Value 8–10

Importance Very high

6–7.9

High

4–5.9 2–3.9

Medium Low

0–1.9

Very low

Meaning The impact is inconsequential in comparison with the aims and objectives of the project in question The impact is not highly relevant in comparison with the aims and objectives of the project in question No intensive corrective or protective measures are required The impact necessitates the restoration of environmental conditions through corrective or protective measures. The necessary recovery time is an extended period The impact exceeds the acceptable threshold. There is a permanent loss of quality in environmental conditions. There is NO possibility of recovery

be assigned to facilitate understanding [53]. It is also important to consider whether the impacts are reversible or irreversible, according to the opinion of experts at the time of the evaluation.

4.3

Contingency Plan

To mitigate the environmental impact of an HKT project, a contingency plan should be developed to identify potential risks and outline how to address them. The plan should consider both intentional and unintentional human-caused risks and should involve the local community in finding suitable solutions. The plan should be structured to document the project processes and conduct a risk assessment analysis to identify the most significant impacts that could occur during the project’s execution, operation, and decommissioning [54]. Project in Execution During the construction phase, it is essential to prepare a detailed contingency plan due to the potential environmental impacts. This plan should consider a range of possible scenarios to anticipate and manage any unforeseen events that could affect the environment [54]. This allows for responsible and sustainable construction practices that can minimize the environmental impact. Air Quality To mitigate air pollution and dust generation during construction, several preventive measures should be taken. These measures include: • Watering the active construction site daily, especially during the dry season, to reduce dust. • Properly covering trucks that transport soil, sand, and other loose construction materials to prevent particle dispersal • Using transportation vehicles that comply with legislative regulations regarding particulate matter emissions.

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

155

Water Quality To prevent the discharge of pollutants from the excavated soil and construction materials that are piled up near water bodies, it is essential to store them in appropriate locations and manage them in an environmentally friendly manner. This prevents pollutants from entering nearby water bodies during the rainy season, minimizing their environmental impact. Soil Quality, Landscape, and Aesthetic View To mitigate the loss of vegetation resulting from land use for the project, consider planting similar species of trees or conducting a re-vegetation process. Trees should be planted in open areas of the project site whenever possible [55]. It’s also important to address potential negative visual impacts to protect historical, natural, and cultural heritage. These impacts might involve changes in the visual appearance or aesthetics of built structures or natural landmarks [56]. Proper site selection, appropriate design, and implementation that minimizes habitat loss and preserves the area’s aesthetic value are essential. Consultation with local communities can help in more effectively addressing the project’s impacts, considering the needs and opinions of those who live in the area. This promotes responsible and sustainable project management that minimizes environmental impact and protects the surrounding environment [30]. Human Health During HKT operations, the use of appropriate measures and technologies is crucial to reduce emissions and protect the environment. The type and amount of pollutants released vary depending on the technology used. Therefore, prioritizing equipment and technologies that minimize their impact on human health and reduce pollutant emissions is essential. Project in Operation Given that the typical operational lifespan of an HKT is around 25 years [47], and considering that turbine operation can have a significant environmental impact, it is crucial to implement contingency measures to minimize these effects. Aquatic Habitat One of the primary benefits of incorporating a hydroelectric power generation system within a turbine is the potential to introduce measures to protect the aquatic fauna within the water body. One such measure is the installation of a mesh or barrier at the system’s entrance, which can significantly reduce species mortality due to collisions with the turbine blades [57]. Furthermore, this structure not only safeguards aquatic biodiversity but also enhances the hydroelectric system’s efficiency by preventing turbine obstructions from the buildup of unwanted materials [58]. Interruption of Community Life To mitigate disruptions to the local economy, particularly in the fisheries sector, an effective strategy is to restock aquatic species. Translocating native fish within the same watershed, from capture to release at stocking sites, can be successfully undertaken to ensure fish population sustainability [59]. Minimizing fish mortality during their time away from their natural habitat is crucial, and reducing their stress is essential. This can be achieved by following these steps:

156

• • • •

B. Martínez et al.

Avoid capturing juvenile specimens. Only collect necessary data from each captured individual. Minimize the time fish spend in aerated containers. Transport specimens in containers with low stocking densities.

Soil Quality To minimize scouring on the water body’s slopes and prevent disasters, engineering works such as dikes, bank protections, and bottom controls are constructed as mitigation measures. The implementation of these measures varies based on factors like the channel, sediments, flow, and others, with the goal of preventing scour development and protecting the environment from adverse effects. Air Quality To control particulate matter emissions during turbine system operation, it is recommended to wet the surface to prevent landslides and use air filters in heavy transport vehicles. Quarterly measurements should be taken to monitor emission levels, using three control points—one at the source and two in nearby towns for air quality monitoring. Project Dismantling Given that decommissioning an HKT can have substantial environmental impacts during the dismantling process, it’s crucial to adopt risk management measures to mitigate these effects. Soil Quality, Landscape, and Aesthetic View To minimize vegetation loss due to land use for the plant, it’s essential to consider planting similar tree species or initiating a re-vegetation process. Trees should be planted in open spaces on the project site when feasible [55]. Moreover, it’s important to address potential negative visual impacts from the project, aiming to preserve historical, natural, and cultural heritage. Impacts might include alterations in the appearance or aesthetic value of built or natural heritage [56]. Hence, proper site selection, thoughtful design, and implementation that limits habitat loss and maintains the area’s aesthetic value are crucial. Consulting local communities can be beneficial in effectively addressing project impacts, considering the needs and opinions of residents. This approach ensures responsible, sustainable project management, minimizes environmental impacts, and safeguards the surrounding environment [30]. Air Quality During the dismantling phase, air quality might temporarily worsen due to fuel burning, vehicle emissions from construction waste transport, and equipment material removal. To mitigate air pollution and dust generation, the following measures should be implemented: • Water the active dismantling area daily to minimize dust production. • Cover trucks transporting soil, sand, and other loose construction materials to prevent particle dispersion into the air. • Monitor and enforce transport speed limits to avoid material release into the atmosphere. • Maintain transportation equipment to reduce particulate matter generation.

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

157

Waste Disposal A comprehensive waste management plan is vital for minimizing project-generated construction waste and managing produced waste, including transportation, storage, and disposal through specialized third parties. The waste management plan might include: • An exhaustive analysis of waste generated throughout the HKT’s life cycle. • A specific procedure for handling, transporting, treating, and disposing of each waste type. • Engaging specialized third parties for proper waste disposal. • A risk plan addressing the management of various construction waste types. • Measures to ensure the health and safety of personnel involved in waste management. • Staff training to enhance awareness of waste types and appropriate handling procedures.

5 Conclusions and Recommendations The lack of formal guidelines has motivated the authors to develop a generic methodology for Environmental Impact Assessments (EIA) and impact assessment during the various phases of HKT projects. This includes the creation of contingency plans and monitoring strategies. The information provided here may be beneficial for evaluating the environmental impacts of HKT projects in Colombia and other developing countries. For an effective EIA of HKT projects that can aid policymakers in making informed decisions, proponents and consultants should consider the following points: • • • • • •

Establish guidelines for the EIA of HKT projects. Begin both EIA and project planning simultaneously. Ensure regular interaction between the EIA team and the engineering team. Utilize suitable methods for community engagement. Allocate sufficient time for conducting the EIA. Establish an independent review committee to evaluate the EIA report, ensuring its quality.

The EIA is a valuable tool for environmental assessments of various processes, projects, or activities, as it is widely applicable and efficient. It can be tailored to the specific needs of the activity, leading to more accurate and quicker results. However, mechanisms that promote broader participation in the development of matrices are essential for obtaining more precise, representative, and reliable data.

158

B. Martínez et al.

References 1. Yuce, M.I., Muratoglu, A.: Hydrokinetic energy conversion systems: A technology status review. Renew. Sust. Energ. Rev. 43, 72–82 (2015). https://doi.org/10.1016/j.rser.2014.10.037 2. Owusu, P.A., Asumadu-Sarkodie, S.: A review of renewable energy sources, sustainability issues and climate change mitigation. Cogent Eng. 3(1), 1167990 (2016). https://doi.org/10. 1080/23311916.2016.1167990 3. Montiel-Bohórquez, N.D., Saldarriaga-Loaiza, J.D., Pérez, J.F.: Analysis of investment incentives for power generation based on an integrated plasma gasification combined cycle power plant using municipal solid waste. Case Stud. Thermal Eng. 30, 101748 (2022). https://doi.org/ 10.1016/j.csite.2021.101748 4. Chala, G.T., Ma’Arof, M.I.N., Sharma, R.: Trends in an increased dependence towards hydropower energy utilization—A short review. Cogent Eng. 6(1) (2019). https://doi.org/10.1080/ 23311916.2019.1631541 5. Badrul Salleh, M., Kamaruddin, N.M., Mohamed-Kassim, Z.: Savonius hydrokinetic turbines for a sustainable river-based energy extraction: A review of the technology and potential applications in Malaysia. Sustain. Energy Technol. Assess. 36, 100554 (2019). https://doi. org/10.1016/j.seta.2019.100554 6. Kumar, D., Sarkar, S.: A review on the technology, performance, design optimization, reliability, techno-economics and environmental impacts of hydrokinetic energy conversion systems. Renew. Sust. Energ. Rev. 58, 796–813 (2016). https://doi.org/10.1016/j.rser.2015.12.247 7. Saini, G., Saini, R.P.: A numerical analysis to study the effect of radius ratio and attachment angle on hybrid hydrokinetic turbine performance. Energy Sustain. Dev. 47, 94–106 (2018). https://doi.org/10.1016/j.esd.2018.09.005 8. Vitorino, M.E., Labriola, C.V.M., Moyano, H.A.: Sistemas conversores fluido-Dinámicos de energía renovable para la Patagonia Argentina. Informe Científico Técnico UNPA. 8(2), 113–138 (2016) 9. UNCUYO: Construirán microturbinas para generar energía eléctrica. Universidad Nacional de Cuyo (2015) 10. Maldar, N.R., Ng, C.Y., Patel, M.S., Oguz, E.: Potential and prospects of hydrokinetic energy in Malaysia: A review. Sustain. Energy Technol. Assess. 52, 102265 (2022). https://doi.org/10. 1016/j.seta.2022.102265 11. Romero-Gomez, P., Richmond, M.C.: Simulating blade-strike on fish passing through marine hydrokinetic turbines. Renew. Energy. 71, 401–413 (2014). https://doi.org/10.1016/j.renene. 2014.05.051 12. Al-Nasrawi, F.A., Kareem, S.L., Saleh, L.A.: Using the Leopold Matrix procedure to assess the environmental impact of pollution from drinking water projects in Karbala City, Iraq. IOP Conf. Ser. Mater. Sci. Eng. 671(1), 012078 (2020). https://doi.org/10.1088/1757-899X/671/1/012078 13. Delaney, J., Basta, D.J.: Florida Keys National Marine Sanctuary environmental assessment on the issuance of permit# FKNMS-2007-122 for testing a hydrokinetic turbine in sanctuary waters (2009) 14. Yadav, A., Hema, H.C., Shivakumara, M.J.: Environmental impact assessment on renewable energy: A review. IOP Conf. Ser. Earth Environ. Sci. 573(1), 012048 (2020). https://doi.org/10. 1088/1755-1315/573/1/012048 15. Aung, T.S., Fischer, T.B., Shengji, L.: Evaluating environmental impact assessment (EIA) in the countries along the belt and road initiatives: System effectiveness and the compatibility with the Chinese EIA. Environ. Impact Assess. Rev. 81, 106361 (2020). https://doi.org/10.1016/j. eiar.2019.106361 16. Behrouzi, F., Nakisa, M., Maimun, A., Ahmed, Y.M.: Global renewable energy and its potential in Malaysia: A review of Hydrokinetic turbine technology. Renew. Sust. Energ. Rev. 62, 1270–1281 (2016). https://doi.org/10.1016/j.rser.2016.05.020

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

159

17. Mendoza Zapata, L.A., Pacheco Bustos, C.A., Certain Abraham, W.D.: Assessment of environmental impacts associated with the eventual environmental recovery of quarries with inert construction and demolition waste in Barranquilla and its metropolitan area. Ing. Desarrollo. 39(2), 275–295 (2021) 18. Montoya Ramírez, R.D., Cuervo, F.I., Monsalve Rico, C.A.: Technical and financial valuation of hydrokinetic power in the discharge channels of large hydropower plants in Colombia: A case study. Renew. Energy. 99, 136–147 (2016). https://doi.org/10.1016/j.renene.2016.06.047 19. Ch, F.D.M.: Issues associated with the implementation of wind energy power generation in isolated and non-interconnected rural areas—Case study. WSEAS Trans. Environ. Dev. 14, 436–445 (2018) 20. Rodríguez-Urrego, D., Rodríguez-Urrego, L.: Photovoltaic energy in Colombia: Current status, inventory, policies and future prospects. Renew. Sust. Energ. Rev. 92, 160–170 (2018). https:// doi.org/10.1016/j.rser.2018.04.065 21. Bhutto, A.W., Bazmi, A.A., Zahedi, G.: Greener energy: Issues and challenges for Pakistan— Biomass energy prospective. Renew. Sust. Energ. Rev. 15(6), 3207–3219 (2011). https://doi. org/10.1016/j.rser.2011.04.015 22. Grupo de Investigavión XUÉ: Estado de la Cobertura Eléctrica y las Zonas No Interconectadas en la Región Central (2020) 23. Rolong Ortiz, H., Acevedo Peñaloza, C.H., Valencia, G.: CFD performance analysis for Darrieus Hydrokinetic Turbine. J. Eng. Sci. Technol. Rev. 12(2), 40–45 (2019) 24. Lopez Lezama, J.M., Villada, F., Muñoz Galeano, N.: Effects of incentives for renewable energy in Colombia. Ing. Univ. 21(2) (2017). https://doi.org/10.11144/Javeriana.iyu21-2.eire 25. Tripp, N.G.: Modelaciones matemáticas de turbinas hidrocinéticas en canales de riego (2018) 26. Kumar, A., Saini, R.P., Saini, G., Dwivedi, G.: Effect of number of stages on the performance characteristics of modified Savonius hydrokinetic turbine. Ocean Eng. 217, 108090 (2020). https://doi.org/10.1016/j.oceaneng.2020.108090 27. Arauco Camargo, E.A.: Diseño de una turbina hidrocinética de eje horizontal para la generación de energía eléctrica en zonas ribereñas (2021) 28. Miller, V.B., Landis, A.E., Schaefer, L.A.: A benchmark for life cycle air emissions and life cycle impact assessment of hydrokinetic energy extraction using life cycle assessment. Renew. Energy. 36(3), 1040–1046 (2011). https://doi.org/10.1016/j.renene.2010.08.016 29. Ministerio de Ambiente y Desarrollo Sostenible: Resolución 1519 de 2017. Bogotá (2017) 30. Milanez, B.: Dialogues between social and natural sciences: Contribution to the debate on socio-environmental conflicts. An. Acad. Bras. Cienc. 87(4), 2335–2348 (2015). https://doi.org/ 10.1590/0001-3765201520140724 31. Kabir, Z., Khan, I.: Environmental impact assessment of waste to energy projects in developing countries: General guidelines in the context of Bangladesh. Sustain. Energy Technol. Assess. 37, 100619 (2020). https://doi.org/10.1016/j.seta.2019.100619 32. Banyal, S., Aggarwal, R., Bhardwaj, S.: A review on methodologies adopted during environmental impact assessment of development projects. J. Pharmacogn. Phytochem. 8(4), 2108–2119 (2019). https://doi.org/10.22271/phyto.2019.v8.i4aj.9273 33. Bishoge, O.K., Mvile, B.N.: A critique of the EIA Report selected from the East African region, taking into consideration what is required in an ideal of EIA Report. J. Appl. Adv. Res., 8–17 (2022) 10.21839/jaar.2022.v7.7478 34. Mouter, N., Dean, M., Koopmans, C., Vassallo, J.M.: Comparing cost-benefit analysis and multi-criteria analysis, pp. 225–254 (2020). https://doi.org/10.1016/bs.atpp.2020.07.009 35. R. E. I. Assessment: Environmental Management Plan. Chittamgondi Bauxite Deposit, Chittamgondi Village, Visakhapatnam District, Andhra Pradesh. Andhra Pradesh Mineral Development Corporation Limited, Secundrabad, India by Geo Experimental Labs, Secundrabad (2022) 36. Alvarez, L.B.: A Beginner’s Guide to Urban Design and Development: The ABC of Quality, Sustainable Design. Taylor & Francis (2023)

160

B. Martínez et al.

37. Josimović, B., Todorović, D., Jovović, A., Manić, B.: Air pollution modeling to support strategic environmental assessment: case study—National Emission Reduction Plan for coal-fired thermal power plants in Serbia. Environ. Dev. Sustain. (2023). https://doi.org/10. 1007/s10668-023-03186-0 38. Josimović, B., Cvjetić, A., Furundžić, D.: Strategic Environmental Assessment and the precautionary principle in the spatial planning of wind farms—European experience in Serbia. Renew. Sust. Energ. Rev. 136, 110459 (2021). https://doi.org/10.1016/j.rser.2020.110459 39. Espina-Valdés, R., Fernández-Jiménez, A., Fernández Francos, J., Blanco Marigorta, E., Álvarez-Álvarez, E.: Small cross-flow turbine: Design and testing in high blockage conditions. Energy Convers. Manag. 213, 112863 (2020). https://doi.org/10.1016/j.enconman.2020. 112863 40. Hassani, S., Bielawski, M., Beres, W., Martinu, L., Balazinski, M., Klemberg-Sapieha, J.E.: Predictive tools for the design of erosion resistant coatings. Surf. Coat. Technol. 203(3–4), 204–210 (2008). https://doi.org/10.1016/j.surfcoat.2008.08.050 41. Crawford, R.H., Mathur, D., Gerritsen, R.: Barriers to improving the environmental performance of construction waste management in remote communities. Proc. Eng. 196, 830–837 (2017). https://doi.org/10.1016/j.proeng.2017.08.014 42. Quaranta, E., Davies, P.: Emerging and innovative materials for hydropower engineering applications: turbines, bearings, sealing, dams and waterways, and ocean power. Engineering. 8, 148–158 (2022). https://doi.org/10.1016/j.eng.2021.06.025 43. Sonebi, M., Ammar, Y., Diederich, P.: Sustainability of cement, concrete and cement replacement materials in construction. In: Sustainability of Construction Materials, pp. 371–396. Elsevier (2016). https://doi.org/10.1016/B978-0-08-100370-1.00015-9 44. Verburg, P.H., et al.: Land system science and sustainable development of the earth system: A global land project perspective. Anthropocene. 12, 29–41 (2015). https://doi.org/10.1016/j. ancene.2015.09.004 45. Gatersleben, B., Griffin, I.: Environ. Stress, 469–485 (2017). https://doi.org/10.1007/978-3319-31416-7_25 46. Tan, K.W., Kirke, B., Anyi, M.: Small-scale hydrokinetic turbines for remote community electrification. Energy Sustain. Dev. 63, 41–50 (2021). https://doi.org/10.1016/j.esd.2021. 05.005 47. Lata-García, J., Jurado, F., Fernández-Ramírez, L.M., Sánchez-Sainz, H.: Optimal hydrokinetic turbine location and techno-economic analysis of a hybrid system based on photovoltaic/ hydrokinetic/hydrogen/battery. Energy. 159, 611–620 (2018). https://doi.org/10.1016/j. energy.2018.06.183 48. Hammar, L., et al.: A probabilistic model for hydrokinetic turbine collision risks: Exploring impacts on fish. PLoS One. 10(3), e0117756 (2015). https://doi.org/10.1371/journal.pone. 0117756 49. Ramírez-Mendoza, R., et al.: Laboratory study on the effects of hydro kinetic turbines on hydrodynamics and sediment dynamics. Renew. Energy. 129, 271–284 (2018). https://doi.org/ 10.1016/j.renene.2018.05.094 50. Rivas, C.: Piensa un minuto antes de actuar: gestión integral de residuos sólidos. Ministerio de Ambiente (2018) 51. Christensen, T.H., et al.: Attenuation of landfill leachate pollutants in aquifers. Crit. Rev. Environ. Sci. Technol. 24(2), 119–202 (1994). https://doi.org/10.1080/10643389409388463 52. Hosseini, S.E., Wahid, M.A., Aghili, N.: The scenario of greenhouse gases reduction in Malaysia. Renew. Sust. Energ. Rev. 28, 400–409 (2013). https://doi.org/10.1016/j.rser.2013. 08.045 53. Arboleda González, J.A.: Metodología para la identificación y evaluación de impactos ambientales, p. 224. Empresas Públicas de Medellín, Medellín (2003) 54. Gavouneli, M.: Regional arrangements for contingency planning and response: The EU regime. In: Managing the Risk of Offshore Oil and Gas Accidents: The International Legal Dimension. Edgar (2019)

Guidelines for Environmental Impact Assessment of Hydrokinetic Turbines. . .

161

55. Andreassen, N.: Arctic Energy Summit Executive Summary 2015. Institute of the North (2015) 56. Laín, S., Contreras, L.T., López, O.: A review on computational fluid dynamics modeling and simulation of horizontal axis hydrokinetic turbines. J. Braz. Soc. Mech. Sci. Eng. 41(9), 375 (2019). https://doi.org/10.1007/s40430-019-1877-6 57. Brown, E., Sulaeman, S., Quispe-Abad, R., Müller, N., Moran, E.: Safe passage for fish: The case for in-stream turbines. Renew. Sust. Energ. Rev. 173, 113034 (2023). https://doi.org/10. 1016/j.rser.2022.113034 58. Tong, C.: Advanced materials and devices for hydropower and ocean energy. In: Introduction to Materials for Advanced Energy Systems, pp. 445–501. Springer, Cham (2019). https://doi.org/ 10.1007/978-3-319-98002-7_7 59. Quaranta, E., et al.: Hydropower case study collection: Innovative low head and ecologically improved turbines, hydropower in existing infrastructures, hydropeaking reduction, digitalization and governing systems. Sustainability. 12(21), 8873 (2020). https://doi.org/10.3390/ su12218873

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas Garcés Anggie, Ramírez Charles, Juan Peralta-Jaramillo, Emérita Delgado-Plaza, Jorge Abad-Moran, Jorge Hurel, Guido Abril, and Ian Sosa

1 Introduction Livestock plays a significant role in human-generated carbon dioxide (CO2) emissions, contributing to around 9% of the total. Studies from the Nature Geoscience journal highlight the substantial impact of cow manure, a major source of atmospheric nitrous oxide and particularly harmful greenhouse gases. This waste also triggers soil degradation and contaminates precious water resources [1]. This project aims to create a biodigester model for treating organic cattle waste and generating biogas in rural settings. The biogas produced serves various purposes, from domestic use to commercial and thermal applications like grain drying and even the production of value-added items like fruits [2, 3]. At the heart of the biodigester lies the anaerobic digestion (AD) process, enabling the breakdown of organic matter to generate biogas—a gas mixture primarily composed of CO2 and methane (CH4). AD also yields biofertilizer, consisting of solid and liquid components called biol [4, 5]. Functioning as a container, the biodigester houses organic matter and microorganisms responsible for AD. The key to its efficiency is maintaining an oxygen-free environment, thus creating anaerobic conditions [6]. Various basic biodigester models exist, providing a foundation for detailed designs. For rural biogas production, fixed dome, floating dome, and tubular models G. Anggie · R. Charles · J. Peralta-Jaramillo · E. Delgado-Plaza (*) · J. Abad-Moran · J. Hurel · G. Abril Escuela Superior Politécnica del Litoral ESPOL, Centro de Desarrollo Tecnológico Sustentable, Facultad de Ingeniería en Mecánica y Ciencias de la Producción, ESPOL Polytechnic University, Guayaquil, Ecuador e-mail: [email protected] I. Sosa Electrical Engineering Department, Instituto Tecnológico de Sonora, Ciudad Obregón, Mexico © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_10

163

164

G. Anggie et al.

prove most common. The fixed dome biodigester, constructed underground to stabilize temperature fluctuations, employs sturdy materials such as concrete or bricks. Its unique feature is a dome connected to the digestion chamber in a single structure. Biogas generated during AD accumulates within the crown before extraction via the upper valve. The Chinese biodigester exemplifies this design [7]. In contrast, the floating dome biodigester incorporates a rigid digestion chamber and a movable dome to store biogas. As biogas production begins, pressure against the crown causes it to rise. When biogas is extracted, the dome descends, maintaining consistent internal pressure. The Indian biodigester model is a widely used example [6]. The tubular biodigester, boasting a straightforward design and cost-effective construction, primarily utilizes cattle manure. Lack of thermal insulation exposes it to temperature fluctuations, prompting its construction in greenhouses or temperature-controlled environments. This adaptable design is the flexible or Taiwanese biodigester [8].

2 Methodology Before proceeding with the design, knowing the climatic conditions was necessary. So, it was considered a place for the application of the biodigester, the rural area of Balzar in the province of Guayas in Ecuador. This area has environmental temperatures of around 21
C for the lowest and 33
C for the highest. In addition, it has an ambient pressure of 1.01058 bar [9]. Cattle manure was considered raw material, which is mixed with water to form the substrate that will enter the biodigester. Given the conditions, it is recommended that there is a water–manure ratio of 2:1. That is, the substrate’s volume contains two parts of water and one part of manure [10]. This work is based on searching and collecting information from similar results by other authors under conditions identical to those proposed in this project. To estimate the most significant parameters involved in the biodigester design, the works of five authors were consulted, listed in Table 1. The fixed dome model was selected because it provides considerable advantages for the conditions raised (Table 2).

Table 1 Comparison of essential variables for the design of a biodigester Parameters kg/day Manure: Water Ratio HRT m3 biogas/kg manure Temperature
C

Author 1a 7.5 1 49 1.8E-06 13.35

Author 1a 7.5 1 38 2.8E-06 18.17

Author 2b 7.5 1 50 0.00072 14.27

Author 3c 10 1 50 0.032 22

Author 4d 10.67 1:0.75 28 0.026 26

Author 5e 4.42 1:3 40 0.039 24.53

Note: aLara Guilcapi [11], bCoronel Durazno [12], cCastillo Monar [13], dGordón and Samaniego [14], eBarzallo Bravo [15]

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

165

Table 2 Measures to size the biodigester [16] Volume m3 Total Load 7.60 6.14 9.84 7.78 12.46 9.72 14.80 11.23 19.01 14.51

L 660 680 700 720 740

r 120 130 140 150 160

H 220 230 240 250 260

Measurements (cm) R1 R2 h 255 150 30 266 160 32 296 170 35 321 180 37 346 190 40

h1 120 130 140 150 160

h2 60 70 80 90 100

Fig. 1 Biodigester shape design [16]

Considering the reference data proposed by Groppelli Eduardo in Table 1, the following values are estimated: • A cow can produce approximately 7.9 kg of manure per day. • Production of 0.016 m3 of biogas per kilogram of manure. • Hydraulic retention time (HRT) of 43 days. Considering the rising demand for livestock-derived products, farmers are inclined to expand their livestock population. Consequently, the aim is to ensure that the biodigester can accommodate a substantial quantity of manure produced. Therefore, the model containing 19.01 m3 was selected (Fig. 1). Since manure collection can vary depending on ground conditions, cow dispersion, and the method used, we have decided to make different graphs that show the behavior of the other variables on which the percentage of manure collection depends. The most critical conditions are below. It is important to mention that all the graphs show the results considering a time of 43 days, that is, the HRT. It was assumed that the consumption of LPG tanks in rural areas is two tanks every 43 days. Based on this, we proceeded to estimate the savings when starting to consume biogas depending on the manure biomass collected. To establish an energy equivalence between LPG and biogas, it was decided to relate them utilizing their calorific values. In this way, an estimate of how much

166

G. Anggie et al.

biogas is required to obtain the same amount of energy as an LPG tank can be obtained. The calorific value of LPG is 10,830 [kcal/kg]. [17] On the other hand, the calorific value of biogas with low methane content is estimated at 4700 [kcal/ m3] [18]. 4700 Equivalence ¼

kcal m3 Biogas

10830



1 1:2

kcal kg LPG

kg Biogas m3 Biogas

¼ 0:36

kg LPG kg Biogas

ð1Þ

Therefore, it is estimated that 0.36 [kg] of LPG produces the same amount of energy as 1 [kg] of biogas. The estimation of biogas and biol production was made based on the equations shown in the biogas manual and the theoretical–practical guide on biogas and biodigesters, both belonging to the catalog of the Food and Agriculture Organization of the United Nations (also known as FAO). To estimate the production values, it was considered that the cattle have a minimum of 10 cows and a maximum of 15.

3 Results Table 3 shows the variables that must be considered to carry out the biodigester design. Among these are the TRH, the amount of manure produced, the amount of substrate that will enter the biodigester, and the estimation of the production of biogas and biol. It was considered to design a dome of stainless steel that is fastened by selfdrilling bolts to the concrete structure of the biodigester, and this design aims to make it easier to maintain the biodigester. Table 4 highlights the biodigester’s internal pressure and the metal dome’s thickness. Figure 2 shows that the higher the percentage of manure collection, the higher the production of biol and, therefore, the profit. It should be noted that the amount of profit shown in the graph includes the cost of transportation and biol storage tanks. For the price of transportation, a route of 150 [km] and a consumption of 35 l of diesel per 100 km were estimated, with the cost of diesel at $ 1.9 per gallon. In addition, it was considered that the biol is stored in tanks of 55 gallons, which cost $30. The value set for the sale of biol was estimated at $0.50 for each liter of biol. However, this value may vary depending on the buyer. Considering the sale price, it can be analyzed that 41.65% is the minimum percentage necessary for manure collection for the client to recover the investment in both tanks and travel. Figure 3 shows that the percentage of manure collection needed to recover the investment in the storage and transport of biol is lower than that shown in Fig. 1.

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

167

Table 3 Estimates of biogas and biol production

Daily manure production per cow Daily manure production, considering all cows Amount of manure collected Hydraulic retention time Total amount of manure collected Manure volume Amount of water needed Substrate volume Amount of substrate in kg Estimation of biogas production per kg of manure Biogas production Biogas production in kg Biofertilizer production Biol production

Minimum (10 cows) 7.9 79

Maximum (15 cows) 7.9 118.5

39.5 43 1705 1.7 3.4 5.1 5115 0.016

59.25 43 2558 2.5 5.1 7.6 7673 0.016

27.7 33.33 5082 4.57

41.6 50 7623 6.86

Units kg manure/day kg manure/day kg manure/day Days Kg m3 m3 m3 Kg m3 biogas/kg manure m3 Kg Kg m3

Table 4 Estimation of pressure and dome thickness The molecular weight of biogas Number of moles of biogas Internal pressure of the biodigester Safety factor Dome thickness

Minimum (10 cows) 32.82 1015 178.7 2.5 3

Maximum (15 cows) 32.82 1523 342 2.5 5

Units kg/kmol mol kPa – mm

This is due to the increase in the number of cows present in the cattle, which causes an increase in the production and collection of manure. That is why collecting a lower percentage of all the manure produced is required to obtain the necessary amount to recover the investment for the commercialization of biol. At the time of completion of this project, in Ecuador, the price of the LPG tank of 15 [kg] was $1.60. Because a domestic application for biogas is sought, it is necessary to know how many LPG tanks can be saved. In Fig. 4, for a manure collection above 62.22%, a biogas production equivalent to an LPG tank is estimated, so it is considered that there is a saving of $1.60. Figure 5 shows the savings in LPG tanks for cattle of 15 cows. The most notable difference concerning Fig. 3 is the possibility of saving two LPG tanks if enough manure is collected. Additionally, Fig. 5 shows that a lower percentage of manure collection is required to generate enough biogas to replace a 15 [kg] LPG tank. Figure 6 shows that the lower the manure collection, the lower the production of biogas and, in turn, a greater volume available for storage, producing a lower internal pressure in the biodigester.

168

G. Anggie et al.

Fig. 2 Percentage of manure collection vs biol production and profit. For the critical case of analysis (10 cows)

Fig. 3 Percentage of manure collection vs Biol production and profit (15 cows)

It can also be observed that at 32.01% of manure collection, the minimum working pressure is obtained. It is then needed for the biogas to reach the customer’s kitchen at the established distance of 20 [m], setting the collection limit that allows the use of biogas in 43 days. Figure 7 shows that the variation in the percentage of manure collection necessary to achieve the working pressure for Fig. 4 is minimal; this difference is 1.04%. This indicates that the internal pressure does not increase significantly when the

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

169

Fig. 4 Percentage of manure collection vs. Savings in LPG tanks. For the critical case of analysis (10 cows)

Fig. 5 Manure collection percentage vs. Savings in LPG tanks (15 cows)

temperature rises. However, it should be emphasized that the increase in biogas production that occurs when the temperature of the substrate increases has not been considered. Figure 8 shows the behavior of biol production and internal pressure within the biodigester for cattle with 15 cows. The most notable difference concerning Fig. 5 is that, as the number of cows increases, the amount of manure produced increases

170

G. Anggie et al.

Fig. 6 Percentage of manure collection vs biogas production and internal pressure. For the critical case of analysis (10 cows and temperature of 21
C)

Fig. 7 Percentage of manure collection vs biogas production and internal pressure (10 cows and temperature of 33
C)

considerably so that the minimum harvesting percentage to achieve the working pressure is significantly reduced. Figure 9 also shows a reduction in the percentage of manure collection; however, this reduction is only 0.7% because only the temperature of the substrate has changed.

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

171

Fig. 8 Percentage of manure collection vs biogas production and internal pressure (15 cows and temperature of 21
C)

Fig. 9 Percentage of manure collection vs biogas production and internal pressure (15 cows and temperature of 33
C)

Figure 10 shows the minimum case of 10 cows, and the biogas production is insufficient to meet the demand for two LPG tanks of 15 [kg], whose biogas equivalence is 82.95 [kg]. Figure 11 shows that by increasing the number of cows present in cattle, enough biogas can be created to meet the demand for two 15 [kg] LPG tanks. Specifically, it is required to collect 82.96% of the manure produced by 15 cows to meet the demand.

172

G. Anggie et al.

Fig. 10 Percentage of manure collection vs Biogas production and consumption for the critical case of analysis (10 cows)

Fig. 11 Percentage of manure collection vs. Biogas production and consumption (15 cows)

It is known that the higher the percentage of manure collection, the greater the amount of biogas, so the internal pressure of the biodigester would increase. The dome would require a greater thickness to withstand this pressure without the material failing. Also, the internal pressure increases when the temperature inside the biodigester increases. Figure 12 shows a polynomial relationship between the percentage of manure collection and the minimum dome thickness necessary for the

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

173

Fig. 12 Manure collection percentage vs. Minimum thickness for the dome. For the critical case of analysis (15 cows and temperature of 33
C)

Fig. 13 Percentage of manure collection vs. Minimum thickness for the dome (10 cows and temperature of 21
C)

proper functioning of the biodigester, requiring greater thickness as the percentage of manure collection increases. In Fig. 13, you can see the minimum thickness of the dome necessary to withstand the internal pressure of the biodigester for ten cows with an internal temperature of 21 [
C], the one with the lowest pressure. In Fig. 14, it is observed that there is a relationship with Fig. 12 because both cases consider some cattle with 15 cows. However, the internal pressures are similar

174

G. Anggie et al.

Fig. 14 Percentage of manure collection vs. Minimum thickness for the dome (15 cows and temperature of 21
C)

Fig. 15 Percentage of manure collection vs. Minimum thickness for the dome (10 cows and temperature of 33
C)

at varied analysis temperatures in both cases. Hence, the minimum thickness values for the dome are also close in both cases. Figure 15 presents a function similar to that shown in Fig. 13 because both graphs are based on analyses carried out for some cattle with ten cows but differ in their working or operating temperature. The curves for minimum thicknesses for the dome have the same behavior as the curves of the internal pressure of the biodigester, and this is because both values are directly related.

Optimization of the Design of a Pilot Biogas Production Unit for Rural Areas

175

4 Conclusions For the analysis of the minimum case involving ten cows, the requisite minimum manure collection percentage to achieve a savings of $1.60, equivalent to the cost of an LPG tank, is 62.22%. Contrarily, in the maximum case involving 15 cows, the minimum manure collection percentage for an approximate profit of $1.60 stands at 41.43%, while 82.96% is necessary to realize an estimated gain of $3.20. Within the confines of the biodigester, the minimum design operation pressure hinges upon achieving a collection percentage of no less than 32.01%. However, to secure profits per unit of biol, the minimum percentage of manure collection must be 41.65%. Operating below this threshold results in transport and storage costs of biol surpassing the potential profit under the assumed conditions. The project’s comprehensive viability needs a minimum manure collection percentage of 62.22%. At this level, the generation of biogas is substantial enough to obviate the need for purchasing an LPG tank and facilitate the production of biol for commercial purposes. In the case of 10 cows, the biogas output falls short of fulfilling household consumption requirements, equivalent to two LPG tanks (82.95 kg of biogas). In contrast, the maximum case with 15 cows necessitates a minimum manure collection percentage of 82.96% to meet household consumption needs. As the percentage of manure collection increases, or temperatures rise, a thicker dome becomes imperative for the biodigester’s proper functioning. This adaptation is crucial due to the augmented biogas production within the biodigester, increasing internal pressure.

References 1. Whitelightskyes, El Clima en Balzar. https://whitelightskyes.com/administrative-area/4381561balzar/ (2022) 2. Salazar Abad, J.B., Arias Bonilla, J.L.: Diseño y construcción de un biodigestor para producción de biogás a partir de estiércol vacuno en la finca Isabel de la parroquia Taracoa, provincia de Orellana, Riobamba (2016) 3. Lara Guilcapi, M.F.: Diseño de un biodigestor para la producción de biogás generado por las excretas de ganado vacuno, en el criadero Jersey Chugllin, Riobamba (2016) 4. Coronel Durazno, D.A.: Valoración de estiércol bovino y porcino en la producción de biogás en un biodigestor de producción por etapas, Cuenca (2018) 5. Castillo Monar, R.U.: Diseño de un biodigestor para una finca del recinto San Luís de las Mercedes del Cantón Las Naves—Provincia de Bolívar, Guayaquil (2009) 6. Gordón Zuleta, J.E., Samaniego Manchay, J.A.: Diseño y construcción de un biodigestor chino anaerobio a partir del estiércol vacuno en la finca “Los 5 hermanos” de la parroquia el dorado, Riobamba (2014) 7. Barzallo Bravo, L.A.: Diseño, construcción y estandarización operativa de biodigestor anaerobio para finca productora de leche, Quito (2018) 8. Groppelli, E.S., Giampaoli, O.A.: El camino de la biodigestión: Ambiente y tecnología socialmente apropiada. Universidad Nacional del Litoral (2001)

176

G. Anggie et al.

9. Lojagas: HOJA DE SEGURIDAD DEL GAS LICUADO DE PETROLEO. http://lojagas.com/ nueva/wp-content/uploads/2017/08/G-99.-HOJA-DE-SEGURIDAD-GAS-LICUADO-DEPETROLEO.pdf (2017) 10. Moncayo Romero, G.: ¿Qué es el biogás? Uelzen (2017) 11. Aneja, V., Schlesinger, W., Erisman, J.W.: Farming pollution. Nat. Geosci., 409–411 (2015) 12. Baredar, P., Khare, V., Nema, S.: Chapter 2—Optimum sizing and modeling of biogas energy system. In: Design and Optimization of Biogas Energy Systems, pp. 33–78. Academic, London (2020) 13. Cotrina Lezama, R.F., Villanueva Vigo, G.: Biodigestores tubulares unifamiliares: Cartilla práctica para instalación, operación y mantenimiento. Block Grant, Lima-Perú (2015) 14. Deng, L., Liu, Y., Wang, W.: Rural Household Digesters. In: Biogas Technology, pp. 31–67. Springer, Singapore (2020) 15. Poudel, R.C.: Small scale biogas production. In: Biogas, Biofuel and Biorefinery Technologies, vol. 6, pp. 437–448. Springer, Cham (2018) 16. Castro, L., Escalante, H., Jaimes-Estévez, J., Díaz, L., Vecino, K., Rojas, G., Mantilla, L.: Low cost digester monitoring under realistic conditions: Rural use of biogas and digestate quality. Bioresour. Technol. 239, 311–317 (2017) 17. FAO: Manual de biogás, Santiago de Chile (2011) 18. Casanovas, G., Della Vecchia, F., Reymundo, F., Serafini, R.: Guía Teórico-Práctica sobre el biogás y los biodigestores. FAO, Buenos Aires (2019)

Techno-economic Analysis for the Valorization of Palm Kernel Shell via Hydrothermal Carbonization and Anaerobic Digestion Carolina Rueda, Sebastián Ponce, and Herman Murillo

1 Introduction Palm oil and palm kernel oil are vegetable oils widely used in the food industry. In 2016, 50 million tons of palm oil were produced globally; by 2022, it had reached 73 million tons [1]. In Ecuador, 470,000 tons of palm kernel oil were built in 2022, of which 58% were exported [2]. Extracting palm oil and palm kernel oil generates different types of waste, including solid residues such as the palm fruit stalk and kernel shell (PKS) [3]. Mainly, PKS, as residue, shows exciting properties, including porosity, moisture content, lignocellulose content, and carbon content. Nevertheless, nowadays, it is mainly used for decoration in gardening, or it is also disposed of in landfills or burned agriculturally, causing environmental pollution and wasting its potential for generating new products. Therefore, the valorization of palm kernel shells plays a significant role in promoting a circular economy. Hydrothermal carbonization (HTC) is a promising thermochemical conversion technology among gasification, pyrolysis, and torrefaction processes, primarily due to its lower temperature operation [4]. Namely, operating within a temperature range of 180–260 °C, HTC offers the advantage of utilizing lignocellulosic biomass as a viable feedstock option. In an HTC process, the biomass undergoes hydrolysis, where hemicellulose, cellulose, and lignin are partially decomposed. These decomposition products undergo dehydration, decarboxylation, and demethanation reactions, forming the solid product (i.e., hydrochar) through polymerization and aromatization reactions [5]. Various gases such as methane or carbon dioxide and a liquid phase called liquor or process water are also produced.

C. Rueda · S. Ponce · H. Murillo (✉) Universidad San Francisco de Quito, Quito, Ecuador e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3_11

177

178

C. Rueda et al.

The HTC processing of PKS for hydrochar production and the by-products mentioned above are available in the literature [6]. The most common application of this hydrochar is in the field of renewable energy used as a solid fuel [7]. Due to its mineral species content, it can also serve as a fertilizer for agricultural amendment. Another application is the adsorption of gases like carbon dioxide, which can contribute to environmental pollution. Furthermore, it can be used for water treatment, functioning as an adsorbent to remove or degrade contaminants present in water, such as dyes, heavy metals, and pharma [8–10]. On the other hand, a large quantity of process water is obtained containing organic compounds (10–64 gCOD/L) dissolved during HTC, making its recovery challenging. Thus, several research groups have demonstrated the possible valorization of HTC process water in anaerobic digestion (AD) for energy production [11, 12]. AD is a biological process where organic compounds are transformed into biogas without oxygen. Moreover, AD might contribute to the energy requirements of the hydrothermal plant, leading to a positive energy balance. In this sense, this work presents a techno-economic analysis evaluating a production plant for an adsorbent material through hydrothermal carbonization of palm oil kernel shell followed by steam activation. At the same time, the HTC spent liquors are coupled to biological treatment via anaerobic digestion, so biogas and digestate-based fertilizers can also be commercialized as main products. Namely, the three products to be offered to customers from this project are the hydrochar-based adsorbent, biogas, and digestate-based fertilizer. This HTC-AD approach can be considered a brief biorefinery concept that may contribute to the valorization of agroindustrial waste that is not adequately managed. Unlike most recent reports on techno-economic analysis concerning HTC that focus on hydrochar use as fuel [13–15], this study encompasses another useful application for hydrochar in adsorbent material production, specifically in Ecuador, where wastewater from different industries is a critical issue. As a first approach, the adsorption capacity of hydrochar for dye removal was evaluated toward wastewater treatment from the textile sector. Nevertheless, hydrochar has also been demonstrated to be used not only in textiles [16] but also in the adsorption of other pollutants such as pharmaceuticals [17], phenols [18], and metals [19].

2 Methodology 2.1

Process Design

Among thermochemical conversion techniques, hydrothermal carbonization emerges as a promising strategy for the valorization of lignocellulosic residues. Unlike traditional pyrolysis, a lower temperature range is required for HTC, as previously mentioned. The latter is one of the main reasons to consider HTC for this work, as it reduces the project energy demand. However, as water is a reactant, biomass decomposition products can migrate to a liquid phase that must be treated

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

179

Fig. 1 Summarized block diagram for the valorization of PKS via hydrothermal carbonization coupled with anaerobic digestion

instead of releasing these liquors to water bodies. Interestingly, the organic matter in HTC process water can be further valorized via anaerobic digestion so that this PKS-based biorefinery concept can obtain two new products. A brief block diagram is shown in Fig. 1 to summarize the proposed process in this work.

2.2

Mass Balance

The production flow rate of the plant was defined based on the available amount of raw material, mainly PKS. Considering the palm kernel oil production in Ecuador, it is determined that 41% and 6% by mass will be oil and PKS from the palm nut, respectively [20]. Accordingly, the generated palm kernel shells will be 14,000 tons annually. From this amount, it has been considered that the plant’s flow rate of palm kernel shells will be 10% of the shell production, resulting in 480 kg per hour. To perform the mass balance calculations, the following feed rates were considered: 480 kg of palm kernel shells, 5414 kg of water, 262.52 kg of methanogenic bacteria, and 65.53 kg of culture medium per hour. The process was determined to produce 524.85 kg of activated hydrochar, 2597.45 kg of biogas, and 10,751.35 kg of fertilizer per hour, respectively. The SuperPro Designer software V11, based on the vapor–liquid equilibrium model, was employed for the mass balances of both inoculation and fermentation processes. On the other hand, the HTC-related mass balances were conducted based on laboratory experimental results for the grinding, hydrothermal carbonization, filtration, and drying methods. To this end, a 500 mLhigh pressure reactor (model TGYF-B-500ML) was used for HTC experiments.

180

2.3

C. Rueda et al.

Data Collection for Hydrothermal Carbonization and Anaerobic Digestion

For the HTC process, various tests were conducted in the hydrothermal carbonization reactor at the USFQ laboratory. Different operating conditions were considered, focusing on the HTC temperature effect. Previously, ground palm kernel shells weighing 30 g and 300 g of water were added based on a determined PKS-towater ratio of 1:10 [21]. In contrast, a residence time of 60 min was established. The temperature levels investigated were 180 °C, 200 °C, 220 °C, and 240 °C, respectively. Subsequently, the products were filtered by vacuum filtration using Whatman filters (pore size: 11 μm, and filter diameter: 125 mm). The obtained hydrochar was dried for 24 h at 105 °C in a drying oven. Additionally, adsorption tests using methylene blue (MB) as a target pollutant were conducted to evaluate the adsorption capacity of each hydrochar. Fifty mL-batch experiments were carried out with an MB initial concentration of 0.05 mM, including a hydrochar addition of 2 g/L. About the AD process, simulations were conducted in the SuperPro Designer V11 software for the anaerobic digestion process of the liquors obtained from the HTC process, as aforementioned. The liquor composition, mainly acetic acid, formic acid, and fructose, was considered [22]. Moreover, the CAS registry numbers were identified for one of the most used methanogenic bacteria in AD and the culture medium, sodium hydrosulfite. It was also assumed that the process would be conducted at 40 °C and the liquor-to-inoculum ratio would be 1:2 [23]. This process was simulated using two digesters due to the reactions that occur during anaerobic digestion. The following reactions were considered for anaerobic digestion: Fructose þ Ammonia → Glutamic Acid þ CO2 þ H2 O

ð1Þ

Acetic Acid þ Ammonia → CH4 þ CO2 þ H2 O þ Biomass

ð2 Þ

Finally, this simulation enabled the determination of the necessary data for the anaerobic digestion process. Therefore, the selected methods for the production plant are suitable for obtaining the products above. Figure 2 shows the simulation of this process.

Fig. 2 Flow diagram scheme used for the SuperPro Designer simulation of the AD process

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

2.4

181

Equipment Sizing and Energy Demand Estimation

The plant equipment includes a mill, four agitated tanks for the HTC, activation, inoculation, and fermentation processes, and auxiliary equipment such as pumps, heat exchangers, separators, and mixers. For instance, the HTC reactor volume was determined according to the mass input and the feed density (based on the CHNO atomic contents of raw PKS), following Eq. (3). Once the volume was calculated, an oversizing factor of 0.15 was considered to keep the reactor operation safe. V HTC reactor = Σðmi =ρi Þ = mC =ρC þ mH =ρH þ mN =ρN þ mO =ρO

ð3Þ

Likewise, the activation tank volume was calculated based on mass and density values; however, the CHNO contents correspond to the hydrochar set to be activated. Again, an oversizing factor of 0.15 was deemed herein. On the other hand, the rest of the equipment sizing was carried out by considering the previous mass balance calculations and assisted by SuperPro Designer V11 simulations (please refer to Fig. 6 for an example showing the hydrochar drying process).

2.5

Economic Analysis

The cost estimation of the equipment entails defining the price, supplier, localization factor according to the country, the required quantity of equipment, construction material, and an element based on the selected material for each piece of equipment [24]. The price was multiplied by the mentioned factors, as well as the quantity, to determine the total cost of the equipment. The ISBL costs, which refer to construction and plant setup costs, were also calculated. Three different methods were considered for this purpose: the Lang Factor Method, the Hang Factor Method, and the Detailed Factorial Method (MFD). For the Lang Factor Method, different factors were considered depending on the type of processing, with a factor of 3 for solids, 5 for fluids such as liquids or gases, and 4 for a solid–fluid mixture. Therefore, the calculated total for each piece of equipment was multiplied by the corresponding factor. For the Hang Factor Method, a factor was considered depending on the type of equipment, such as compressors, heat exchangers, and pressure vessels. Each of these factors was multiplied by the previously calculated total equipment cost. Different factors for pipes, materials, and civil work, among others, were defined depending on the type of process: solids, fluids, or solid–fluid mixture. These factors were determined for each piece of equipment, and their sum reflects the detailed installation factor multiplied by the previously calculated total equipment cost. All the factors mentioned above were obtained from the literature [24]. Once the ISBL costs have been calculated, it is possible to estimate the fixed capital investment, representing the funds required for project implementation. This

182

C. Rueda et al.

involves the sum of ISBL, OSBL, engineering, construction, and contingency costs. The OSBL costs, encompassing plant infrastructure expenses, were calculated as 20% of the ISBL costs. The engineering and construction costs were determined as 5% of the previously estimated ISBL costs. Finally, the contingency costs were calculated as 10% of ISBL and OSBL costs [24]. To assess whether the project will be profitable, it is necessary to calculate the cash flow by subtracting expenses from revenues. Fees are calculated for each year, considering the values of fixed capital investment, fixed costs, variable costs, and working capital costs. For this project, the profitability indicators calculated are the Net Present Value (NPV), Internal Rate of Return (IRR), and Payback Period (PP). Finally, a brief sensitivity analysis was also provided.

3 Results and Discussion 3.1

Process Design

Figure 3 shows the flow diagram for the proposed process, where the lines and arrows represent the process streams, and the numbers on them correspond to the code indicating the composition of each stream. Detailed information about the equipment and stream codes can be found in Tables 8 and 9. The process begins with feeding the palm kernel shells into the mill to reduce their particle size. The ground shells and water are then introduced into the hydrothermal carbonization reactor, operating at subcritical conditions with temperatures

Fig. 3 Proposed flow diagram for the entire valorization of palm kernel shell via hydrothermal carbonization

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

183

and pressures below the critical point of water. The PKS-to-water ratio is 1:10 [5]. The output of this process passes through a filter to separate the hydrochar from the liquors. The wet hydrochar is then dried in a dryer at 180 °C. The dried hydrochar is transferred to an activation tank, where water vapor is used at a hydrochar-to-vapor ratio of 1:2 and a temperature of 600 °C for activation [25]. Meanwhile, the spent liquors are separated and transferred to the inoculation reactor, where methanogenic bacteria and sodium hydrosulfite medium are added as the culture medium. The inoculum is then fed into the anaerobic digestion process, entering the first fermentation reactor and mixing with another liquor stream. In this reactor, fructose is transformed into acetic acid. The biogas and digestate products are mixed and transferred to the second fermentation reactor, where acetic acid fermentation produces methane and other gases, generating biogas and digestate that can be utilized as fertilizer [26]. Therefore, the required process streams and equipment, such as pumps and valves, among others, and the mass balance have been determined. Additionally, data from the literature were considered for the activation process, such as a hydrochar-to-vapor ratio of 1:2 and a yield of 56.9% for activated hydrochar [25]. Thus, the quantities of the products obtained were quantified based on the amount of raw material, and the results are presented in Table 10. A slight decrease in hydrochar yields was observed with temperature since mass yields of 65.6%, 62%, 57%, and 55.1% were obtained concerning starting PKS for each temperature level, respectively. This means that more biomass constituents are decomposed at higher temperatures. It is also known that at higher temperatures, since more volatile compounds are released from hydrochar, a porous structure is obtained, which is reflected by the improvement in the hydrochar surface area [5]. This is thus favorable for dye adsorption.

3.2

Adsorption Capacity Tests

The hydrochar obtained at 240 °C showed the highest adsorption capacity after assessing the amount of MB adsorbed with time (qt: in mg of MB retained per g of adsorbent) as indicated in Fig. 4. It can be attributed to a higher surface area and oxygenated functional groups provided to hydrochar at higher HTC temperatures, which promotes the adsorption of different pollutants including dyes [27]. Unlike the minor disparities in mass yield observed earlier, a significant variation in adsorption capacity at 240 °C becomes apparent. This disparity holds crucial implications for enhancing adsorption efficiency while maintaining a cost-effective hydrochar production approach suitable for extensive industrial applications.

184

C. Rueda et al.

Fig. 4 qt vs. t plot for MB adsorption on hydrochar obtained at different temperatures

Table 1 Capacity and energy demand for simulated equipment

3.3

Equipment Mill Filter Dryer Digestion tank 1 Digestion tank 2 Pump Heat exchanger Mixer

Capacity 480 10.32 11.58 9764.73 9924.39 4825.02 0.05 4449.6

Units kg/h m2 m2 L L L/h m2 kg/h

Energy demand (kW) 58.6 – 49.42 89.55 97.65 0.19 80.4 –

Equipment Sizing and Energy Demand Estimation

The construction material for each piece of equipment was chosen based on the material it would handle. Carbon steel was considered for the mill and pump, while stainless steel was selected for all other equipment. The sizing of most equipment was determined using simulation in SuperPro Designer V11. However, the sizing of the HTC and activation tanks was done considering their operating conditions. These tanks require an agitator and a heating jacket to achieve turbulent flow, and the activation tank needs to maintain a temperature of 500 °C [28]. To assess the energy requirements of the process, energy balances were conducted for each piece of equipment within the plant. These balances were obtained through simulation in SuperPro Designer V11 for most of the used equipment. However, for the HTC reactor, the activation energy of the reaction was considered. For the activation tank, the energy used in the process was considered. These energies were multiplied by the respective input flow rate. The activation energy for the HTC process was estimated at 4212 kJ/kg [29]. The energy used in the activation process was estimated at 5680 kJ/kg [25]. Energy demand values are also summarized in Table 1.

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

3.4

185

Economic Analysis

Cost Estimation Equipment Therefore, the ISBL costs for each method were USD 20.2 million for the Lang Factor Method, USD 18.2 million for the Hang Factor Method, and USD 12.4 million for the MFD. This information is summarized in Table 2 (detailed information is available in Table 11).

Production Cost The production cost estimation involves the sum of both fixed and variable costs. Fixed costs represent the expenses incurred by the company regardless of the plant’s production quantity. These costs include worker and supervisor wages, plant maintenance, property taxes and insurance, land rental, environmental permits, and a miscellaneous cost category encompassing other relevant expenses. The sum of these costs amounts to approximately USD 1.7 million. On the other hand, variable costs are dependent on the production quantity and may vary accordingly. This estimation considers the total costs of raw materials, electricity consumption costs derived from energy balances, and the fuel required for generating the necessary steam in the activation process, which is diesel. These costs are up to USD 68.6 million. Notice that raw material costs are included in variable costs. The raw materials are PKS and water for the HTC process, water vapor for activation, methanogenic bacteria, and HS medium (sodium hydrosulfite medium) that serves as the culture medium for bacterial reproduction, aiding in the creation of Table 2 ISBL costs are based on different methodologies Equipment Mill Pump HTC reactor Filter Heat exchanger Dryer Activation tank Incubation reactor Mixer/separator AD reactor Total (USD)

Lang factor (USD) 484,500.00 95,000.00 7,538,440.00 266,760.00 222,300.00 1,022,580.00 7,538,440.00 1,003,017.60 14,523.60 2,004,009.80 20,189,571.00

Hang factor (USD) 403,750.00 76,000.00 7,538,440.00 166,725.00 155,610.00 852,150.00 7,538,440.00 501,508.80 7261.80 1,002,004.90 18,241,890.50

Detailed factorial (USD) 242,250.00 32,300.00 4,975,370.40 146,718.00 86,252.40 511,290.00 4,975,370.40 489,472.59 5635.16 977,956.78 12,442,615.73

186 Table 3 Summary of the production cost

C. Rueda et al.

Fixed costs Workforce Supervision Maintenance Taxes/insurance Land rent Others Environmental licenses Total (USD) Variable costs Raw materials Electricity Fuel for producing steam Total (USD) Production cost (fixed + variable) (USD)

555,750.00 138,937.50 511,391.51 170,463.84 139,357.30 85,231.92 136,371.07 1,737,503.12 4,376,605.79 64,086,815.52 16,6033.80 68,629,455.11 70,366,958.23

the inoculum for anaerobic digestion of liquors derived from the HTC process, as shown in Table 12. The PKS can be obtained from palm kernel oil extractors in Ecuador, as it is a by-product. Its composition comprises 26% hemicellulose, 7% cellulose, and 54% lignin [30]. Water and water vapor are auxiliary services provided to the plant. The bacteria and culture medium obtain the necessary inoculum for anaerobic digestion. These products are sourced from industrial suppliers in the United States and China. Consequently, the production costs are obtained by combining the fixed and variable costs, resulting in an approximate total of $70.3 million, as indicated in Table 3. It should be noted that these costs will aid in estimating the working capital. Another essential aspect that has been considered is the location. It has been determined that the hydrothermal carbonization plant in Ecuador will be located in Santo Domingo. This city is characterized by suitable climatic conditions for producing African palm, including the extraction of palm oil and palm kernel oil [31]. Therefore, the quantity of palm kernel shell waste required for the process can be found in this area.

Fixed Capital Investment and Work Capital Estimation All these costs were summed up to obtain the fixed capital investment results, which amount to $25.5 million for the Lang factor method, $24.8 million for the Hang factor method, and $17.05 million for the MFD method. Therefore, the subsequent calculations and estimations were conducted considering the MFD method. (Please refer to Table 4 for further information). Regarding the estimation of working capital, the following factors were considered: raw material cost for 2 weeks of production, the production cost for 2 weeks to estimate product costs, the production cost for 1 week to calculate cash in hand, raw

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

187

Table 4 Fixed capital investment details Costs ISBL OSBL Design and engineering Contingency Total (USD) Table 5 Selling prices

Lang factor (USD) 17,204,304.30 5,161,291.29 860,215.22 2,236,559.56 25,462,370.36

Product Adsorbent Biogas Fertilizer

Hang factor (USD) 16,749,257.15 5,024,777.15 837,462.86 2,177,403.43 24,788,900.58

Detailed factorial (USD) 12,442,615.73 2,488,523.15 622,130.79 1,493,113.89 17,046,383.55

Selling price (USD/kg) 4.00 1.11 6.00

material cost for 4 weeks of production to estimate credit costs and spare parts cost equal to 1.1% of the sum of ISBL and OSBL costs. These factors resulted in a working capital cost of $4.3 million.

Profitability Assessment Project Cash Flow Cash flow refers to the amount of money generated in the project, both inflows and outflows, over the years. It is calculated by considering the annual revenues and expenses caused by the company, as the difference between revenues and expenses yields the cash flow. This is used to assess the feasibility of project implementation. Gross revenues are determined based on production costs and product prices to calculate the project’s net income (selling prices are detailed in Table 5). The net income is then obtained by subtracting 12% of taxes from the gross revenues. This calculation is performed for each product over 5 years, and the net gains received for each year are summed to determine the net income of the plant. Additionally, the cumulative cash flow for 15 years is determined. For this, the cash flow and incremental cash flow for the first year are assumed to be the same. From the second year onward, the cumulative cash flow is calculated as the sum of the incremental cash flow from the previous year plus the cash flow for the current year. Finally, the results obtained can be observed in Fig. 5. These results will help determine the net present value, internal rate of return, and investment payback period. It should be noted that negative cash flows indicate that the cash outflow is greater than the cash inflow, while positive values indicate the opposite. Additionally, in the figure, it can be observed that the cash flows will be harmful for the first 8 years. However, from the ninth year onward, they become positive, resulting in a payback period of approximately 9 years. Notice that the project’s first 2 years

188

C. Rueda et al.

Fig. 5 Cash flow chart

correspond to the time range necessary to set the project up. In other words, the time required for plant construction, equipment installation, etc. It means the products (adsorbent, biogas, and fertilizer) will be available in the market from the third year.

Profitability Indicators NPV is used to assess the business’s profitability, as it represents the sum of all future cash flows of the project. To calculate NPV, cash flows and a discount factor based on the Minimum Acceptable Rate of Return (MARR), also known as the discount rate, are considered [32]. Once both elements are defined, their product determines the present value for each of the estimated years, and the sum of all the calculated values gives the NPV of the project. Additionally, it is essential to note that the interpretation of NPV values for a project is as follows: a positive value indicates that the cash inflows are more significant than the outflows, indicating a good investment with profits. A negative value indicates the opposite situation, and if NPV is equal to zero, it means there are no losses or gains in the project. This project’s NPV of USD 89.5 million indicates profitability and expected returns. Once NPV is determined, it is possible to calculate the IRR. This measure helps determine the potential profitability of investments in the project. It is the discount rate at which the NPV of all cash flows equals zero. A higher IRR indicates a better investment option if it is greater than the MARR. In this case, the project is considered profitable. For this project, with an MARR of 20%, the calculated IRR is 59%, indicating that the project will be beneficial. Another important indicator for the project is the PP, which estimates the time it takes to recover the investment cost. This is crucial for investors who know when they will recoup their money [24]. Therefore, the shorter the payback period, the more attractive the project is to investors. In this project, a PP of 8.7 years has been

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . . Table 6 Profitability indicators

NPV MARR IRR PP

189

89.5 million USD 20% 59% 8.7 years

Table 7 Raw materials’ price changes Raw material PKS Water Bacteria Culture medium Fuel for steam production

Initial prices (USD/kg) 0.28 0.00124 32.00 201.56 0.75

Maximum prices (USD/kg) 4.50 0.01984 498.50 3124.25 12.00

determined, primarily due to the significant amount of money required for project implementation, considering the high costs of the adsorbent, biogas, and fertilizer production plant equipment. These indicators are summarized in Table 6. While the project appears to yield profits over 8 years, its viability could face challenges within Ecuador. The persisting cycle of protests and social upheaval in the country introduces a notable concern due to the market’s and prices’ inherent volatility. As a general guideline, it’s advisable to prioritize projects with payback periods of less than 5 years in such circumstances.

Sensitivity Analysis For this section, two different scenarios have been considered. The first scenario involves varying only the selling prices of the products until an NPV of zero is obtained. The second scenario involves altering the raw materials’ prices until an NPV of zero is achieved. This analysis aims to determine the products’ minimum selling price and the raw materials’ maximum purchase price. In the first scenario, the prices of the products varied from USD 4.00/kg to USD 14.60/kg for the adsorbent, USD 1.11/kg to USD 5.45/kg for the biogas, and USD 6.00/kg to USD 23.15/kg for the fertilizer. This means the maximum selling prices must be lower than the new prices mentioned, as these are the price limits to ensure the project is profitable, namely, maintaining a positive NPV. On the other hand, considering the second scenario, where the prices of the raw materials are changed, the prices varied, as shown in Table 7. In this latter table, the first column displays the initial prices. In contrast, the second column represents the adjustments made for the flexibility analysis, showing the maximum prices for acquiring raw materials to achieve a net present value of zero. These prices indicate the limits one would pay to ensure the project’s profitability.

190

C. Rueda et al.

The study’s outcomes highlight the crucial need to focus on the availability of raw biomass. It is imperative to consider the impact of natural occurrences like “El Niño” and dry seasons, as these factors can significantly influence the availability and prices of biomass resources.

4 Conclusion This work deals with the valorization of agro-waste, specifically, palm kernel shell, a biomass residue from the palm oil extraction industry. Thermochemical conversion of PKS was conducted via HTC, which converts biomass in the presence of subcritical water to attain hydrochar-based adsorbent materials. A brief biorefinery concept was developed since the spent liquors from hydrothermal processing were integrated via biological valorization (i.e., anaerobic digestion) so that biogas and digestate-based fertilizer can also be produced and commercialized alongside hydrochar. From laboratory-scale experiments, the feasibility of using hydrochar as an adsorbent material was determined after performing batch experiments for methylene blue removal from aqueous solutions. However, the plant design included an activation process for hydrochar, so its adsorption capacity can be further improved on an industrial scale, including steam activation. This physical activation was conducted instead of chemical to avoid releasing alkaline or acid wastewater (e.g. when activating with KOH or HCl). On the other hand, the biological treatment of HTC spent liquors was simulated using SuperPro Designer V11 to determine biogas and fertilizer production. A techno-economic analysis to process 10% of the PKS production in Ecuador was performed as a first approach. The plant is intended to be in Santo Domingo de los Tsachilas Province, which is a location where most of the PKS is generated. The process design entails the production of 524.85 kg of activated hydrochar, 2597.45 kg of biogas, and 10,751.35 kg of fertilizer per hour, including selling prices of USD 4.00, USD 1.11, and USD 6.00 per kg, respectively. Based on these results, a positive NPV and an IRR higher than the interest rate proposed by financial institutions reflect that the project is profitable, with a Payback Period of 8.7 years. In conclusion, this project reflects that the proposed biorefinery concept for the PKS valorization can remarkably contribute to the circular economy trend, which seeks the conversion of residual biomass to value-added products, especially those residues from the agricultural sector that represent one of the most important economic activities in Ecuador. Therefore, due to the significant amount of agrowaste available in the country, valorization strategies must be developed to avoid inefficient disposal of these residues in landfills or, worse, on stubble burning.

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

191

5 Supporting Information (Fig. 6 and Tables 8, 9, 10, 11 and 12)

Fig. 6 SuperPro Designer simulation for the hydrochar drying process

Table 8 Equipment codes related to the process flow diagram are indicated in Fig. 3 Equipment code DR-101 F-101 GR-101 HE-101 HE-102 MX-101 R-101 R-102 R-103 R-104 R-105 S-101

Equipment name Dryer Filter Mill Heat exchanger Heat exchanger Mixer HTC batch reactor Tank Incubation reactor First AD reactor Second AD reactor Separator

Operating conditions 105 °C, 1 bar 25 °C, 1 atm 25 °C, 1 atm 90 °C 90 °C 36 °C, 1 bar 240 °C, 1 MPa 600 °C 36 °C, 1 bar 36 °C, 1 bar 36 °C, 1 bar 25 °C, 1 bar

Reference SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation [5] [13] SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation SuperPro Designer simulation

192 Table 9 Stream codes related to the process flow diagram indicated in Fig. 3

C. Rueda et al. Stream code 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137

Description PKS fresh feed Milled PKS Water feed Hydrochar/HTC spent liquor mixture Gas exit from R-101 Water input Water output Hydrochar HTC spent liquor from R-101 Air Steam input to HE-101 Hot air Steam output from HE-101 Dry hydrochar Hot air exit Water Steam input to HE-102 Steam Steam output from HE-102 Hydrochar-based adsorbent Gas phase from R-102 Water input Water output Methanogenic bacteria Culture medium (hydrosulfite) Inoculum Water input Water output Digestate from R-104 Gas phase from R-104 Water input Water output Gas/digestate mixture Fertilizer Biogas Water input Water output

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

193

Table 10 Mass balance results for all the processes involved (data are expressed in kg per batch) Input Milling Raw PKS Hydrothermal carbonization Milled PKS Water Filtration Hydrochar/liquor mixture Drying Wet hydrochar Air Activation Dry hydrochar Steam Anaerobic digestion (first stage) Liquor Acetic acid Ammonia Biomass CO2 Formic acid Fructose Methane Water Inoculum

Output 480

Milled PKS

480

480 4800

Hydrochar/liquor mixture Gas phase

5003.86 276.14

5003.86

Wet hydrochar Liquor

554.26 4449.60

554.26 4622.62

Dry hydrochar Gas phase

307.47 4869.41

307.47 614.94

Activated hydrochar Effluent

524.85 397.56

4449.60 911.72 177.98 177.98 836.52 662.99 756.88 836.52 88.99 8899.2

Biogas CO2 Methane Digestate Acetic acid Ammonia Biomass CO2 Formic acid Fructose Methane Water

1922.23 598.20 1324.03 11426.57 2735.52 53.70 609.04 2391.58 1989.37 1106.09 1985.94 555.33

Biogas CO2 Methane Digestate Acetic acid Ammonia Biomass CO2 Formic acid Fructose Methane Water

2597.45 604.17 1993.28 10751.35 2705.04 1.08 625.73 2416.90 1989.00 1117.07 1327.79 568.75

Anaerobic digestion (second stage) Biogas 1922.23 598.20 CO2 Methane 1324.03 Digestate 11,426.57 Acetic acid 2735.52 Ammonia 53.70 Biomass 609.04 2391.58 CO2 Formic acid 1989.37 Fructose 1106.09 Methane 1985.94 555.33 Water

Equipment Mill Pump HTC reactor Filter Heat exchanger Dryer Activation tank Incubation reactor Mixer/separator AD reactor

Value (USD) 85,000 10,000 763,000 27,000 9000 138,000 763,000 81,216 588 81,134

Table 11 Equipment costs

Supplier country China China China China China China China China China China

Localization factor 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9

Qty 1 1 1 1 2 1 1 1 2 2

Construction material Carbon steel Carbon steel Stainless steel Stainless steel Stainless steel Stainless steel Stainless steel Stainless steel Stainless steel Stainless steel

Material factor 1 1 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 Total (USD)

Total (USD) 161,500 19,000 1,884,610 66,690 44,460 340,860 1,884,610 200,603.52 2904.72 400,801.96 5,006,040.20

194 C. Rueda et al.

Techno-economic Analysis for the Valorization of Palm Kernel Shell. . .

195

Table 12 Raw materials and their corresponding prices Raw material PKS Water Bacteria HS medium

Quantity (kg/h) 480 5414.94 262.52 65.63

Price (USD/kg) 0.28 0.0012 32.00 201.56

Annual cost (USD/year) 48384.00 2417.23 1680128.00 2645676.56

Supplier country Ecuador Ecuador China USA

References 1. Producción Agrícola Mundial H Producción Mundial Aceite de Palma 2022/2023: http://www. produccionagricolamundial.com/cultivos/aceitedepalma.aspx (2023). Accessed 30 Apr 2023 2. Ministerio de Comercio Exterior, Informe Sobre el Sector Palmicultor Ecuatoriano 2017 (2019.): https://www.produccion.gob.ec/wp-content/uploads/2019/06/informe-palmaespañol-.pdf. Accessed 12 May 2023 3. Paucar, W., Rubio, R.: Caracterización físico-química del cuesco y fibra obtenidos del procesamiento de palma africana para un aprovechamiento eficiente de la energía térmica en calderas. Rev. Perspect. 6, 110–126 (2021) 4. González-Arias, J., Sánchez, M.E., Cara-Jiménez, J., Baena-Moreno, F.M., Zhang, Z.: Hydrothermal carbonization of biomass and waste: a review. Environ. Chem. Lett. 20, 211–221 (2022) 5. Heidari, M., Dutta, A., Acharya, B., Mahmud, S.: A review of the current knowledge and challenges of hydrothermal carbonization for biomass conversion. J. Energy Inst. 92, 1779–1799 (2019) 6. Hammud, H.H., Karnati, R.K., Al Shafee, M., Fawaz, Y., Holail, H.: Activated hydrochar from palm leaves as efficient lead adsorbent. Chem. Eng. Commun. 208, 197–209 (2019) 7. Ameen, M., Zamri, N.M., May, S.T., Azizan, M.T., Aqsha, A., Sabzoi, N., Sher, F.: Effect of acid catalysts on hydrothermal carbonization of Malaysian oil palm residues (leaves, fronds, and shells) for hydrochar production. Biomass Conv. Bioref. 12, 103–114 (2022) 8. Kimbi Yaah, V.B., Zbair, M., Botelho de Oliveira, S., Ojala, S.: Hydrochar-derived adsorbent for the removal of diclofenac from aqueous solution. Nanotechnol. Environ. Eng. 6, 1–12 (2021) 9. Yek, P.N.Y., Liew, R.K., Wan Mahari, W.A., Peng, W., Sonne, C., Kong, S.H., Tabatabaei, M., Aghbashlo, M., Park, Y.K., Lam, S.S.: Production of value-added hydrochar from single-mode microwave hydrothermal carbonization of oil palm waste for de-chlorination of domestic water. Sci. Total Environ. 833, 154968 (2022) 10. Fagnani, H.M.C., da Silva, C.T.P., Pereira, M.M., Rinaldi, A.W., Arroyo, P.A., de Barros, M.A. S.D.: CO2 adsorption in hydrochar produced from waste biomass. SN Appl. Sci. 1, 1–10 (2019) 11. Merzari, F., Langone, M., Andreottola, G., Fiori, L.: Methane production from process water of sewage sludge hydrothermal carbonization. A review. Valorising sludge through hydrothermal carbonization. Crit. Rev. Environ. Sci. Technol. 49, 947–988 (2019) 12. Langone, M., Basso, D.: Process waters from hydrothermal carbonization of sludge: characteristics and possible valorization pathways. Int. J. Environ. Res. Public Health. 17(18), 6618 (2020) 13. Yan, M., Liu, Y., Song, Y., Xu, A., Zhu, G., Jiang, J., Hantoko, D.: Comprehensive experimental study on energy conversion of household kitchen waste via integrated hydrothermal carbonization and supercritical water gasification. Energy. 242, 123054 (2022) 14. Sangaré, D., Moscosa-Santillan, M., Aragón Piña, A., Bostyn, S., Belandria, V., Gökalp, I.: Hydrothermal carbonization of biomass: experimental study, energy balance, process simulation, design, and techno-economic analysis. Biomass Conv. Bioref. 1, 1–16 (2022)

196

C. Rueda et al.

15. Carrasco, S., Pino-Cortés, E., Barra-Marín, A., Fierro-Gallegos, A., León, M.: Use of hydrochar produced by hydrothermal carbonization of lignocellulosic biomass for thermal power plants in Chile: a techno-economic and environmental study. Sustainability. 14, 8041 (2022) 16. Tran, T.H., Le, A.H., Pham, T.H., Nguyen, D.T., Chang, S.W., Chung, W.J., Nguyen, D.D.: Adsorption isotherms and kinetic modeling of methylene blue dye onto a carbonaceous hydrochar adsorbent derived from coffee husk waste. Sci. Total Environ. 725, 138325 (2020) 17. Escudero-Curiel, S., Pazos, M., Sanromán, A.: Facile one-step synthesis of a versatile nitrogendoped hydrochar from olive oil production waste, “alperujo”, for removing pharmaceuticals from wastewater. Environ. Pollut. 330, 121751 (2023) 18. Pei, T., Shi, F., Liu, C., Lu, Y., Lin, X., Hou, D., Yang, S., Li, J., Zheng, Z., Zheng, Y.: Bamboo-derived nitrogen-doping magnetic porous hydrochar coactivated by K2FeO4 and CaCO3 for phenol removal: governing factors and mechanisms. Environ. Pollut. 331, 121871 (2023) 19. Navas-Cárdenas, C., Caetano, M., Endara, D., Jiménez, R., Lozada, A.B., Manangón, L.E., Navarrete, A., Reinoso, C., Sommer-Márquez, A.E., Villasana, Y.: The role of oxygenated functional groups on cadmium removal using pyrochar and hydrochar derived from Guadua angustifolia residues. Water. 15, 525 (2023) 20. Sánchez Alfonso, R.A., Durán Peralta, H.A., Aguiar Urriago, L.M., Uribe Aldana, N., Rojas Forero, A.Y.V.: Modelo para la gasificación del cuesco de palma aceitera. Ingenium. 18, 81–100 (2017) 21. Rodriguez Correa, C., Hehr, T., Voglhuber-Slavinsky, A., Rauscher, Y., Kruse, A.: Pyrolysis vs. hydrothermal carbonization: understanding the effect of biomass structural components and inorganic compounds on the char properties. J. Anal. Appl. Pyrolysis. 140, 137–147 (2019) 22. Hoekman, S.K., Broch, A., Robbins, C., Zielinska, B., Felix, L.: Hydrothermal carbonization (HTC) of selected woody and herbaceous biomass feedstocks. Biomass Con. Bioref. 32(3), 113–126 (2012) 23. Murillo, H.A., Pagés-Díaz, J., Díaz-Robles, L.A., Vallejo, F., Huiliñir, C.: Valorization of oat husk by hydrothermal carbonization: optimization of process parameters and anaerobic digestion of spent liquors. Bioresour. Technol. 343, 126112 (2022) 24. Towler, G., Sinnot, R.: Chemical Engineering Design: Principles, Practice and Economics of Plant and Process Design. Butterworth-Heinemann, Amsterdam (2008) 25. Azargohar, R., Dalai, A.K.: Steam and KOH activation of biochar: experimental and modeling studies. Microporous Mesoporous Mater. 110, 413–421 (2008) 26. Harun, N., Othman, N.A., Zaki, N.A., Mat Rasul, N.A., Samah, R.A., Hashim, H.: Simulation of anaerobic digestion for biogas production from food waste using SuperPro Designer. Mater. Today Proc. 19, 1315–1320 (2019) 27. Jain, A., Balasubramanian, R., Srinivasan, M.P.: Hydrothermal conversion of biomass waste to activated carbon with high porosity: a review. Chem. Eng. J. 283, 789–805 (2016) 28. Cancino, H.: Diseño de reactor, producción y caracterización de carbón activado de cáscaras de nuez para uso en separación de cianuros metálicos. Pontificia Universidad Católica de Valparaíso, Valparaíso (2011) 29. Nizamuddin, S., Jayakumar, N.S., Sahu, J.N., Ganesan, P., Bhutto, A.W., Mubarak, N.M.: Hydrothermal carbonization of oil palm shell. Korean J. Chem. Eng. 32, 1789–1797 (2015) 30. Ikumapayi, O.M., Akinlabi, E.T.: Composition, characteristics and socioeconomic benefits of palm kernel shell exploitation-an overview. J. Environ. Sci. Technol. 11, 220–232 (2018) 31. CONGOPE: Estrategia de cambio climático de la provincia de Santo Domingo de los Tsáchilas con enfoque de género. Proyecto Acción Provincial frente al Cambio Climático, Quito (2019) 32. Urbina, G.: Evaluación de proyectos, 6th edn. Instituto Politécnico Nacional, México D.F (2010)

Index

A Airfoil profile, 48, 88, 89 Anaerobic digestion (AD), 163, 164, 178–180, 183, 185, 186, 190, 191, 193, 194 Angle of attack, 43–47, 50, 52–55, 74–79 Ansys Fluent, 48, 88, 89

B Biodigesters, 163–167, 169, 172–175 Biogas, 102, 163–172, 175, 178, 179, 183, 187–190, 192, 193 Biol, 163, 166–169, 175

C Colombia, 88, 91–94, 100, 122, 123, 136, 139–142, 145, 157 Curved buildings, 74, 80 CYME, 6, 11, 15

D Demand management systems, 5 Developing countries, 24, 141, 157

E Economic analyses, 59, 134, 181–182, 185–190 Electric vehicles, 3, 14–16, 18 Energy efficiency, 3, 24, 26, 28, 43, 58, 71, 100, 122, 129, 130

Energy sources, 4, 6, 24, 43, 57, 58, 99–101, 140 Energy system modeling, 122 Environmental impact assessment (EIA), 140–141, 145–148, 157

G Geographic information systems (GIS), 25, 26, 30, 32 Geothermal heat pump (GHP), 121, 122, 134–136 Green hydrogen, 100–102, 104–112, 114, 115, 117, 118 Grid connected, 57, 95

H H Type Darrieus, 88–90, 92, 94, 95 Hydrochars, 177–181, 183, 184, 190–193 Hydrokinetic turbines (HKT), 94, 140–143, 145–157 Hydrothermal carbonization (HTC), 177–186, 190–194

I Industrial hydrogen, 100, 107, 115

K K-means, 6–9, 11, 13

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024 M. Espinoza-Andaluz et al. (eds.), Congress on Research, Development, and Innovation in Renewable Energies, Green Energy and Technology, https://doi.org/10.1007/978-3-031-52171-3

197

198 N NACA 1412, 44, 45, 48–55

P Particle Swarm Optimization (PSO), 5, 6, 9 Photovoltaic (PV) systems, 57 Ponds, 123, 136 Power generation, 99, 103, 105–107, 118, 140–142, 145, 153, 155 Process design, 178–179, 182–183, 190

Index Self-starting, 88, 93, 95 Soils, 130, 131, 133, 135, 137, 146, 148–156, 163

T Temperatures, 23, 62, 93, 102, 114, 121–123, 125–137, 152, 164, 168–175, 177, 178, 180, 182–184

U Urban environments, 59, 71, 72, 88 R Renewable energies, 4–6, 24, 43, 57, 61, 95, 99, 100, 103, 112, 115, 139, 140, 150, 178

S Self-consumption, 57–61, 63–66

W Wind energy, 71, 115 Wind power, 71, 74, 81, 82, 102, 103