DIREC TIONS IN DE VELOPMENT Energy and Mining Shedding Light on Electricity Utilities in the Middle East and North Africa Insights from a Performance Diagnostic Daniel Camos, Robert Bacon, Antonio Estache, and Mohamad M. Hamid Shedding Light on Electricity Utilities in the Middle East and North Africa DIREC TIONS IN DE VELOPMENT Energy and Mining Shedding Light on Electricity Utilities in the Middle East and North Africa Insights from a Performance Diagnostic Daniel Camos, Robert Bacon, Antonio Estache, and Mohamad M. Hamid © 2018 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 21 20 19 18 This work is a product of the staff of The World Bank with external contributions. The findings, interpreta- tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Camos, Daniel, Robert Bacon, Antonio Estache, and Mohamad M. Hamid. 2018. Shedding Light on Electricity Utilities in the Middle East and North Africa: Insights from a Performance Diagnostic. Directions in Development. Washington, DC: World Bank. doi:10.1596/978-1-4648-1182-1. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content con- tained within the work. The World Bank therefore does not warrant that the use of any third-party- owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1182-1 ISBN (electronic): 978-1-4648-1183-8 DOI: 10/1596/978-1-4648-1182-1 Cover photo: © Anna Om / Shutterstock. Night cityscape in Lebanon. Used with the permission. Further permission required for reuse. Cover design: Debra Naylor, Naylor Design, Washington, DC Library of Congress Cataloging-in-Publication Data have been requested. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Contents Acknowledgments xv About the Authors xvii Abbreviations xix Introduction 1 The Region’s Electricity Challenge 1 The New MENA Electricity Database 2 The Structure and Content of the Report 4 Notes 6 References 6 Part 1 How Do MENA’s Electricity Utilities Perform? 9 Chapter 1 Quasi-Fiscal Deficits in MENA’s Power Sector 11 Economy-Level Results 12 Utility-Level Results 15 What Can Be Done about Underpricing in MENA Economies? 22 Conclusion 25 Notes 26 References 26 Chapter 2 Comparing the Region’s Performance with the Rest of the World 29 Summary of Results and Overall Assessment 31 Detailed Comparisons for Selected Indicators 32 Conclusion 55 Notes 55 Shedding Light on Electricity Utilities in the Middle East and North Africa   v   http://dx.doi.org/10.1596/978-1-4648-1182-1 vi Contents Chapter 3 A Dynamic Look at MENA Performance over Five Years 57 Data Challenges 57 Indicator Trends with All Utilities Aggregated 59 Indicator Trends Disaggregated by Utility Type 60 Conclusion 62 Notes 63 Chapter 4 A Multi-Indicator Approach to Analyzing Utility Performance 65 Methodology 65 Data Considerations 66 Distribution Utilities: Average Rank Score 68 Generation Utilities: Average Rank Score 70 Vertically Integrated Utilities: Average Rank Score 71 Conclusion 72 Notes 72 Chapter 5 Drivers of Utility Performance: Institutional and Contextual Characteristics 73 Potential Determinants of Utility Performance 75 Summary of Results 78 Statistically Significant Differences between Subgroups of Characteristics 84 Conclusion 92 Notes 95 References 96 Part 2 What Do the Country Case Studies Tell Us? 97 Chapter 6 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 101 Electricity Generation 102 Electricity Transmission 103 Electricity Distribution 104 Comparison of Egyptian Generation Utilities 105 Comparison of Egyptian Distribution Utilities 107 Evolution of Egypt’s Electricity Sector since 2014 110 Conclusion 111 Notes 113 References 113 Chapter 7 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 115 Electricity Generation 116 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Contents vii Electricity Transmission 117 Electricity Distribution 117 Electricity Tariffs between Utilities 118 Comparison of Jordanian Generation Utilities 119 Comparison of Jordanian Distribution Utilities 122 Evolution of Jordan’s Electricity Sector since 2014 124 Conclusion 125 Notes 126 References 127 Chapter 8 Benefits and Challenges of Multiservice Providers: The Case of Morocco 129 Electricity Generation 129 Electricity Transmission 132 Electricity Distribution 132 Comparison of Moroccan Generation Utilities 134 Comparison of Moroccan Distribution Utilities 136 Evolution of Morocco’s Electricity Sector since 2014 139 Conclusion 140 Notes 142 References 142 Chapter 9 A Remarkably Sophisticated Power Market: The Case of Oman 145 Electricity Generation 146 Electricity Transmission 147 Electricity Distribution 147 Comparison of Generation Utilities in Oman 148 Comparison of Distribution Utilities in Oman 151 Evolution of Oman’s Electricity Sector since 2014 152 Conclusion 154 Notes 155 References 155 Chapter 10 Synopses of the Case Studies 157 Introduction 157 Arab Republic of Egypt: An Urgent Need for Sector Reforms 157 Jordan: Harvesting Results from a Restructuring of the Power Sector 159 Morocco: Benefits and Challenges of Multiservice Providers 160 Oman: A Remarkably Sophisticated Power Market 162 Note 164 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 viii Contents Conclusion 165 Cutting Hidden Costs in the Power Sector Is Key to Financing Sorely Needed Investment 165 Underpricing Is the Major Source of Inefficiencies, Although Otherwise Inefficiencies Are Economy and Utility Specific 166 MENA’s Power Sector Must Match Its Technical Success with Improvements in Commercial and Financial Management 167 Well-Targeted Institutional and Economic Reforms Would Boost MENA’s Power Sector 168 The Case Studies Yield Valuable Insights on the Variety and Nature of Reform Paths 169 More Systematic Monitoring of Power Sector Performance Is Needed 172 Appendix A Manual of Indicators and Data Sources 175 Appendix B Utilities Considered and Their Basic Characteristics 183 Appendix C Quasi-Fiscal Deficit: Hypothesis and Methodology 197 Appendix D Methodology for the Analysis of Drivers of Performance 215 Appendix E Core Values for MENA Indicators 219 Box I.1 The MENA Electricity Database 3 Figures 1.1 The Quasi-Fiscal Deficit as a Percentage of GDP, 14 MENA Economies, 2013 13 1.2 Comparison of Average End-User and Cost-Recovery Tariffs in MENA, 2013 (or most recent year with data, 2009–12) 23 2.1 OPEX per Connections for Distribution and Vertically Integrated Utilities in MENA ($), 2013 (or most recent year with data, 2009–12) 34 2.2 OPEX per Kilowatt Hour Sold ($), MENA, 2013 (or most recent year with data, 2009–12) 35 2.3 Residential Connections per Full-Time Equivalent Employee for Distribution and Vertically Integrated Utilities, MENA, 2013 (or most recent year with data, 2009–12) 37 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Contents ix 2.4 Distribution Losses of Distribution Utilities and Vertically Integrated Utilities in MENA (%), 2013 (or most recent year with data, 2009–12) 38 2.5 Energy Sales Volume per Connection for Distribution and Vertically Integrated Utilities in MENA (kWh), 2013 (or most recent year with data, 2009–12) 40 2.6 Total Billing per Connection for Distribution and Vertically Integrated Utilities in MENA ($), 2013 (or most recent year with data, 2009–12) 41 2.7 Collection Rates for Distribution and Vertically Integrated Utilities in MENA (%), 2013 (or most recent year with data, 2009–12) 43 2.8 OPEX Recovery from Sales for Distribution and Vertically Integrated Utilities, MENA (%), 2013 (or most recent year with data, 2009–12) 44 2.9 Sales as a Share of Total Costs for Distribution and Vertically Integrated Utilities, MENA (%), 2013 (or most recent year with data, 2009–12) 46 2.10 Accounts Receivable to Sales for Distribution and Vertically Integrated Utilities Utilities in MENA (days), 2013 (or most recent year with data, 2009–12) 47 2.11 Ratio of Debt to Equity across Utility Types in MENA (%), 2013 (or most recent year with data, 2009–12) 49 2.12 Ratio of Current Assets to Current Liabilities: Selected Utilities of All Types, MENA (%), 2013 (or most recent year with data, 2009–12) 51 2.13 Return on Assets: Selected Utilities of All Types, MENA (%), 2013 (or most recent year with data, 2009–12) 53 2.14 Return on Equity for Selected Utilities of All Types in MENA (%), 2013 (or most recent year with data, 2009–12) 54 6.1 Electricity Sector Organization, Arab Republic of Egypt 102 6.2 Share of Technology Type in Generating Electricity, Arab Republic of Egypt, 2013 103 6.3 Energy Sold from Distribution Utilities by Sector (medium- and low-voltage consumers), Arab Republic of Egypt, 2013 104 7.1 Electricity Sector Organization, Jordan, 2014 116 7.2 Share of Fuel Type in Electricity Generation, Jordan, 2009–13 117 7.3 Volume of Energy Distributed by Sector, Jordan, 2013 118 8.1 Electricity Sector Organization, Morocco 130 8.2 Generated Electricity in Morocco, by Technology Share, 2013 131 8.3 Share of Volume of Energy Distributed, by Sector, Morocco, 2013 133 9.1 Electricity Sector Organization, Oman 146 9.2 Share of Energy Distributed, by Consumer Sector, Oman, 2013 148 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 x Contents Tables 1.1 Quasi-Fiscal Deficit Calculations at the Economy Level, 2013 (except as noted) 13 1.2 Quasi-Fiscal Deficit Calculations at the Utility Level, Selected Utilities across MENA, 2013 (or most recent year with data, 2009–12) 17 1.3 Comparison of Utility- and Economy-Level Quasi-Fiscal Deficits for Economies with One Utility, 2013 (or most recent year with data, 2009–12) 19 1.4 Comparison of Economy- and Utility-Level Quasi-Fiscal Deficits for Economies with Multiple Utilities, 2013 (or most recent year with data, 2009–12) 21 1.5 Average Electricity Tariffs for MENA Economies and Comparison with Non-MENA Economies 23 1.6 Drivers of Electricity Tariff Design in MENA Economies 24 2.1 Comparing the Median Performance of Selected MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 31 2.2 OPEX per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 32 2.3 OPEX per kWh Sold for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 34 2.4 Residential Connections per Full-Time Equivalent Employee for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 36 2.5 Distribution Losses in MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 37 2.6 Volume of Energy Sold per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 39 2.7 Total Billing per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 41 2.8 Collection Rates for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 42 2.9 OPEX Recovery as a Share of Sales (%) for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 43 2.10 Energy Sales as a Share of Total Costs (%) for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 45 2.11 Ratio of Accounts Receivable to Sales in MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 46 2.12 Ratio of Debt to Equity for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 48 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Contents xi 2.13 Ratio of Current Assets to Current Liabilities for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 50 2.14 Return on Assets for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 50 2.15 Return on Equity for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) 52 3.1 Median Values of Ratio of Current Assets to Current Liabilities for Utilities (%), 2009–14 58 3.2 Median Values of Ratio of Current Assets to Current Liabilities for Utilities (%), 2009–14 58 3.3 Estimated Trend of Indicators for Utilities, 2009−14 59 3.4 Estimated Trends for Median Ratio of Current Assets to Current Liabilities, by Utility Type, 2009–13 60 3.5 Estimated Trends, by Utility Type, 2009–13 61 3.6 Estimated Trends for Ratio of Current Assets to Current Liabilities for Generation Utilities in Oman, 2009–13 62 4.1 Ranked Performance of MEDC (Oman) on Various Indicators 67 4.2 Trade-Off between Number of Distribution Utilities and Number of Indicators Common to All MENA Utilities 68 4.3 Ranks and Average Rank Score for Distribution Utilities, MENA 68 4.4 Ranks and Average Rank Score for Generation Utilities, MENA 70 4.5 Ranks and Average Rank Score for Vertically Integrated Utilities, MENA 71 5.1 Breakdown of Sample Utilities by Size, Ownership, Presence of a Separate Regulator, and Income, MENA, 2013 (or most recent year with data, 2009–12) 74 5.2 Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities 80 5.3 Number of Indicators with a Significant Relation to Each Factor, MENA Utilities 82 5.4 Number and Percentage of Significant Results, by Indicator Category 83 5.5 Categories of Indicators Whose Drivers of Performance Show Significant Results for a Substantial Proportion of the Indicators in that Category 84 6.1 Generation Mix, Arab Republic of Egypt, 2013 102 6.2 Electricity Transmission Data, Arab Republic of Egypt, 2013 103 6.3 Electricity Distribution Data, Arab Republic of Egypt, 2013 104 6.4 Comparing the Performance of Generation Utilities across Indicators, Arab Republic of Egypt, 2012/13 106 6.5 Comparing the Performance of Distributors across Indicators, Arab Republic of Egypt, 2012/13 108 7.1 Generation Mix, Jordan, 2013/14 116 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 xii Contents 7.2 Electricity Transmission Data, Jordan, 2013 117 7.3 Electricity Distribution Data, Jordan, 2013 118 7.4 Comparing the Performance of Generation Utilities across Indicators, Jordan and MENA Median, 2013 120 7.5 Comparing the Performance of Distributors across Indicators, Jordan and MENA Median, 2013 123 8.1 Generation Mix, Morocco, 2013 131 8.2 Electricity Transmission Data, Morocco, 2013 132 8.3 Electricity Distribution Data, Morocco, 2013 133 8.4 Comparing the Performance of Moroccan Generators across Indicators and against Egypt’s Upper Egypt Production Company and the MENA Median, 2013 134 8.5 Comparing the Performance of Moroccan Distributors across Indicators and against the MENA Median, 2013 137 9.1 Generation Mix, Oman, 2013 147 9.2 Electricity Transmission Data, Oman, 2013 147 9.3 Electricity Distribution Data, Oman, 2013 148 9.4 Comparing the Performance of Oman’s Generation Utilities across Indicators and against the MENA Median, 2013 149 9.5 Comparing the Performance of Oman’s Distributors across Indicators and against the MENA Median, 2013 152 CL.1 Comparing Median Utility Performance in the MENA Region and Elsewhere 167 CL.2 Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities 170 A.1 Descriptions of the 36 Core Indicators 176 A.2 Number of Indicator Points and Number of Utilities, by Type of Utility and Region 179 A.3 Number of Indicator Points Collected, 2009–13 180 A.4 Number of Data Points Collected, 2009–13 180 B.1 Summary of the Electricity Sector for 14 MENA Economies, 2013 184 B.2 Names and Abbreviations of MENA Utilities 187 B.3 Characteristics of MENA Utilities 188 B.4 Names and Abbreviations of Non-MENA Utilities 192 C.1 Sources of Data Used for the Economy-Level QFD Calculations 198 C.2 List of Alternate Sources for the Utility-Level QFD 200 C.3 Descriptions and Assumptions of Economy-Level QFD Components 201 C.4 Data and Sources Used for Calculating Collection Rates 203 C.5 Share of Energy Mixes Used in the Calculation of Tc (%) 204 C.6 LCOE Values Used to Calculate the Cost-Recovery Tariffs and Their Sources 204 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Contents xiii C.7 Components Used to Calculate the CAPEX According to Technology Type 206 C.8 Components Used to Calculate the CAPEX of the T&D Network 206 C.9 Values and Methodology Used in Calculating Labor Costs for the Arab Republic of Egypt 208 C.10 Utilities and Data Available for Jordanian Utilities 208 C.11 Calculating the Unit Labor Cost for Jordan 209 C.12 Calculating the Cost of Labor for the Two Utilities with Missing Values for Jordan 209 C.13 Calculating the Total Labor Costs for Jordan 209 C.14 Utilities and Data Available for Moroccan Utilities 210 C.15 Calculating the Unit Labor Cost for Morocco 210 C.16 Calculating the Cost of Labor for the Utilities with Missing Values for Morocco 210 C.17 Calculating the Total Labor Costs for Morocco 211 C.18 Calculating Average Monthly Earning Based upon ILO Data for the Republic of Yemen 211 C.19 Calculating the Cost of Labor for the Republic of Yemen 211 E.1 Indicator Names and Their Abbreviations, as Used in Tables E.2–E.4 of This Appendix 220 E.2a Technical and Operational Indicators 221 E.2b Technical and Operational Indicators 224 E.3 Financial Indicators 227 E.4 Commercial Indicators 230 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Acknowledgments This study was written by Daniel Camos, Robert Bacon, Antonio Estache, and Mohamad Mahgoub Hamid. The core team included Bipul Singh, Adnan Sirajee, and Mark Njore. It was carried out under the guidance of Erik Fernstrom and Charles Cormier. Vivien Foster provided substantial support and feedback throughout the exercise. The team gratefully acknowledges peer review com- ments received from Luis Andres, Sudeshna Banerjee, Vivien Foster, Victor Loksha, Marcelino Madrigal, Elvira Morella, and Sameer Shukla. We are grateful to Franck Bousquet and to Jonathan Walters, who provided guidance and advice, the latter in the concept stage and the former in the final stage. Electricity utilities, line ministries, and regulators are gratefully acknowledged for sharing data. This study also benefited from numerous exchanges with the Arab Union of Electricity, the Arab Electricity Regulatory Forum, and the Regional Center for Renewable Energy and Energy Efficiency, which also pro- vided support in gathering data from publicly available annual reports and utility financial statements. We are grateful to the following World Bank colleagues who provided advice and helped in data collection and validation: Waleed Alsuraih, Husam Beides, Roger Coma, Ferhat Esen, Mohab Hallouda, Ashish Khanna, Fanny Missfeldt-Ringius, Alejandro Moreno, Tara Shirvani, Simon Stolp, Manaf Touati, Chris Trimble, and Jianping Zhao. The data questionnaire was developed by Richard Schlirf, with support from Daniel Camos, Bipul Singh, Sudeshna Banerjee, and Marcelino Madrigal, all under the guidance of Vivien Foster. The data collection process was led by Daniel Camos and Bipul Singh. Jorge Sneij provided invaluable advice on infor- mation architecture and managing heavy datasets. The following people provided excellent support during data collection: Manaf Touati and Badr El Ahrari for Algeria; Hafez El Salmawy, Fatma Mostafa, and Bipul Singh for the Arab Republic of Egypt; Georges Dib and Bipul Singh for Bahrain; Aboubakar Hassan and Roger Coma for Djibouti; Simon Stolp and Adnan Sirajee for Iraq; Salah Tayeh and Usaimah Khalifeh for Jordan; Georges Zammar for Lebanon; Manaf Touati, Tayeb Amegroud, and Badr El Ahrari for Morocco; Hassan Taqi and Zahra Al Obaidani for Oman; Ahmad Esmaeel Al Mutawkel for the Republic of Yemen; Adnan Sirajee for Shedding Light on Electricity Utilities in the Middle East and North Africa   xv   http://dx.doi.org/10.1596/978-1-4648-1182-1 xvi Acknowledgments Qatar; Mansour Helal Al-Anazi and Debasish Ghosh for Saudi Arabia; Ezzedine Khalfallah for Tunisia; and Reem Muhsin Yusuf for the West Bank. Alberto Cena and Tayeb Amegroud provided substantial support in verifying and interpreting the data. The country case studies were led by Bipul Singh and Fatma Mostafa (Egypt), Imad Nejdawi and Mohamad Mahgoub Hamid (Jordan), Tayeb Amegroud and Richard Schlirf (Morocco), and Kirstin Morrison and Adnan Sirajee (Oman). Steven Kennedy provided excellent editing support, and Jewel McFadden effectively managed publication aspects. The financial and technical support by the Energy Sector Management Assistance Program (ESMAP) is gratefully acknowledged. ESMAP—a global knowledge and technical assistance program administered by the World Bank— assists low- and middle-income countries to increase their know-how and insti- tutional capacity to achieve environmentally sustainable energy solutions for poverty reduction and economic growth. ESMAP is funded by Australia, Austria, Denmark, the European Commission, Finland, France, Germany, Iceland, Italy, Japan, Lithuania, Luxembourg, the Netherlands, Norway, the Rockefeller Foundation, Sweden, Switzerland, the United Kingdom, and the World Bank. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 About the Authors Daniel Camos is a senior infrastructure economist at the World Bank, where he has worked in the energy and water global practices leading both operations and analytical work. Previously, he worked for the European Commission, the United Nations, and nongovernmental organizations. He currently works in the MENA region and has previous experience in Latin America and the Caribbean and in Sub-Saharan Africa. He has a training in economics and engineering, including a PhD in economics from the Paris School of Economics and the Université libre de Bruxelles; an MPA in international development from Harvard Kennedy School; and an industrial engineering degree from the Polytechnic University of Catalonia. Robert Bacon is a specialist in the economics of the energy sector. In recent years, he has been a consultant with the World Bank and the African Development Bank. Before that, he was the manager of the Oil and Gas Policy Division of the World Bank. Previously, for 30 years he was on the faculty of economics at the University of Oxford and a visiting fellow at the Oxford Institute for Energy Studies. His work on the energy sector includes studies on the assessment of progress and problems of reform of power sectors in low- and middle-income countries, econometric analyses of pricing and demand for petroleum products, and evaluation of the generation of employment throughout an economy by the actions of the energy sector. Antonio Estache is professor of economics at the Université libre de Bruxelles where he holds the Bernard Vanommeslaghe Chair aimed at increasing aware- ness throughout Europe about the importance of regulation and competition issues in public service industries. He is also a member of the European Center for Advanced Research in Economics and Statistics in Brussels. Previously, he spent 25 years at the World Bank, where he worked on infra- structure restructuring, procurement​, regulation, and public sector and tax reform. He has published extensively on these topics and ​ continues to work as a policy adviser to international organizations, governments, and parlia- ments around the world. Shedding Light on Electricity Utilities in the Middle East and North Africa   xvii   http://dx.doi.org/10.1596/978-1-4648-1182-1 xviii About the Authors Mohamad Mahgoub Hamid has an engineering background in mechanics and energetics and has recently been working as a consultant for the World Bank. His experience in the MENA region includes research and project coordination at the Regional Centre for Renewable Energy and Energy Efficiency in Cairo. As an energy policy analyst, he worked on projects with the League of Arab States and several United Nations agencies. Mohamad has followed trainings in several energy-related fields, including an energy statistics course at the International Energy Agency in Paris. He has a masters in engineering and a masters in manage- ment from the University of Aix-Marseille. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Abbreviations AER Authority for Electricity Regulation AICD Africa Infrastructure Country Diagnostic ANRE Agence Nationale de Régulation de l’Electricité BOOT build-own-operate-transfer CAIDI Customer Average Interruption Duration Index CAPEX capital expenditure CREG Commission de Régulation de l’Electricité et du Gaz (Electricity and Gas Regulatory Commission) DRSC Direction des Régies et Services Concédés DSP distribution service provider DU distribution utility ECRA Electricity and Cogeneration Regulatory Authority EEP electric energy production Egypt ERA Egypt Electric Utility and Consumer Protection Regulatory Agency EHV extra high voltage EMRC Energy and Minerals Regulatory Commission FTE full-time equivalent GCC Gulf Cooperation Council GDP gross domestic product GNI gross national income GU generation utility HFO heavy fuel oil HIC high-income country HV high voltage IAS international accounting standards IEA International Energy Agency IFRS international financial reporting standards IPP independent power producer IRENA International Renewable Energy Agency Shedding Light on Electricity Utilities in the Middle East and North Africa   xix   http://dx.doi.org/10.1596/978-1-4648-1182-1 xx Abbreviations ISO independent system operator IWPP independent water and power producer JREEF Jordan Renewable Energy and Energy Efficiency Fund LAC Latin America and the Caribbean LMIC low- and middle-income country MASEN Moroccan Agency for Solar Energy MEMDD Ministère de l’Enérgie, des Mines, et du Développement Durable (Ministry of Energy, Mines, and Sustainable Development) MEMR Ministry of Energy and Mineral Resources MENA Middle East and North Africa MHEW Ministry of Housing, Electricity and Water MIS main integrated system MoERE Ministry of Electricity and Renewable Energy NREA New and Renewable Energy Authority OEB Ontario Energy Board OECD Organisation for Economic Co-operation and Development OFGEM Office of Gas and Electricity Markets OPEX operational expenses PERC Palestinian Electricity Regulatory Council PERG Programme d’Electrification Rurale Global (Global Rural Electrification Program) PHES pumped hydroelectric energy storage PPA power purchase agreement PPP purchasing power parity QFD quasi-fiscal deficit RE renewable energy RFP request for proposal RISE Readiness for Investment in Sustainable Energy ROA return on assets ROE return on equity SAIDI System Average Interruption Duration Index SAIFI System Average Interruption Frequency Index SAOG Société Anonyme Omanaise Générale (Omani Public Limited Company) SCADA supervisory control and data acquisition SPC Services Permanents de Contrôle (Local Monitoring Units) TL transmission lines TPA third-party access TSO transmission system operator Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Abbreviations xxi TU transmission utility UMIC upper-middle-income country VIU vertically integrated utility All dollar amounts are U.S. dollars unless otherwise indicated. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Introduction The Region’s Electricity Challenge The electricity sector in the Middle East and North Africa (MENA) is in the grip of an apparent paradox. Although the region continues to hold the world’s larg- est oil and gas reserves and has been able to maintain electricity access rates of close to 100 percent in most of its economies, it may not be in a position to cater to the future electricity needs of its fast-growing population and their business activities. Primary energy demand in the region is expected to continue to rise at an annual rate of 1.9 percent through 2035, requiring a significant increase in generating capacity. Investments have not been rising fast enough to meet that requirement. The annual electricity investments needed to keep up with demand have been estimated at about 3 percent of the region’s projected gross domestic product (GDP) (Ianchovichina and others 2012). However, in most of the economies of the region, the ability to make those investments has been limited by fiscal con- straints. The region’s 2015 fiscal deficits averaged 9.3 percent of GDP, and the economies with the largest deficits were also those where electricity is most heavily subsidized. It seems unavoidable that, as economies adjust to their fiscal situation, they will continue to cut financing for the sector. To bridge the widen- ing financing gap, the electricity sector must find its own financing sources, and it must do so quickly to keep pace with demand. This work demonstrates that the solution is readily available: by improving the management and performance of the region’s utilities, more than enough resources could be freed up to make the investments needed to meet demand and operate at lower cost. These management and policy changes would make the production and consumption of electricity more affordable for the region’s taxpayers and could even make it more affordable for the poorest. They would also ease the transition toward renewable energy sources, reducing the depen- dency on imports for some economies and, for the economies that export oil and gas, extending the asset life of their nonrenewable resources. The essence of the solution is not surprising. It involves cutting costs and improving revenue. But the report provides detailed evidence of the size of the Shedding Light on Electricity Utilities in the Middle East and North Africa   1   http://dx.doi.org/10.1596/978-1-4648-1182-1 2 Introduction potential gain. In short, efficiency improvements could generate financing equal to twice the sector’s investment needs. That said, the optimal mix of cost-cutting and revenue-enhancing solutions is economy-specific, since cost and revenue- efficiency margins vary substantially across the region. For that reason, wherever several utilities share the responsibility to produce, transmit, and distribute elec- tricity within a given economy, the analysis and the evidence identify the major cost drivers and the sources of revenue losses at the utility level. The New MENA Electricity Database This quantitative assessment of electricity utilities’ performance has four main goals: • To provide a recent, detailed snapshot of technical and operational, commer- cial, and financial indicators for a large sample of electricity utilities in the MENA region, based on a major effort to collect original data for the region • To use these data to estimate the quasi-fiscal deficit (QFD) of the power sec- tor in the economies of the region, and to determine what proportion of the deficit can be attributed to underpricing (setting tariffs below costs), collection losses (failure to bill or collect revenues due to the utility), transmission and distribution losses, and overstaffing (employing more labor than an efficient utility of the same size and characteristics would do) • To assess the utilities’ relative performance on a wide variety of indicators in MENA and beyond, as well as the scope for improvements of MENA electric- ity utilities, both at the utility and economy levels • To assess the relevance of key factors on operators’ performance—that is, the degree to which performance is affected by (a) vertical integration; (b) utility size; (c) utility ownership; (d) the presence or absence of a regulator; and (e) the level of development of a given economy. • To distill useful lessons from four country case studies for the region to improve the performance of electricity utilities. To provide answers to these questions, we surveyed the power utilities in the region and established the MENA Electricity Database (box I.1). Before this survey, information on the region’s power sector was very uneven. The database thus forms a valuable public resource for policy makers as they reconcile the multiple dimensions of utility management performance with key policy con- cerns at the sector level. A limitation of the analysis is that the database’s baseline is 2013, and the power sectors of some MENA economies have changed consid- erably since then. The target audiences are managers of electricity utilities, regulators, policy makers, and other stakeholders (including members of civil society) concerned with the performance of specific utilities. The analysis is likely to be useful both at the sector level, since it highlights directions in which the sector may want to evolve in the region and in specific economies, and at the macroeconomic level, Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Introduction 3 Box I.1  The MENA Electricity Database This study is based on collection and analysis of primary data on 36 performance indicators in the Middle East and North Africa (MENA) Electricity Database. It covers 67 electricity utilities in 14 economies of the region: Algeria, the Arab Republic of Egypt, Bahrain, Djibouti, Iraq, Jordan, Lebanon, Morocco, Oman, Qatar, the Republic of Yemen, Saudi Arabia, Tunisia, and the West Bank.a It also relies on a sample of comparable non-MENA economies. The data were collected by means of a standardized survey completed by utilities and reg- ulatory agencies, covering indicators of technical, commercial, and financial performance. In some economies, the data were collected with support from local consultants or the public authorities. For the non-MENA economies, the data were collected from publicly available international databases. The sample of MENA operators comprises 12 vertically integrated utilities (VIUs), 29 distribution utilities (DUs), 23 generation utilities (GUs), and 3 transmission utilities (TUs). Data were collected from 2009 to 2013, with 2013 as the base year. Although the database contains much partial information, it also contains 945 base-year entries validated across 14 MENA economies and 3,832 entries for the period 2009–13. Source: World Bank compilation. a. Not included in the study are Libya, the Syrian Arab Republic, and the Islamic Republic of Iran. The utilities analyzed in this study are listed in appendix B. since it highlights the main drivers of the fiscal costs of the sector. At the utility level, the data (where they are detailed enough) allow managers and regulators to evaluate performance features, which can them weigh the trade-offs involved in making utilities more cost-effective and client-oriented. For regulators and the other stakeholders concerned with the need to improve governance of the sector, the overall analysis highlights significant information gaps. Without data, poor management and poor policy decisions are unlikely to be addressed, imposing a significant cost on users and taxpayers. The quality of the available data is also important. As a preliminary quality control measure, we asked utilities or economies to provide information on their accounting practices. First, we asked utilities about their adoption of (and com- pliance with) international accounting standards (IAS) or international financial reporting standards (IFRS): 60 percent responded positively, 10 percent nega- tively, and 30 percent did not respond. Second, we asked utilities whether they relied on cost-accounting systems; only one-third answered affirmatively—the other two-thirds were split between a negative answer and a nonresponse. Finally, the survey asked utilities if they relied on the supervisory control and data acqui- sition (SCADA) system of software and hardware elements to control processes locally or at remote locations or to monitor, gather, and process real-time data. Again, only one-third responded positively; the other two-thirds were split between a negative answer and a nonresponse. In sum, the quality of part of the available data—particularly that related to financial indicators—may be compro- mised by accounting practices. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 4 Introduction The Structure and Content of the Report The report is divided into two parts and several appendices. Part I (chapters 1–5) focuses on the region. Part II (chapters 6–10) consists of four country studies (Arab Republic of Egypt, Jordan, Morocco, and Oman) and a synopsis of all four. A short conclusion evokes the main themes and lessons from the entire report. Across the report, information at the utility level drawn from the MENA Electricity Database forms the basis of the analysis. Chapter 1 calculates the QFD (or hidden costs) of the power sector in each of the 14 MENA economies studied, a first attempt to quantify the hidden costs of power sector inefficiencies in the region. QFDs are presented at the economy level and at the utility level. The hidden costs of financial, technical, commercial, and labor-related inefficiencies contribute to the already delicate fiscal situation of most economies in the MENA region and cause financial strains when they accumulate over several years. The QFD (or hidden-cost) approach has been used in numerous analyses as a powerful tool to communicate with policy makers. It also has been applied to other infrastructure sectors, notably water.1 The QFD was computed for 28 utilities, of which 11 are VIUs and 17 DUs. A limitation of this exercise was that it was not possible to compute the QFD for GUs and TUs, for lack of data on the price at which they sell electricity (a generation utility might sell to a TU or to a single buyer or VIU, depending on the structure of the market in the economy in which it operates). Chapter 2 provides a snapshot of key performance indicators of MENA power utilities for which international comparisons are possible. These compari- sons are made between the 14 MENA economies as well as with countries outside the region for which data were readily available. The MENA data are taken from the MENA Electricity Database. Comparisons are made for 14 tech- nical, financial, and commercial indicators to highlight possible differences in performance among regions. Within MENA, further comparisons are made between utilities to highlight strong and weak performers for the indicator in question. Ideally, comparisons for every indicator would be based on the same set of utilities within the region and on the same countries or utilities from outside it. However, this ideal is not yet attainable. The database has varying coverage for the 36 indicators included in the survey, for two reasons. First, certain indicators are relevant only to certain types of utilities. Second, many utilities did not report data on certain indicators, even when relevant: for example, only 46 of the 67 utilities surveyed reported data on their return on assets. Chapter 3 examines performance indicators over time. Where governments have introduced power sector reform, policy makers should examine the reform’s effects based on certain indicators. Changes should be expected to be gradual rather than sudden. To construct the MENA Electricity Database, we asked utilities to provide information for 2009–14. Because of the number of trend calculations to be made and the brevity of the data series, we decided to construct aggregates across utilities, indicator by indicator, and to carry out trend Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Introduction 5 analysis on these aggregates for the few years of data available. Data were further disaggregated by utility type (distribution, generation, and vertically integrated) to check whether they revealed different trends. Chapter 4 considers the relative overall performance of utilities within the MENA region when more than one indicator is considered. Understanding the reasons behind a strong overall performance can suggest policies that could be applied elsewhere, just as understanding poor performances can suggest ways to ameliorate the problem. The variability of performance across indicators suggests that some form of average performance measure is required if utilities are to be seen in a context of overall strengths and weaknesses. The challenge in creating a multi-indicator approach is to combine indicators measured in very different units. This is done by the use of the “average rank score”: the average of the ranked positions among utilities over a number of indicators. Average rank scores were calculated for each of the 17 DUs for which we had data on five indicators, 13 GUs for which we had data on three indicators, and 8 VIUs for which we had data on three indicators. This exercise made it possible to identify the best- and worst-performing utilities within the groups analyzed. Chapter 5 investigates whether certain organizational differences are corre- lated with differences in performance. Policy choices, such as the unbundling of the sector, the introduction of private ownership, the size of utilities, or the intro- duction of a separate regulatory authority, have been suggested as key steps in improving the overall performance of the electricity sector (see, for example, Bacon and Besant-Jones 2001). Because evidence of the benefits of power sector reform has not been overwhelming, policy responses to underperformance are being reevaluated.2 Evidence of the impact of various reform strategies can help to inform the debate. The data collected for the analysis of performance in the MENA region contribute to this discussion. In the present study, the availability of data drawn from a large number of utilities exhibiting different characteristics provides the opportunity to test for the effects of various reform strategies in a new way. If the average performance of all public utilities on various indicators is poorer than that of the average for all private utilities on the same indicators, then this supports the argument that privatization helps to improve performance. In making such comparisons we recognize that many factors contribute to performance on a given indicator, so that differences between public and private utilities are not due solely to owner- ship status. However, a significant difference in performance by ownership type would support the argument that ownership matters, whereas lack of significant difference would suggest that mode of ownership alone does outweigh all the other factors in determining performance on the indicator. Part II focuses on detailed analysis of four countries that have taken very dif- ferent approaches to the power sector: Egypt, Jordan, Morocco, and Oman (chapters 6 to 9). Each country study provides an overview of the national power sector and an analysis of utility performance (comparing these with Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 6 Introduction regional median values) to identify potential areas of improvement. The narrative and figures presented in these chapters focus on the year 2013. Their characteristics and challenges of the case study countries are repre- sentative of the 14 MENA economies in this study, though each has a unique story to tell, whether related to its dependence on fossil fuel imports, its population and geographical size, or the initial and organizational structure of its electricity sector. Several themes relevant to the region as a whole emerge from the case studies. First, all four countries have undertaken significant reforms of their electricity sectors over the past decades. Second, factors exogenous to the electricity sector have had an impact on utility performance. These include political instability; disruptions in primary-fuel supply; and spillovers from regional conflicts. Third, the four case-study countries have addressed in different ways the link between water and energy, a very salient matter in the MENA region. Fourth, some of the case studies deal with the introduction of renewable sources to the energy mix in a region where fossil fuels remain the dominant source of electricity. As with the report as a whole, the case studies are limited by the availability of data. Yet they represent a good start toward the more consistent and devel- oped analysis needed to meet the major challenges identified in this report. Key points raised in the case studies are presented, country by country, in chapter 10. Notes 1. For example, the methodology used to compute utility QFDs in this chapter was largely inspired by Trimble and others (2016). Another example of the use of QFD is Eberhard and others (2008). 2. Vagliasindi and Besant-Jones (2013) show that unbundling can deliver performance improvements, but not for all indicators. References Bacon, R., and J. Besant-Jones. 2001. “Global Electric Power Reform, Privatization and Liberalization of the Electric Power Industry in Developing Countries.” Annual Review of Energy and Environment 26 (1): 331–59. Also as: Energy and Mining Sector Board Discussion Paper 2, World Bank, Washington, DC. Cambini, C., and D. Franzi. 2013. “Independent Regulatory Agencies and Rules Harmonization for the Electricity Sector and Renewables in the Mediterranean Region.” Energy Policy 60 (September): 179–91. Eberhard, A., V. Foster, C. Briceño-Garmendia, F. Ouedraogo, D. Camos, and M. Shkaratan. 2008. “Underpowered: The State of the Power Sector in Sub-Saharan Africa.” Africa Infrastructure Country Diagnostic (AICD), summary of Background Paper No. 6, World Bank, Washington, DC. Ianchovichina, E., A. Estache, R. Foucart, G. Garsous, and T. Yepes. 2012. “Job Creation through Infrastructure Investment in the Middle East and North Africa.” Policy Research Working Paper No. 6164, World Bank, Washington, DC. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Introduction 7 Trimble, C., M. Kojima, I. P. Arroyo, and F. Mohammadzadeh. 2016. “Financial Viability of Electricity Sectors in Sub-Saharan Africa: Quasi-Fiscal Deficits and Hidden Costs.” Policy Research Working Paper 7788, World Bank, Washington, DC. Vagliasindi, M., and J. Besant-Jones. 2013. Power Market Structure: Revisiting Policy Options. Directions in Development Series. Washington, DC: World Bank. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 PA R T I How Do MENA’s Electricity Utilities Perform? Shedding Light on Electricity Utilities in the Middle East and North Africa   9   http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 1 Quasi-Fiscal Deficits in MENA’s Power Sector Estimating the power sector’s quasi-fiscal deficit (QFD) provides a first attempt at quantifying the hidden costs originating from sector inefficiencies. When incurred by utilities over years, hidden costs can cause financial strain. This, in turn, can worsen an already delicate fiscal situation. A majority of economies in the Middle East and North Africa (MENA) region are at risk of financial strain. The QFD, used in numerous analyses of various infrastructure sectors, includ- ing electricity, is a powerful tool for communicating with policy makers.1 Four types of inefficiency contribute to the QFD: • Financial, as measured by the size of the gap between the average tariff and the cost-recovery rate (underpricing) • Technical, or the difference between actual transmission and distribution (T&D) losses and those of an ideal utility2 • Commercial, or the share of bills not collected (collection losses) • Labor, as estimated by comparing the number of customers per utility employee against an efficiency benchmark3 (overstaffing) All four inefficiences can be expressed in absolute monetary terms or as a percentage of gross domestic product (GDP) or of a utility’s revenue. Their cal- culation is illustrated in equation 1.1: Qe Tc (lm − ln )  NC  413 − Qe (Tc − Te ) + + Qe Te (1 − Rct ) +  NE  CL 1 − lm    413  (1.1) Financial Technical Commercial Labor inefficiency inefficiency inefficiency inefficiency Shedding Light on Electricity Utilities in the Middle East and North Africa   11   http://dx.doi.org/10.1596/978-1-4648-1182-1 12 Quasi-Fiscal Deficits in MENA’s Power Sector Qe = end-user consumption (kilowatt-hours, kWh) lm = technical loss rate (%) Tc = cost-recovery tariff ($/kWh) Rct = collection rate (%) Te = average end-user tariff ($/kWh) NE = number of employees ln = normative loss rate (%) 413 = benchmark number of customers NC = number of customers per employee CL = cost of labor ($) per employee This chapter seeks to estimate the hidden costs of the power sector in MENA. It represents the first time—to the best of our knowledge—that QFDs have been calculated in a context where utilities are not fully vertically integrated. We esti- mate QFDs at both the economy and utility level. In economies where there is only one vertically integrated utility (VIU), one might expect both estimates to be similar. But where the fuel used to generate electricity is subsidized, a signifi- cant gap between the two is created. Meanwhile, in economies with some degree of unbundling and more than one utility, the economy and utility-level QFDs differ as expected, and a number of methodological considerations and hypoth- eses need to be considered. In theory, the sum of the QFDs of an economy’s utilities should equal the economy’s own QFD. Although both QFDs can be expressed as a percentage of GDP, the utility-level QFD can also be expressed as a fraction of the utility’s revenues. Of the 14 MENA economies considered in this report, all except the Arab Republic of Egypt, Jordan, Morocco, Oman, and the West Bank are treated as having just one VIU. Several sources were used in an effort to acquire a maximum amount of data for the calculations of the economy- and utility-level QFDs. Most data come from the MENA Electricity Database, the World Development Indicators (WDIs), and reports from the Arab Union of Electricity. Appendix C provides the sources used for each indicator. This appendix also explains how several methodological chal- lenges related to data availability were solved. Often, data for all 14 economies were not available in a single source, requiring further collection and verification. Particular challenges were faced, for instance, in gathering utilities’ bill-collection rates—necessary to calculate the QFD’s commercial inefficiency component—or estimating their labor costs in economies with several utilities. Economy-Level Results As can be seen in figure 1.1, half of the 14 MENA economies studied have a QFD above 4 percent of their GDP. Of particular concern is the fact that Lebanon, Djibouti, Bahrain, and Jordan have a QFD between 7.5 percent and 9 percent of GDP. Table 1.1 lists QFDs for the 14 economies (expressed both in absolute terms and as a percentage of GDP) as well as the four individual components (as a percentage of GDP). Five economies have a QFD below 3 percent of GDP (West Bank, Morocco, Tunisia, Qatar, and Algeria); another four have a QFD Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficits in MENA’s Power Sector 13 Figure 1.1  The Quasi-Fiscal Deficit as a Percentage of GDP, 14 MENA Economies, 2013 10 9 8 7 6 Percent 5 4 3 2 1 0 –1 . p. ep co n n an a q an ria sia r k ti ta bi no ai Ira an Re ou oc ,R rd Om ge ni hr Qa ra tB ba ab Tu Jo or ib en iA Al Ba Le es Dj M Ar m ud W Ye t, Sa yp Eg Source: World Bank calculations. Note: GDP = gross domestic product; MENA = Middle East and North Africa. Table 1.1  Quasi-Fiscal Deficit Calculations at the Economy Level, 2013 (except as noted) QFD as QFD components as share of GDP (%) Absolute QFD share of Economy (US$ million) GDP (%) Underpricing T&D losses Collection losses Overstaffing Lebanon 3,826 8.9 8.20 0.41 0.21 0.03 Djibouti 101 8.2 0.98 1.08 5.24 0.88 Bahrain 2,640 8.0 7.86 0.02 0.02 0.13 Jordan 2,608 7.8 5.96 0.84 0.75 0.21 Egypt, Arab Rep. 18,219 6.4 5.61 0.42 0.06 0.28 Saudi Arabia 38,467 5.2 4.81 0.11 0.17 0.07 Yemen, Rep. 1,494 4.2 3.16 0.81 0.08 0.11 Iraq 7,888 3.6 2.44 0.83 0.13 0.21 Oman 2,496 3.2 2.70 0.22 0.18 0.10 Algeria 4,720 2.3 1.46 0.37 0.10 0.32 Qatar 3,224 1.6 1.47 0.02 0.10 0.01 Tunisia 655 1.4 0.34 0.39 0.54 0.15 Morocco 948 1.0 0.65 0.33 0.20 −0.21 West Bank −13 −0.1 −0.84 0.30 0.30 0.13 Source: World Bank calculations. Note: The year is 2013 for all except the following: 2012 for Lebanon, Iraq, Morocco, and the West Bank; and 2011 for Djibouti. This variation reflects data availability. GDP = gross domestic product; MENA = Middle East and North Africa; QFD = quasi-fiscal deficits; T&D = transmission and distribution. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 14 Quasi-Fiscal Deficits in MENA’s Power Sector between 3 percent and 6 percent of their GDP (Oman, Iraq, the Republic of Yemen, and Saudi Arabia), and five economies have a QFD between 6 percent and 9 percent of GDP (Egypt, Jordan, Bahrain, Djibouti, and Lebanon). In short, the QFD’s share of GDP is relatively small in Maghreb economies, and large in some Mashreq and Gulf Cooperation Council (GCC) economies. In absolute terms, the highest QFDs are to be found in Saudi Arabia ($38 billion), Egypt ($18 billion), and Iraq ($8 billion) and the lowest in the West Bank (with a negative QFD of $13 million), Djibouti ($101 million, despite having the second-highest QFD when expressed as a percentage of GDP), and Tunisia ($655 million). These values strongly correlate to the size of the economy and to the consumption levels of its population. As seen in table 1.1, we obtain negative values for overstaffing in Morocco and underpric- ing in the West Bank. This simply means that Morocco’s ratio of customers to employees is better than the efficiency benchmark (413:1) used in this report, and that the West Bank’s cost-recovery tariff is smaller than the average end-user tariff (based on the energy mix of Israel, given that the West Bank imports all its electricity from there). Underpricing appears to be the main factor behind high QFD values: in 8 of the 14 economies, this component represents more than three-quarters of the QFD. In as many as 11 economies, it represents two-thirds. Underpricing does not, in itself, help disentangle the two common types of subsidies: that is, subsidies (a) of elec- tricity and (b) of the fuels used to generate electricity. This is because the cost- recovery tariff used to estimate the economy-level QFD is based on levelized energy costs, computed as weighted averages of each economy’s energy mix, to which a factor was added to account for T&D costs. The utility-level QFDs pre- sented in table 1.2 and their comparison with economy-level QFDs allow us to better differentiate the two types of subsidies. The reason Djibouti and the West Bank are notable exceptions to the trend of underpricing as a driving force of the QFD is that both economies have high average end-user tariffs: $0.31 per kilowatt- hour (kWh) and $0.16 per kWh, respectively. Technical inefficiencies (represented by T&D losses) are an important part of some economies’ QFDs: they represent more than one-fifth of the total QFDs in Iraq, Morocco, Tunisia, the West Bank, and the Republic of Yemen. Commercial inefficiencies (represented by bill collection losses) represent as much as two-thirds of the QFD in Djibouti, more than one-third in Tunisia, and a substantial share in Morocco and the West Bank. Uncollected bills do not appear to be a key QFD component in the remaining 10 economies. Finally, labor inefficiencies represent between 10 percent to 15 percent of the QFDs in Algeria, Tunisia, and Djibouti. Expressed as a percentage of GDP, they represent 1.0 percent in Djibouti and between 0.2 percent and 0.3 percent in Egypt and Jordan. Addressing this type of inefficiency may be a delicate act for governments, because it implies reducing the size of state-owned enterprises (SOEs). Providing public jobs—and subsidized basic services—has been part of the social contract in the region for the past several decades, in exchange for social stability. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficits in MENA’s Power Sector 15 How do economy QFD results for MENA compare with other regions? A recent study that computed QFDs for 17 Sub-Saharan African economies obtained values that ranged between −0.3 percent and 6.0 percent of GDP (Trimble and others 2016). Again, our values for MENA range from −0.1 percent in the West Bank to 8.9 percent in Lebanon (see table 1.1 for QFDs, including values for each of the four components, as a share of GDP). This indicates that MENA’s utilities are more inefficient than Africa’s: the median QFD in the 17 African economies is 0.8 percent of GDP, whereas it is close to 4.0 percent in the 14 MENA economies. Interestingly, although the MENA QFD appears to be driven mostly by financial inefficiency, in the case of Sub-Saharan Africa, techni- cal and commercial inefficiencies play the largest role overall. Another study (Ebinger 2006) of water and electricity sectors in 16 econo- mies of Europe and Central Asia (ECA) found that tariffs set below cost-recovery rates were the biggest culprits behind the energy sector’s hidden costs (the study did not consider overstaffing). QFDs as a share of GDP were as high as 14 percent in Tajikistan and 8 percent in Moldova—that is, comparable to the highest values in our sample of MENA economies. Utility-Level Results We computed QFDs for VIUs and distribution utilities (DUs) when sufficient data were available. This was done for all 14 economies but Qatar (because of insufficient information on that country’s VIUs, the Qatar General Electricity and Water Corporation, KAHRAMAA). In total, QFDs were computed for 28 utili- ties, of which 11 are VIUs and 17 are DUs. A limitation of this exercise is that we were not able to compute QFDs for generation utilities (GUs) and transmission utilities (TUs). The reason is that, although we had data on the end-user tariffs set by VIUs and DUs (Arab Union of Electricity 2014), we did not have this data for GUs selling electricity (be it to a TU, a single buyer, or a VIU, depending on the market structure of the economy) or for TUs selling electricity. This gap pre- vented us from computing the financial inefficiency component at the utility level for GUs and TUs, which is a key component of the economy-level QFD in much of MENA. The formula used to compute the QFD at the utility level is the same as that used at the economy level. However, two important differences merit clarification, because they drive much of the difference between the two types of QFD: • Qe at the economy level is the end-user consumption (taken from the WDIs), whereas Qe at the utility level represents the amount of energy billed (taken from the MENA Electricity Database). • Whereas Tc at the economy level is based on the energy mix and levelized costs of each generation technology (in addition to a T&D component), Tc at the utility level is based on the investment and operating costs of the utility annualized (taken from the MENA Electricity Database). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 16 Quasi-Fiscal Deficits in MENA’s Power Sector To compute the QFD of a utility, we used the year 2013 or, depending on data constraints, another year in the study period (2009–13). In some cases, to get around data gaps, we used observations across a few years. The meth- odology is described in appendix C, as well as the assumptions considered and sources of data. The results of our exercise—that is, utility-level QFDs, pre- sented as a fraction of GDP and as a fraction of revenue of utilities—are presented in table 1.2. The QFDs of VIUs tend to be higher than those of DUs. Of the 11 VIUs analyzed, 3 have a QFD above 4.0 percent of GDP (reaching as much as 8.4 percent in the case of Electricité de Djibouti), 6 are between 1.0 percent and 2.5 percent, and 2 are below 0.5 percent. These figures differ significantly from the 17 QFDs computed for DUs: 3 are at 0.8 percent of GDP or above (reaching as much as 2.3 percent in the case of the Jordan Electric Power Company), 6 are between 0.3 percent and 0.5 percent, and 8 are at 0.2 percent or below (and 6 of this last subset are in Egypt). One way to adjust for the differences in types of utilities is to look at the QFD as a percentage of utility revenue, which provides revealing results. The Northern Electricity Distribution Company (NEDCO) in the West Bank is the DU with the lowest QFD as a proportion of its revenue (25 percent), whereas the Iraqi Ministry of Electricity’s proportion is 1,267 percent (in other words, the monetary value of its inefficiencies is more than 12 times its revenue). Some other outliers on the upper end include Jerusalem District and Tubas District Electricity Companies (448 percent and 193 per- cent, respectively) and Electricité du Liban (372 percent). Of the 28 utilities analyzed, 13 have inefficiencies that are higher than their revenues. In other words, these utilities would double their revenue if they were to maximize their efficiency. When one looks at the QFD components at the utility level, the results are—unsurprisingly—similar to those seen at the economy level: underpricing is by far the main driving force, except in Djibouti where collection losses play this role. For DUs, underpricing is the main driving force in Jordan, Morocco, and Oman. The West Bank is the exception, because T&D or collection losses are the driving force here. This is because the average end-user tariff in the West Bank is relatively high. Table 1.3 compares the economy QFD with the utility QFD for economies with only one utility. The right-hand column discusses the observed differences and provides plausible explanations for them. In economies with one utility, we would expect to obtain similar values for both the economy-level and the utility- level QFD (with minor differences due to methodological considerations or dif- ferent sources of data). However, in four of nine economies with one utility, we observe economy QFDs that are between 1.5 and 3.0 times higher than the utility QFD. Two main factors explain this: (a) the subsidies of fuel used for elec- tricity generation (as in Saudi Arabia or Lebanon) artificially diminish the VIU's cost-recovery tariff, because the latter is based on artificially low operating Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Table 1.2  Quasi-Fiscal Deficit Calculations at the Utility Level, Selected Utilities across MENA, 2013 (or most recent year with data, 2009–12) QFD as share QFD as share QFD components as share of GDP (%) Economy Type Utility of GDP (%) of revenues (%) Underpricing T&D losses Collection losses Overstaffing Djibouti Vertically integrated Electricité de Djibouti 8.4 93 1.20 1.10 5.24 0.88 Lebanon Vertically integrated Électricité du Liban 5.7 372 5.27 0.26 0.11 0.03 Bahrain Vertically integrated Electricity and Water Authority 4.1 125 3.96 0.01 0.01 0.13 Tunisia Vertically integrated Société Tunisienne de l’Électricité et du Gaz 2.9 73 1.67 0.56 0.51 0.15 Jordan Distribution Jordan Electric Power Company 2.3 71 1.92 0.33 0.05 −0.01 Iraq Vertically integrated Ministry of Electricity 2.3 1,267 1.45 0.50 0.10 0.21 Algeria Vertically integrated Société Nationale de l’Électricité et du Gaz 1.9 129 1.16 0.33 0.10 0.32 Saudi Arabia Vertically integrated Saudi Electricity Company 1.7 131 1.40 0.04 0.17 0.07 Yemen, Rep. Vertically integrated Public Electricity Corporation 1.4 152 0.79 0.47 0.06 0.11 West Bank Distribution Jerusalem District Electricity Company 1.1 448 0.37 0.55 0.08 0.08 Morocco Vertically integrated Office National de l’Électricité et de l’Eau Potable 1.1 38 0.74 0.24 0.18 −0.08 Jordan Distribution Electricity Distribution Company 0.8 88 0.45 0.07 0.24 0.04 Morocco Distribution LYDEC 0.5 70 0.29 0.01 0.15 0.09 Oman Distribution Muscat Electricity Distribution Company 0.4 61 0.32 0.03 0.07 −0.01 table continues next page 17 18 Table 1.2  Quasi-Fiscal Deficit Calculations at the Utility Level, Selected Utilities across MENA, 2013 (or most recent year with data, 2009–12) (continued) QFD as share QFD as share QFD components as share of GDP (%) Economy Type Utility of GDP (%) of revenues (%) Underpricing T&D losses Collection losses Overstaffing Oman Distribution Mazoon Electricity Distribution Company 0.4 64 0.31 0.03 0.05 0 Oman Distribution Majan Electricity Company 0.3 62 0.21 0.03 0.04 0.02 Egypt, Arab Rep. Distribution South Cairo Electricity Distribution Company 0.3 95 0.18 0.01 0.01 0.06 Oman Vertically integrated Rural Areas Electricity Company 0.3 121 0.20 0.01 0.01 −0.04 Egypt, Arab Rep. Distribution Canal Electricity Distribution Company 0.3 100 0.19 0 0 0.05 Egypt, Arab Rep. Distribution North Cairo Electricity Distribution Company 0.2 95 0.14 0.01 0 0.04 Oman Vertically integrated Dhofar Power Company 0.2 72 0.11 0.01 0.02 0 West Bank Distribution Tubas District Electricity Company 0.2 193 −0.03 0.01 0.11 0 West Bank Distribution Northern Electricity Distribution Company 0.2 25 0.03 0.05 0.06 0.02 Egypt, Arab Rep. Distribution Upper Egypt Electricity Distribution Company 0.1 125 0.08 0.02 0.01 0.03 Egypt, Arab Rep. Distribution Middle Egypt Electricity Distribution Company 0.1 103 0.08 0.01 0 0.03 Egypt, Arab Rep. Distribution Alexandria Electricity Distribution Company 0.1 115 0.07 0.01 0 0.04 Egypt, Arab Rep. Distribution North Delta Electricity Distribution Company 0.1 92 0.08 0.01 0.01 0.03 Egypt, Arab Rep. Distribution El-Behera Electricity Distribution Company 0.1 99 0.07 0.01 0 0.03 Source: World Bank calculations. Utilities selected based on data availability. Note: GDP = gross domestic product; MENA = Middle East and North Africa; QFD = quasi-fiscal deficits; T&D = transmission and distribution. Table 1.3  Comparison of Utility- and Economy-Level Quasi-Fiscal Deficits for Economies with One Utility, 2013 (or most recent year with data, 2009–12) Economy QFD Utility QFD Economy Utility (% of GDP) (% of GDP) Factors explaining differences between the two QFDs Algeria Socièté Nationale de 2.3 1.9 Only underpricing and T&D losses (economy QFD is higher by 0.3% and 0.04%, respectively). Specifically, l’Electricité et du Gaz Tc is different for both cases, given different sources and methodologies used (see appendix C for (SONELGAZ) more details): it is 12.0¢/kWh for the economy, and 10.6¢/kWh for the utility. In addition, there is a minor difference in losses: 18.4% for the economy (World Development Indicators, or WDI) and 18.8% for the utility (MENA Electricity Database, or MED). Bahrain Electricity and Water 8.0 4.1 Underpricing, T&D losses, and collection losses (economy QFD is higher by 3.9%, 0.01%, and 0.01%, Authority respectively). The main variable driving differences is Qe, and Tc to a much lesser extent. Specifically, (EWA) Tc is 11.3¢/kWh for the economy and 10.5¢/kWh for the utility; Qe is 24.6 TWh for the economy (WDI), and 13.4 TWh for the utility (MED). The reason for this important difference is that EWA represents only 63% of installed capacity, because the other 37% (2,249 MW)a is the self-generation of an aluminum smelting company, Alba. Djibouti Electricité de Djibouti 8.2 8.4 Underpricing and T&D losses, particularly underpricing, which is 0.2% higher for the utility-level QFD. (EDD) This difference is driven by a slightly higher Tc for the utility (35.5¢/kWh) than for the economy (34.7¢/kWh). The small difference between the two values is due to methodology. Both QFDs correspond to the year 2011 because there were insufficient data for 2013. Iraq Ministry of Electricity 2.5 2.3 Underpricing (0.18% higher in the economy) and to a smaller extent T&D losses. The main (MOE) factor driving these differences is Tc, which is 11.9¢/kWh for the economy and 9.9¢/kWh for the utility. Both QFDs correspond to the year 2012 because of insufficient data for 2013. Lebanon Electricité du Liban 8.9 5.7 Differences observed in all but overstaffing components, but mainly in underpricing (economy QFD is (EdL) 2.93% higher). This is primarily driven by Tc and Qe differentials. Tc is 29.0¢/kWh for the economy and 34.8¢/kWh for the utility. Qe is 7.2 TWh for the utility (2012) and 13.8 TWh for the economy (2012). The difference in Qe can be attributed to the fact that about 35% of energy consumption in Lebanon is self-generation for own consumption.b Both QFDs correspond to the year 2012 because of insufficient data for 2013. Qatar Qatar General Electricity and 1.6 — n.a. Water Corporation (KAHRAMAA) table continues next page 19 20 Table 1.3  Comparison of Utility- and Economy-Level Quasi-Fiscal Deficits for Economies with One Utility, 2013 (or most recent year with data, 2009–12) (continued) Economy QFD Utility QFD Economy Utility (% of GDP) (% of GDP) Factors explaining differences between the two QFDs Saudi Arabia Saudi Electricity Company 5.2 1.7 Underpricing (economy-level QFD higher by 3.41%) and to a lesser extent T&D losses (economy QFD (SEC) higher by 0.07%). The difference is driven mainly by different Tc values: 14.9¢/kWh for the economy and 5.4¢/kWh for the utility. This significant differential can be attributed to the importance of fuel subsidies, which are included in the economy Tc but not in the utility Tc, because these subsidies would not be reflected in the OPEX of the utility. In 2013, the fuel subsidy for electricity generation given to SEC was US$40 billion.c If we were to add this value to the OPEX of SEC, we would then see an increase in its Tc (5.4¢/kWh to 21.0¢/kWh), and consequently the utility QFD would also increase (from 1.7% to 7.2%) to a value much closer to the economy QFD. The remaining differences between the economy and utility QFDs can be attributed to methodology. Tunisia Société Tunisienne de 1.4 2.9 Underpricing (1.3% higher in utility), and to some lesser extent T&D losses (0.2% higher in utility). l’Electricité et du Gaz Tc is the main driver in differences because it is 11.2¢/kWh for the economy and 15.5¢/kWh (STEG) for the utility. Yemen, Rep. Public Electricity Corporation 4.2 1.4 Differences observed in all but overstaffing components, and most in underpricing (2.37% higher in the (PEC) economy), and to a lesser extent T&D losses (0.34% higher in the economy). Tc is the main driver: 21.1¢/kWh for the economyd and 8.8¢/kWh for the utility. This significant differential can be attributed to the importance of fuel subsidies, which are included in the economy Tc but not in the utility Tc, as these subsidies would not be reflected in the OPEX of the utility. Fuel subsidies for electricity generation given to PEC amounted to about $1.1 billion.e If we were to add this value to the OPEX of PEC, we would then see an increase in its Tc (8.8¢/kWh to 30.6¢/kWh), and consequently the utility QFD would also increase (from 1.4% to 5.6%). The remaining differences between the economy and utility QFD can be attributed to methodology. Source: World Bank calculations. Specific sources and additional information in notes. Note: GDP = gross domestic product; kWh = kilowatt-hour; MENA = Middle East and North Africa; MW = megawatts; OPEX = operating expenses; Qe = end-user consumption (kWh); QFD = quasi-fiscal deficit; T&D = transmission and distribution; Tc = cost-recovery tariff (US¢/kWh); TWh = terawatt-hours; — = negligible (insufficient data available); n.a. = not applicable. a. Alba’s website: http://www.albasmelter.com/About%20Alba/Factsfigures/Pages/default.aspx. b. World Bank 2009, 18. c. Jeddah Chamber of Commerce 2015, 11. d. Of the 68 percent of electricity that is fuel (WDI), we assume that half (34 percent of total) is based on diesel self-generation, and that the remaining half (34 percent of total) is produced by PEC, using in equal amounts heavy fuel oil (HFO) and diesel, as per the Arab Union of Electricity (AUE) Manual of Power Stations. e. Fattouh and El-Katiri 2012: 30. Although this figure corresponds to 2008, the cost of fuel purchases of PEC remained stable between 2009 and 2012 according to PEC reported figures. Quasi-Fiscal Deficits in MENA’s Power Sector 21 expenses; and (b) even in economies with a vertically integrated electricity market, self-generation can be widespread among residential consumers (as in Lebanon or the Republic of Yemen) or even industrial consumers (as with an aluminum smelting company in Bahrain). Table 1.4 compares the economy-level QFD with the utility-level QFD for economies with multiple utilities. In such economies, we expected to see higher values for economy QFDs than for the sum of utilities’ QFDs because we did not have sufficient data to compute the QFDs of all the utilities present. This assumption holds true in the case of Egypt, Jordan, and Oman. Interestingly, in the West Bank and in Morocco we see the opposite happening. In the West Bank, this could be because all the electricity consumed is imported from Israel, and the selling price may be delinked from the levelized cost of energy in Israel. In Morocco, this could be attributed to understaffing: Morocco’s Office National de l’Electricité et de l’Eau Potable (ONEE) has one employee per 557 customers, higher than the benchmark of 413. In addition to being a VIU, ONEE is also the single buyer of electricity in Morocco and sells mainly to the 11 DUs. Table 1.4  Comparison of Economy- and Utility-Level Quasi-Fiscal Deficits for Economies with Multiple Utilities, 2013 (or most recent year with data, 2009–12) Economy QFD Utility QFD Economy (% GDP) (% GDP) Factors explaining differences between the two types of QFDs Egypt, Arab Rep. 6.4 1.4 The QFD of utilities corresponds only to the nine Egyptian DUs, because data were not sufficient to compute it for the TUs and GUs. Tc for the economy is 12.6¢/kWh, and for the DUs it oscillates between 3.8¢/kWh and 4.5¢/kWh. This presumably indicates that the electricity that DUs buy is subsidized. Jordan 7.8 2.8 The QFD of utilities corresponds to only two Jordanian DUs, because data were not sufficient to compute it for the one remaining DU, the six GUs, and the one TU. Tc for the economy is 19.8¢/kWh, and for the two DUs it is 12.4¢/kWh and 10.9¢/kWh. This presumably indicates that the electricity that the DUs buy is subsidized. Morocco 1.0 1.5 The utility QFD corresponds to the vertically integrated utility ONEE because data for the 11 DUs were insufficient. The Tc for the vertically integrated utility ONEE, which is the single buyer of electricity and generates 42% of the total electricity supplied, is 4.4¢/kWh, although the economy Tc is 3.6¢/kWh. As a single buyer, ONEE sells about 44% of its electricity to the DUs, at an average price of 10.5¢/kWha, and the rest is sold directly to consumers (average tariff is calculated at 11.3¢/kWh). Both QFDs correspond to 2012 because data were not sufficient for 2013. table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 22 Quasi-Fiscal Deficits in MENA’s Power Sector Table 1.4  Comparison of Economy- and Utility-Level Quasi-Fiscal Deficits for Economies with Multiple Utilities, 2013 (or most recent year with data, 2009–12) (continued) Economy QFD Utility QFD Economy (% GDP) (% GDP) Factors explaining differences between the two types of QFDs Oman 3.2 1.3 The QFD of utilities corresponds to only one VIU and three DUs, because data were not sufficient to compute it for the remaining one DU, one TU, one VIU, and 12 GUs. Tc for the economy is 1.6¢/kWh, and Tc for utilities oscillates between 5.4¢/kWh and 6.7¢/kWh, except for the rural VIU for which the Tc goes up to 27.0¢/kWh. The latter is presumably due to the usage of diesel generators to produce electricity. Here again, because the Tc for the economy tends to be above the one for DUs, presumably the electricity that the DUs buy is subsidized. West Bank −0.1 1.4 It is striking that the economy-level QFD is slightly negative. This result is driven by a negative underpricing component (−0.8% of GDP), because Te is 16.4¢/kWh whereas Tc is 11.4¢/kWh. This result is surprising, and may reflect the fact that the cost at which DUs in the West Bank are buying electricity from Israel is delinked from costs. The rest of the QFD economy components oscillate between 0.1% and 0.3% of GDP. The QFDs of utilities corresponds to three DUs, which are all the utilities in the West Bank in this study (most generation comes from Israel). In theory, given that our utility QFD covers all utilities in the West Bank, both types of QFD should be equal. The reason this is not so is because the Tc of utilities is considerably higher and oscillates between 12.7¢/kWh and 19.3¢/kWh. Both QFDs correspond to 2012 because data were not sufficient for 2013. Source: World Bank calculations. Note: DU = distribution utility; GDP = gross domestic product; GU = generation utility; kWh = kilowatt-hours; ONEE = Office National de l’Electricité et de l’Eau Potable; QFD = quasi-fiscal deficit; Tc = cost-recovery tariff (¢/kWh); Te = average end-user tariff (¢/kWh); TU = transmission utility: VIU = vertically integrated utility. a. 0.88 Moroccan dirham per kilowatt-hour average. What Can Be Done about Underpricing in MENA Economies? Underpricing is by far the biggest factor behind the high QFD values observed in the power sector of the MENA region. This is not surprising when looking at figure 1.2, because the average end-user tariff appears to be below the cost- recovery level in all economies but the West Bank. In some cases, this differential is particularly acute—Lebanon and the Republic of Yemen being primary examples. Other economies with relatively high average electricity tariffs are Tunisia, Morocco, and Djibouti. MENA average end-user tariffs are low when compared to the rest of the world. Table 1.5 lists tariffs for MENA economies and basic statistics for econo- mies both in and outside MENA (the latter based on a sample of more than 60 countries). The average and median of non-MENA economies are more than twice and four times the respective MENA values. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficits in MENA’s Power Sector 23 Figure 1.2  Comparison of Average End-User and Cost-Recovery Tariffs in MENA, 2013 (or most recent year with data, 2009–12) 40 35 30 25 US¢ / kWh 20 15 10 5 0 Le ep. M ep. sia W rain an Ar ria ud cco a m an n Ba r k q ti bi ta no an Ira ou Om Ye rd ge ,R ni R ra Qa o h ba tB ab Tu ib Jo or en iA Al Dj es t, Sa yp Eg Average end-user tari (Te) Cost recovery tari (Tc) Source: World Bank calculations. Note: MENA = Middle East and North Africa; kWh = kilowatt-hour. Table 1.5  Average Electricity Tariffs for MENA Economies and Comparison with Non-MENA Economies Economy ¢/kWh Economy ¢/kWh Algeria 5.25 Morocco 11.36 Bahrain 0.8 Oman 2.6 Djibouti 31.0 Qatar 2.2 Egypt, Arab Rep. 1.78 Saudi Arabia 1.33 Iraq 0.9 Tunisia 10.19 Jordan 6.63 West Bank 16.38 Lebanon 3.29 Yemen, Rep. 3.1 MENA basic statistics Non-MENA basic statistics Average 6.92 Average 15.62 Quartile 1 1.89 Quartile 1 8.5 Median 3.2 Median 13.55 Quartile 3 9.3 Quartile 3 18.98 Number of observations 14 Number of observations 61 Source: World Bank calculations. Note: MENA tariffs come from the Arab Union of Electricity 2014; they represent the average domestic tariff of 250 kWh per month (in U.S. cents per kilowatt-hour). Non-MENA tariffs are obtained from Readiness for Investment in Sustainable Energy (RISE) and are mainly from sub-Saharan Africa, Asia, and Latin America. See Banerjee and others (2016). MENA = Middle East and North Africa; ¢/kWh = U.S. cents per kilowatt hour. Tariff reforms could help improve the political viability of efforts to increase cost-recovery rates. Improved cost recovery, subsidy cuts, and better targeting demand a detailed look at the current design of electricity tariffs. Table 1.6 hints at the fact that, in many of the economies, there is at least some scope to improve the tariff design. Cross-subsidies do not systematically favor the poorest, even if Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 24 Table 1.6  Drivers of Electricity Tariff Design in MENA Economies Peak/off-peak Surcharge for heavy Economy rates Use Seasonality nonresidential users Lowest tariff Cross-subsidies Algeria Yes Yes Yes Yes — — Bahrain No Yes No No Residential Yes Egypt, Arab Rep. No Yes No No Agriculture Yes Iraq No Yes No No — No Saudi Arabia Yes Yes Yes No Agriculture Yes Kuwait No No No No — No Jordan Yes (afternoon) Yes No Yes Industry Yes Lebanon No Yes No No Residential and commercial Yes Libya No (Only in residential) No No Residential Yes Morocco Yes Yes Yes No Agriculture Yes Oman Yes Yes Yes No — No Qatar No Yes No No Agriculture and residential No Syria Yes (evening) Yes No No Agriculture and public Yes administration Tunisia Yes Yes Yes Yes Residential Yes United Arab Emirates No Yes (mostly) No No Residential and commercial Yes West Bank No No No No Industry Yes Yemen, Republic No Yes No No Residential Yes Source: Arab Union of Electricity 2014 and utilities’ websites. Note: MENA = Middle East and North Africa. — = not available. Quasi-Fiscal Deficits in MENA’s Power Sector 25 most economies seem to protect farmers and to some extent residential users. If the priority is to cut direct subsidies, then progressive cross-subsidies across user types might offer a way to promote joint efficiency, equity, financial viability, and, when needed, fiscal sustainability. Estimating the scope for improving tariff design would demand a much more thorough analysis than table 1.6 provides. But even though this book largely focuses on the supply side of the electricity business, there are other relevant dimensions to consider. In particular, there is a need to pay equivalent attention to the demand side, including to consumers’ ability and willingness to pay. There is a good case for assessing the extent to which pricing options could do better to improve incentives on both sides of the market. International experience suggests that tariff reforms are likely to be part of a politically viable financing solution, in addition to management improvements needed to reduce the cost inefficiencies documented in the following chapters. We have seen in the previous section that fuel subsidies for electricity generation are also an important part of the underpricing challenge in some MENA economies. For example, the QFD of Saudi Arabia increases by more than 5 percent if we account for fuel subsidies for electricity generation— equivalent to US$40 billion. The Republic of Yemen’s increases by more than 4 percent—​ equivalent to US$1.1 billion. This represents by far the highest source of inefficiencies in these economies’ electricity sectors. We did not have sufficient data to study the impact of subsidies in more detail, leaving an important topic for future analysis. Conclusion The median QFD value in the 14 MENA economies analyzed here is about 4 percent of GDP. This represents more than the average investment needed in the region’s electricity sector, estimated at about 3 percent of GDP. In other words, the sector’s investment gap could be filled simply by removing a fraction of the current level of inefficiency. Indeed, there is heterogeneity across the region, as QFD estimates vary from 8 percent to 9 percent of GDP in Lebanon, Djibouti, or Bahrain to less than 1.5 percent in Tunisia, Morocco, and the West Bank. Underpricing appears to be the main driver of QFD for most economies in the region. This is due to the significant presence of subsidies for electricity and for the fuel used to generate electricity. Also, the cost-recovery tariff in many coun- tries is high due to the significant presence of fuel in the energy mix. The other components of the QFD are T&D losses, collection losses, and overstaffing. These should not be forgotten, their aggregate median value for the 14 MENA econo- mies is 0.8 percent of GDP, but it goes as high as 7.2 for Djibouti for example. Different priorities will need to be identified in different countries to reduce the QFD, incorporating the political economy in certain measures, be it tariff reforms or managing the sector’s labor force. To the best of our knowledge, this chapter represents the first effort to compute and compare QFDs at both the economy and at the utility levels. One advantage of this dual exercise is that the utility-level QFD is useful to Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 26 Quasi-Fiscal Deficits in MENA’s Power Sector utility managers, particularly when there are multiple electricity utilities in a given economy. Another advantage is that even without explicitly computing fuel subsidies for electricity production, we can get a sense of their size by comparing the results of the two types of QFD. Finally, this chapter suffered from several data constraints. The most obvious is that we did not have data on the prices at which GUs were selling electricity, which prevented us from computing QFDs for these utilities. Another limita- tion was the quality of the data collected, particularly for variables such as the bill-collection rate or the cost of labor. Obtaining disaggregated data on the number of employees and the revenues in multiservice utilities also proved to be challenging, for example, in the case of Electricity and Water Authority in Bahrain, Société Tunisienne de l’Electricité et du Gaz in Tunisia, ONEE in Morocco, and Société Nationale de l’Electricité et du Gaz in Algeria. Notes 1. For example, the methodology used for the QFD in this chapter has been greatly inspired by Trimble and others (2016). Another example of the use of QFD is Eberhard and others (2008). 2. T&D is fixed at 5 in this report because the best-performing utilities in our sample are slightly above this value, which we consider to correspond to an “ideal utility.” 3. This inefficiency is estimated at 413 for developing countries, following Trimble and others (2016). References Arab Union of Electricity. 2014. Electricity Tariff in the Arab Countries. Statistical bulletin. Amman: Arab Union of Electricity. Banerjee, S. G., A. Moreno, J. Sinton, T. Primiani, J. Seong. 2016. “Regulatory Indicators for Sustainable Energy: A Global Scorecard for Policy Makers.” Working paper 112828, World Bank, Washington, DC. Eberhard, A., V. Foster, C. Briceño-Garmendia, F. Ouedraogo, D. Camos, and M. Shkaratan. 2008. “Underpowered: The State of the Power Sector in Sub-Saharan Africa.” Africa Infrastructure Country Diagnostic, summary of background paper 6, World Bank, Washington, DC. Ebinger, J. O. 2006. “Measuring Financial Performance in Infrastructure: An Application to Europe and Central Asia.” Policy Research Working Paper 3992, World Bank, Washington, DC. Fattouh, B., and L. El-Katiri. 2012. “Energy Subsidies in the Arab World.” Arab Human Development Report Research Paper Series, United Nations Development Programme, New York. Jeddah Chamber of Commerce. 2015. Sectorial Report on Saudi Arabia—Electricity July 2015. Jeddah: Jeddah Economic Gateway. http://www.jeg.org.sa/data/modules/contents​ /uploads/infopdf/2832.pdf. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficits in MENA’s Power Sector 27 Prasad, T. V. S. N., M. Shkaratan, A. K. Izaguirre, J. Helleranta, S. Rahman, and S. Bergman. 2009. Monitoring Performance of Electric Utilities: Indicators Benchmarking in Sub-Saharan Africa. Washington, DC: World Bank. https://www.esmap​ .org/sites/esmap.org/files/P099234_AFR_Monitoring%20Performance%20of%20 Electric%20Utilities_Tallapragada_0.pdf. Trimble, C., M. Kojima, I. P. Arroyo, and F. Mohammadzadeh. 2016. “Financial Viability of Electricity Sectors in Sub-Saharan Africa: Quasi-Fiscal Deficits and Hidden Costs.” Policy Research Working Paper 7788, World Bank, Washington, DC. World Bank. 2009. Energy Efficiency Study in Lebanon—Final Report. Washington, DC: World Bank. http://climatechange.moe.gov.lb/viewfile.aspx?id=205. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 2 Comparing the Region’s Performance with the Rest of the World This chapter provides a snapshot of key performance indicators for those power utilities in the Middle East and North Africa (MENA) region for which both regional and international comparisons are possible. The MENA data are taken from the MENA Electricity Database (MED), which itself is based on question- naires administered to 67 power utilities in the region. Data are for 2013 except in a few cases where another year between 2009 and 2012 was chosen because of data constraints. The MED covers 12 VIUs, 23 generation utilities (GUs), 29 distribution utilities (DUs), and 3 transmission utilities (TUs) (names and corre- sponding abbreviations for the utilities considered are in appendix B). Ideally, each indicator would be compared across an identical set of utilities within MENA and also against a single set of utilities in countries from other regions. However, this ideal is not attainable at present. The database has varying coverage of the 36 indicators included in the survey for two reasons. First, certain indicators are relevant only to certain types of utilities: for example, generators do not experience distribution losses and hence do not collect such data. Second, many utilities did not provide data on certain indicators, even though these are relevant to utility performance: for example, only 49 of the 67 utilities surveyed reported data on their return on assets (ROA). For data on non-MENA economies the challenge is greater: surveys cover only a subset of the MENA indicators, and this coverage differs across regions. Added to this is the problem of missing data. As a result, the sample of countries and utilities available1 for comparison varies by indicator. To some extent the choice of indicator for global comparisons was dictated by data availability. MENA and non-MENA performance was compared using the first (Q1), second (Q2), and third (Q3) quartiles of the data. The second quartile (median) denotes the level of the indicator at which one half of the Shedding Light on Electricity Utilities in the Middle East and North Africa   29   http://dx.doi.org/10.1596/978-1-4648-1182-1 30 Comparing the Region’s Performance with the Rest of the World observations were smaller and one half were larger. The median is preferred to the mean value because the latter is too sensitive to the dispersion of per- formance levels and the possible existence of extreme outliers in the data.Q1 separates the smallest 25 percent from the other observations, and Q3 separates the largest 25 percent from the other observations. This reduces the impact of extreme observations, which are given equal weight, relative to other obser- vations smaller than Q1 or larger than Q3. The use of Q1 and Q3, in addition to Q2, which is close to the mean, allows a more complete comparison. For instance, the medians for two groups may be close, but the upper quartile may be considerably larger for one region, indicating that the best performers are not comparable. For some indicators, a high value denotes good perfor- mance and a low value poor performance (for example, ROA), whereas for other indicators, a high value denotes poor performance (for example, distri- bution losses). Extensive statistical testing described in chapter 5 indicates that a substantial number of performance indicators are related to a country’s income level (at higher incomes, performance is better). As a result, comparisons between regions may also be affected by differences in regional income levels. For exam- ple, income levels in Sub-Saharan Africa are generally well below those of the MENA region, whereas those in Latin America and the Caribbean (LAC) are nearer to the MENA levels. Hence the median values of performance indicators for Africa are likely to be lower than median levels for MENA. This comparison suggests that median MENA performance might be close to the Q3 performance of non-MENA utilities, allowing for the effect of income levels (for indicators where high values indicate high performance). This effect is particularly the case for vertically integrated utilities (VIUs) because of the predominance of utilities from low-income economies in Sub-Saharan Africa. Comparisons—based on median, Q1, and Q3 values—are made for techni- cal, financial, and commercial indicators to highlight possible differences in performance across regions. Further, MENA utilities have been compared with one another to highlight strong and weak performers (for the indicator in question). For some indicators (for example, those related to operating expenses [OPEX] or total costs), performance cannot be expected to be the same for different types of utilities. For example, a VIU bears generation and transmis- sion costs as well as the distribution costs borne by a DU supplying the same number of customers. However, data for VIUs cannot be disaggregated into various functions, so no comparison can be made between DUs and VIUs for such indicators. For other indicators, such as the ROA, performance depends entirely on efficiency, and so all utility types can be directly compared. The comparison between MENA and non-MENA utilities is split between DUs and vertically integrated utilities only when the indicators have different ­ definitions, depending on the type of utility. In the case of GUs and TUs, there were too few observations to make adequate comparisons both within MENA and across global regions. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 31 Summary of Results and Overall Assessment Table 2.1 summarizes the relative performance of MENA versus non-MENA economies by comparing the median values of indicators across all utilities for which data were available. If an indicator is such that different utility structures can be expected to have similar performance levels, then the median values are used across all the utility types possible. If an indicator is such that performance varies by type, then comparisons are contained accordingly. The main findings can be summarized as follows: • MENA economies tend to perform better than non-MENA ones for about half of the indicators selected. • The financial health of the MENA utilities is questionable. For example, the MENA value for accounts receivable over sales is almost three times that of non-MENA economies, and the ratio of current assets to current liabilities is lower than the non-MENA median—and lower than 100 percent. This is reinforced by a very high debt-to-equity ratio (almost four times the non-MENA median), leaving utilities highly exposed to external shocks. Although the ROA is above the non-MENA level, it is still low, suggesting that improvements in performance are required. Table 2.1  Comparing the Median Performance of Selected MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Indicator Utility type MENA median Non-MENA median MENA is superior? OPEX/connection ($) Distribution 346 129 No OPEX/connection ($) Vertically integrated 1,237 594 No OPEX/kWh ($) Distribution 0.10 0.14 Yes OPEX/kWh ($) Vertically integrated 0.07 0.18 Yes Residential connections/employee Distribution 252 367 No Residential connections/employee Vertically integrated 90 157 No Distribution losses (%) All 11 12 Yes Energy sold/connection (kWh) All 4,223 3,405 Yes Total billing/connection ($) All 299 292 Yes Collection rate (%) All 92 94 No Sales/OPEX (%) Distribution 93 98 No Sales/OPEX (%) Vertically integrated 92 87 Yes Sales/total costs (%) Distribution 88 67 Yes Sales/total costs (%) Vertically integrated 56 67 No Accounts receivable/sales (days) All 148 52 No Debt/equity (%) All 357 91 No Current assets/current liabilities (%) All 84 88 No Return on assets (%) All 3 1 Yes Return on equity (%) All 6 0 Yes Source: World Bank calculations. Note: kWh = kilowatt-hours; MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 32 Comparing the Region’s Performance with the Rest of the World • One other area where there is a large difference between MENA and non-MENA economies is that of connections per employee. The low ratio in MENA sug- gests that hiring practices in MENA may need to be reviewed in some cases. Detailed Comparisons for Selected Indicators A more complete comparison between MENA and non-MENA economies is carried out in tables 2.2 to 2.15 by using indicator values at the Q1, Q2, and Q3 levels, both for all utilities and for VIUs, DUs, TUs, and GUs separately, where appropriate. The values of the indicators for all comparable utilities within MENA (for which data were available) are plotted and compared against the MENA and non-MENA median values. Technical and Operational Performance Indicators OPEX per connection ($). OPEX consists mainly of fuel costs, labor costs, maintenance and repair costs, and energy purchases. The relative proportions of these categories can be expected to vary among economies depending on rela- tive prices. In particular, wage rates may vary to a large degree, while unit fuel costs may be similar (though, in fact, these costs can vary considerably in MENA given the importance of subsidy schemes in some economies). For instance, an inefficient utility might be overstaffed because of hiring practices or have excessive fuel bills because of poor dispatch decisions. Expenditure on maintenance may be inadequate because of a desire to cut costs in the short run. If labor costs dominate, then we expect that a more efficient utility would have lower OPEX per connection, other factors being held constant. Table 2.2 presents the data on OPEX per connection. This indicator does not apply to GUs, so those are excluded. Also, VIUs and DUs are separated because, as previ- ously noted, a VIU incurs generation and transmission OPEX in addition to distribution costs. Table 2.2  OPEX per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Number of Quartile 1: best Quartile 3: worst Utility type Region utilities performers ($) Median ($) performers ($) All MENA 36 215 411 864 Non-MENA 70 75 172 416 Distribution MENA 25 157 346 547 Non-MENA 48 67 129 197 Vertically integrated MENA 11 665 1,237 1,547 Non-MENA 22 243 594 852 Source: World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. The table indicates that OPEX per connection is less than $346 in 50 percent of distribution utilities inside MENA and less than $129 in 50 percent outside MENA. Notably, OPEX per connection is more than $547 in 25 percent of distribution utilities in MENA. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 33 The medians are substantially higher for DUs and VIUs inside MENA than outside the region. The best-performing groups in MENA (Q1) and the worst- performing groups (Q3) are also much more pronounced than their non- MENA equivalents. These results suggest that the components of OPEX are substantially larger than their equivalents in non-MENA economies. In the case of VIUs, this could be explained by different fuel mixes used to generate elec- tricity: OPEX for fuel-based generation are higher than for hydro-based genera- tion, and fuel is important to many MENA economies. However, this cannot apply to the differences seen in DUs, which could also be due to higher wage bills in MENA. Figure 2.1a plots the values of OPEX per connection for DUs in MENA. The values for the Muscat Electricity Distribution Company (MEDC) and the Mazoon Electricity Company (MZEC), both in Oman, stand out as being very high, given the amount of variation across the other utilities in MENA, whereas the values for the Egyptian utilities are all low. The latter could indicate an insuf- ficient level of maintenance. Further, consumption levels explain part of these differences: for example, electricity consumption per capita in Oman is almost four times higher than that in the Arab Republic of Egypt. Differences in wage levels could also be a factor. It would be necessary to examine the reasons for these findings before concluding that the Oman utilities are unusually inefficient or the Egyptian utilities are very efficient.2 Figure 2.1b plots the values for the VIUs in MENA. Oman’s Rural Areas Electricity Company (RAECO) is an outlier; its high value can be explained by the fact that it covers only rural areas, whereas the other VIUs listed also cover urban areas and benefit from economies of scale. Here again, differ- ences could be explained by (a) differences in fuel mix (for example, fuel is critical to Djibouti’s economy, and the value for that country’s VIUs is above the median), (b) differences in salaries (for example, none of the VIUs in the Gulf Cooperation Council [GCC] economies has a value below the median) and levels of consumption (which is very low in the Republic of Yemen, for example). OPEX per kWh sold ($). Low values of this indicator generally suggest relatively high levels of efficiency because the utility can supply a given amount of electricity at a relatively low operating cost. Where maintenance and repairs are obviously suboptimal, a low value of this indicator reflects inefficiency. In the case of VIUs, a low value can mask a utility’s energy mix, specifically whether it is more or less based on fossil fuel. Table 2.3 presents the calculations of OPEX per kWh sold, divided by quartile for utilities both in and beyond MENA. For DUs, the median value is lower inside MENA than outside, and this trend is even more defined among the worst performers (Q3). These results suggest that utilities inside MENA have been better able to hold down costs per kWh than utilities outside the region. Meanwhile, as has been noted, OPEX per connection is higher inside MENA, where more kWh are supplied per connection than in other regions. Although the same trends are observed for VIUs, these Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 34 Comparing the Region’s Performance with the Rest of the World Figure 2.1  OPEX per Connections for Distribution and Vertically Integrated Utilities in MENA ($), 2013 (or most recent year with data, 2009–12) a. Distribution utilities b. Vertically integrated utilities Oman - MEDC Oman - RAECO Oman - MZEC Morocco - LYDEC West Bank - TUBAS Djibouti - EDD West Bank - NEDCO Morocco - REDAL Lebanon - EdL Jordan - IDECO Morocco - AMENDIS TAN Qatar - KAHRAMAA Morocco - RAK Morocco - RADEEMA Morocco - RADEEJ Oman - DPC Morocco - RADEEL Morocco - AMENDIS TET Saudi Arabia - SEC Morocco - RADEES Morocco - RADEEF Tunisia - STEG Morocco - RADEM Egypt, Arab Rep. - CEDC Iraq - MOE Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - NCEDC Morocco - ONEE Egypt, Arab Rep. - AEDC Egypt, Arab Rep. - UEEDC Algeria - SONELGAZ Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - NDEDC Yemen, Rep. - PEC Egypt, Arab Rep. - SDEDC 0 500 1,000 1,500 2,000 0 1,000 2,000 3,000 4,000 5,000 6,000 U.S. dollars U.S. dollars MENA Median Non-MENA Median Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. Table 2.3  OPEX per kWh Sold for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Number of Quartile 1: best Quartile 3: worst Utility type Region utilities performers ($) Median ($) performers ($) All MENA 36 0.04 0.10 0.13 Non-MENA 28 0.08 0.14 0.24 Distribution MENA 26 0.04 0.10 0.13 Non-MENA 19 0.06 0.14 0.23 Vertically integrated MENA 10 0.05 0.07 0.18 Non-MENA 6 0.09 0.18 0.25 Source: World Bank calculations. Note: kWh = kilowatt-hours; MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 35 results should be treated with caution, keeping in mind that the sample of non-MENA VIUs is small. Figure 2.2a shows the values of OPEX per kWh for DUs in MENA. The val- ues for Lyonnaise des Eaux de Casablanca (LYDEC) in Morocco and the Jerusalem District Electricity Company (JDECO) in the West Bank are notably high. As is the case for OPEX per connection, Egypt’s utilities have much lower values than those in other economies, a finding that requires further research to understand. Figure 2.2b plots the values of OPEX per kWh for VIUs in MENA. The value for Electricité de Djibouti (EDD) in Djibouti is a clear outlier, and Electricité du Liban (EdL) in Lebanon and RAECO in Oman are also well above the median values for both MENA and non-MENA economies. The values for EDD and EdL may be explained by exceptionally high fuel costs and various types of inefficien- cies. RAECO’s value could be explained because it covers only rural areas. Figure 2.2  OPEX per Kilowatt Hour Sold ($), MENA, 2013 (or most recent year with data, 2009–12) a. Distribution utilities b. Vertically integrated utilities Morocco - LYDEC West Bank - JDECO Djibouti - EDD Morocco - REDAL West Bank - NEDCO Lebanon - EdL Morocco - AMENDIS TET Jordan - JEPCO Morocco - RADEES Oman - RAECO Morocco - RADEEL Morocco - AMENDIS TAN Morocco - ONEE Morocco - RAK Morocco - RADEEF Morocco - RADEM Bahrain - EWA West Bank - TUBAS Morocco - RADEEMA Jordan - IDECO Iraq - MOE Morocco - RADEEJ Oman - MJEC Yemen, Rep. - PEC Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - AEDC Algeria - SONELGAZ Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - EEDC Oman - DPC Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - NDEDC Egypt, Arab Rep. - SDEDC Saudi Arabia - SEC Egypt, Arab Rep. - MEEDC 0 0.05 0.10 0.15 0.20 0.25 0 0.10 0.20 0.30 0.40 0.50 U.S. dollars U.S. dollars MENA median Non-MENA median Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 36 Comparing the Region’s Performance with the Rest of the World Residential connections per full-time equivalent employee (FTE). Relatively efficient utilities are expected to sustain a greater number of connections per employee. A VIU would be expected to have a lower value than a DU with the same number of connections because supplying the additional generation and transmission requires extra labor. Table 2.4 presents the values for this indicator. The median value for DUs outside MENA is about 50 percent greater than in MENA, suggesting that MENA’s utilities are overstaffed from a purely technical perspective. Among both the worst performers and the best performers, non- MENA utilities outperform those in MENA by a similar factor, suggesting that overstaffing is a regionwide phenomenon (there are very few observations for the Table 2.4  Residential Connections per Full-Time Equivalent Employee for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Number of Quartile 1: worst Quartile 3: best Utility type Region utilities performers Median performers All MENA 24 124 184 311 Non-MENA 68 225 336 654 Distribution MENA 19 151 252 317 Non-MENA 57 272 367 691 Vertically integrated MENA 5 44 90 173 Non-MENA 11 102 157 284 Source: World Bank calculations. Note: MENA = Middle East and North Africa. MENA VIUs and, accordingly, results for this subgroup should be treated with caution). Figure 2.3a shows the values for residential connections per employee among DUs in MENA. REDAL in Morocco has the highest value by far, and this is well up in the range of best performers outside MENA. At the other extreme, the worst performer—Tubas, in the West Bank—has an exceptionally low value. This may well be due to a social policy to increase employment even when this may drive up costs. Figure 2.3b plots the values for residential connections per employee for VIUs in MENA. The values for Saudi Electricity Company (SEC) in Saudi Arabia and Dhofar Power Company (DPC) in Oman are above the MENA and non-MENA medians and suggest that these utilities are appropriately staffed. The value for RAECO in Oman is very low, which is consistent with the fact that this is the only rural VIU in our sample. Distribution losses (percent). Higher losses indicate lower efficiency levels. Because there is no a priori reason that the losses should be different between DUs and VIUs, both types of utility can be directly compared. Table 2.5 shows the quartile data for distribution losses (percent) inside and outside MENA. The median MENA value is slightly lower than the non-MENA value. The best- performing group in MENA is also slightly better than the equivalent non- MENA group. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 37 Figure 2.3  Residential Connections per Full-Time Equivalent Employee for Distribution and Vertically Integrated Utilities, MENA, 2013 (or most recent year with data, 2009–12) a. Distribution utilities b. Vertically integrated utilities Morocco - REDAL Morocco - RADEEJ Saudi Arabia - SEC Oman - MEDC Jordan - JEPCO Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - NDEDC Oman - DPC Jordan - IDECO Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - NCEDC Yemen, Rep. - PEC Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - AEDC Djibouti - EDD West Bank - NEDCO Jordan - EDCO Oman - MZEC Oman - RAECO Oman - MJEC West Bank - TUBAS 0 200 400 600 800 1,000 1,200 0 50 100 150 200 Residential connections per FTE Residential connections per FTE MENA median Non-MENA median Source: MENA Electricity Database and World Bank calculations. Note: FTE = full-time employee; MENA = Middle East and North Africa. Table 2.5  Distribution Losses in MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: best Quartile 3: worst Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 114 9 12 18 MENA 37 8 11 14 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 38 Comparing the Region’s Performance with the Rest of the World Figure 2.4 illustrates the performance of the DUs and VIUs in the MENA region. The performance of the VIUs is generally poor. The worst cases—the Ministry of Electricity (MOE) in Iraq, the Public Electricity Corporation (PEC) in the Republic of Yemen, EdL in Lebanon, and JDECO in the West Bank—have losses far above the median values for both MENA and non-MENA, and even the Figure 2.4  Distribution Losses of Distribution Utilities and Vertically Integrated Utilities in MENA (%), 2013 (or most recent year with data, 2009–12) Iraq - MOE Yemen, Rep. - PEC Lebanon - EdL West Bank - JDECO Algeria - SONELGAZ West Bank - TUBAS Oman - DPC Morocco - ONEE Tunisia - STEG Jordan - JEPCO West Bank - NEDCO Oman - MJEC Jordan - EDCO Egypt, Arab Rep. - AEDC Oman - MZEC Morocco - AMENDIS TET Jordan - IDECO Egypt, Arab Rep. - MEEDC Oman - RAECO Egypt, Arab Rep. - EEDC Morocco - AMENDIS TAN Egypt, Arab Rep. - NCEDC Oman - MEDC Egypt, Arab Rep. - NDEDC Morocco - RADEEL Egypt, Arab Rep. - SCEDC Morocco - RAK Morocco - REDAL Morocco - RADEM Morocco - LYDEC Egypt, Arab Rep. - CEDC Morocco - RADEEMA Saudi Arabia - SEC Morocco - RADEEJ Morocco - RADEES 0 10 20 30 40 Percent Distribution utilities Non-MENA median MENA median Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 39 value of the third quartile for non-MENA countries. This is a variable that can be tackled directly and quickly and is an obvious target for any efforts to improve utility performance. On the lower end, most Moroccan DUs and SEC in Saudi Arabia appear to be good performers. Commercial Indicators Energy volume sold per connection. This indicator relates to the scale of operations rather than efficiency and is particularly related to the customer base composition (for example, residential, commercial, industrial). As the economy grows over time, industrialization and household incomes rise and the demand for electricity increases accordingly. If there are economies of scale at any stage of production, then average costs of supply fall, and this can be interpreted as a form of efficiency gain. Table 2.6 shows the values of energy sold per connection. Considering all DUs and VIUs, MENA median sales are larger than non-MENA, as are the sales of the Q1 group. However, for the Q3 group (the best performers), values are similar in both MENA and non-MENA. Table 2.6  Volume of Energy Sold per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers (kWh) Median (kWh) performers (kWh) Non-MENA 133 2,103 3,405 5,730 MENA 35 3,551 4,223 5,724 Source: World Bank calculations. Note: kWh = kilowatt-hours; MENA = Middle East and North Africa. Figure 2.5 illustrates the performance of the MENA DUs and VIUs with respect to energy sales per connection. Three VIUs—SEC in Saudi Arabia and DPC and RAECO in Oman—have values far greater than other VIUs or DUs; these are also higher than the median values both inside MENA and outside. This reflects the relatively high income levels of the consumers served by these utilities. Total billing per connection ($). This indicator measures scale effects: a higher ratio of billing to connections suggests that the utility’s operations are on a relatively sustainable path. Again, an important exogenous factor is the composition of the customer base. Everything else being equal, higher tariffs should be associated with higher billing per connection. There is no a priori reason to expect values to be different between VIUs and DUs. Table 2.7 lists the values for billing per connection. MENA utilities in the median quintile perform slightly better than non-MENA utilities and substantially worse in Q1 and Q3. The spread of values across both DUs and VIUs in the MENA region is shown in figure 2.6. RAECO in Oman performs very well, with a value above Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 40 Comparing the Region’s Performance with the Rest of the World Figure 2.5  Energy Sales Volume per Connection for Distribution and Vertically Integrated Utilities in MENA (kWh), 2013 (or most recent year with data, 2009–12) Saudi Arabia - SEC Oman - DPC Oman - RAECO Jordan - JEPCO West Bank - TUBAS Jordan - EDCO West Bank - JDECO Egypt, Arab Rep. - CEDC Algeria - SONELGAZ Morocco - ONEE Jordan - IDECO Lebanon - EdL Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - NCEDC Morocco - AMENDIS TAN West Bank - NEDCO Morocco - LYDEC Morocco - RADEEJ Morocco - RADEEMA Morocco - REDAL Tunisia - STEG Egypt, Arab Rep. - MEEDC Djibouti - EDD Egypt, Arab Rep. - AEDC Egypt, Arab Rep. - UEEDC Morocco - RAK Egypt, Arab Rep. - NDEDC Morocco - RADEEL Morocco - RADEEF Morocco - RADEM Yemen, Rep. - PEC Morocco - RADEES Egypt, Arab Rep. - SDEDC Morocco - AMENDIS TET 0 10,000 20,000 30,000 40,000 Kilowatt-hours Distribution utilities Non-MENA median MENA median Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: kWh = kilowatt-hours; MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 41 Table 2.7  Total Billing per Connection for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers ($) Median ($) performers ($) Non-MENA 72 199 292 531 MENA 27 135 299 439 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Figure 2.6  Total Billing per Connection for Distribution and Vertically Integrated Utilities in MENA ($), 2013 (or most recent year with data, 2009–12) Oman - RAECO Jordan - IDECO Lebanon - EdL Morocco - LYDEC Morocco - AMENDIS TAN Morocco - RADEEMA Morocco - REDAL Morocco - RADEEJ Tunisia - STEG Morocco - RADEEF Morocco - RAK Morocco - RADEES Morocco - RADEM Morocco - AMENDIS TET Egypt, Arab Rep. - CEDC Morocco - ONEE Iraq - MOE Yemen, Rep. - PEC Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - AEDC Egypt, Arab Rep. - UEEDC West Bank - JDECO Egypt, Arab Rep. - NDEDC Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - SDEDC 0 200 400 600 800 1,000 U.S. dollars Distribution utilities Non-MENA median MENA median Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 42 Comparing the Region’s Performance with the Rest of the World the non-MENA Q3. Most of Egypt’s DUs have very low values, for reasons that should be explored. Collection rate (percent). Failure to collect the total amount due is an impor- tant source of inefficiency, because it leads to deficits and the underfunding of future investment needs. Table 2.8 lists the collection rates of utilities inside and outside MENA. It is expected that DUs and VIUs should be able to achieve similar collection rates. The performance of MENA and non- MENA utilities is very similar at the median, Q1, and Q3 values. It might be noted—and it is surprising, given the importance of this indicator for Table 2.8  Collection Rates for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 15 88 94 97 MENA 15 85 92 94 Source: World Bank calculations. Note: MENA = Middle East and North Africa. overall efficiency—that only 15 MENA utilities shared the value for this key indicator. Insights into the poor performance of the weakest MENA utilities are offered by figure 2.7. Most VIUs and DUs have collection rates near the MENA median, but RAECO (Oman), Tubas (West Bank), and especially EDD (Djibouti) have very low collection rates. This points to clear weaknesses in these utilities. Financial Indicators OPEX recovery from sales (percent). This indicator measures the extent to which a utility is recovering operating expenditures from its sales of energy. Higher values indicate better performance, and it is interesting to note if the coverage is greater than 100 percent (full recovery). Table 2.9 presents the results for MENA and non-MENA utilities. DUs and VIUs cannot be compared directly because OPEX cannot be disaggregated by individual function using the MED. The MENA median value for DUs (93 percent) is below that for non-MENA (98 percent), but MENA VIUs have a higher median value than non-MENA VIUs. However, Q3 VIUs in MENA perform worse than those outside. This result may reflect the geographical composi- tion of the non-MENA utilities included in the analysis: most of the VIUs are in Sub-Saharan Africa, whereas the DUs are largely in Latin America. On the other hand, DUs in Q1 appear to be doing slightly better in MENA than beyond. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 43 Figure 2.7  Collection Rates for Distribution and Vertically Integrated Utilities in MENA (%), 2013 (or most recent year with data, 2009–12) Egypt, Arab Rep. - AEDC Bahrain - EWA Jordan - JEPCO Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - MEEDC West Bank - NEDCO Egypt, Arab Rep. - UEEDC Egypt, Arab Rep.- SCEDC Egypt, Arab Rep. - NDEDC Oman - RAECO West Bank - TUBAS Djibouti - EDD 0 20 40 60 80 100 120 Percent Distribution utilities Non-MENA median MENA median Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Table 2.9  OPEX Recovery as a Share of Sales (%) for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Number of Quartile 1: worst Median Quartile 3: best Utility type Region utilities performers (%) (%) performers (%) All MENA 32 85 92 99 Non-MENA 21 77 87 106 Distribution MENA 23 87 93 99 Non-MENA 5 81 98 103 Vertically integrated MENA 9 55 92 99 Non-MENA 16 77 87 109 Source: World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 44 Comparing the Region’s Performance with the Rest of the World Figure 2.8a focuses on DUs in MENA and presents the values for OPEX recovered from energy sales. Utilities in Morocco and Jordan appear to perform relatively well on this indicator, whereas those in Egypt and Oman are at the other end of the spectrum. But these results should be considered with caution. The surveys that feed the MED define energy sales as actual sales, without government transfers. However, some utilities seem to have included the government transfers in their responses. This makes comparison difficult and could explain differences across utilities. For example, looking at figure 2.8a, all Oman’s utilities—MEDC, Majan Electricity Company (MJEC), and MZEC—are well below the MENA median and, importantly, the breakeven point of 100 percent. But if government subsidies are included, these same three utilities’ OPEX recovery values rise to 117 percent, 118 percent, and 126 percent, respectively. Figure 2.8b presents OPEX recovery values for VIUs in MENA. The value of EDD in Djibouti is well above that of other utilities, whereas EdL in Lebanon is well below, for reasons that require further research to understand. The low value Figure 2.8  OPEX Recovery from Sales for Distribution and Vertically Integrated Utilities, MENA (%), 2013 (or most recent year with data, 2009–12) a. Distribution utilities b. Vertically integrated utilities Morocco - RADEEJ Morocco - RADEEMA Djibouti - EDD West Bank - NEDCO Jordan - IDECO Morocco - ONEE Morocco - REDAL Morocco - LYDEC Morocco - RADEEF Saudi Arabia - SEC Jordan - EDCO Morocco - RADEM Qatar -KAHRAMAA Morocco - RAK Jordan - JEPCO Egypt, Arab Rep. - SCEDC Algeria - SONELGAZ Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - EEDC Oman - DPC Morocco - RADEES Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - UEEDC Tunisia - STEG Morocco - RADEEL Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - AEDC Bahrain - EWA Oman - MEDC Oman - MJEC Lebanon - EDL Oman - MZEC 0 50 100 150 0 50 100 150 200 Percent Percent MENA median Non-MENA median Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 45 for Bahrain’s Electricity and Water Authority (EWA) can be explained by the fact that it did not include government transfers. Energy sales as a share of total costs (percent). This indicator is a direct measure of a utility’s ability to cover all its costs. Values less than 100 percent indicate that total costs are not being recovered for various reasons, all of which can be described as inefficiency. DUs and VIUs cannot be directly compared because the distribution costs of VIUs cannot be separated out from other cost categories. Table 2.10 presents results for MENA and non-MENA utilities. The sample size of non-MENA utilities is notably small: 8 in all categories. The MENA sample of 19 is also small when compared with those used for other indicators. At 67 percent, the median value for non-MENA DUs is well below the cost-recovery level; MENA’s is closer, at 88 percentage points. It should be noted, however, that there were only two non-MENA observations, so the values for non-MENA utilities could be underestimates. The VIUs inside MENA are less efficient than those outside MENA, as can be observed by their median values. Table 2.10  Energy Sales as a Share of Total Costs (%) for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: share Quartile 3: share Number of of worst of best Utility type Region utilities performers (%) Median (%) performers (%) All MENA 19 74 87 91 Non-MENA 8 45 67 79 Distribution MENA 12 82 88 93 Non-MENA 2 54 67 79 Vertically integrated MENA 7 43 56 80 Non-MENA 6 50 67 73 Source: World Bank calculations. Note: MENA = Middle East and North Africa. The performance of individual DUs in MENA is shown in figure 2.9a. Only the Northern Electricity Distribution Company (NEDCO) in the West Bank and the El Jadida municipal utility (RADEEJ) in Morocco have a sales-to-total-cost ratio greater than 1. Moroccan DUs tend to perform better, whereas the majority of Egyptian utilities are the furthest ­ from total cost recovery. Figure 2.9b shows the performance of VIUs in MENA. EDD in Djibouti is the only utility covering total costs, whereas SEC in Saudi Arabia and EdL in Lebanon cover only a small fraction of total costs. Accounts receivable to sales (days). This indicator measures the time it would take, at current sales levels, to collect all bills. It is used to estimate the number of times a utility is able to convert its credit sales to cash during a financial year. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 46 Comparing the Region’s Performance with the Rest of the World Figure 2.9  Sales as a Share of Total Costs for Distribution and Vertically Integrated Utilities, MENA (%), 2013 (or most recent year with data, 2009–12) a. Distribution utilities b. Vertically integrated utilities Morocco - RADEEJ Djibouti - EDD West Bank - NEDCO Jordan - IDECO Morocco - ONEE Morocco - REDAL Egypt, Arab Rep. - NDEDC Oman - DPC Morocco - LYDEC Algeria - SONELGAZ Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - NCEDC Tunisia - STEG Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - EEDC Saudi Arabia - SEC Egypt, Arab Rep. - MEEDC Lebanon - EdL Egypt, Arab Rep. - UEEDC 0 50 100 150 0 20 40 60 80 100 120 Percent Percent MENA median Non-MENA median Source: MENA Electricity Database (MED) and World Bank calculations. Note: MENA = Middle East and North Africa. Table 2.11  Ratio of Accounts Receivable to Sales in MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: best Quartile 3: worst Region Number of utilities performers (days) Median (days) performers (days) Non-MENA 7 8 52 140 MENA 26 117 148 202 Source: World Bank calculations. Note: MENA = Middle East and North Africa. The higher the indicator is, the higher collection efficiency and the higher the utility’s liquidity value.3 Table 2.11 presents estimated values inside and outside MENA (where the sample size is very small, at 7). The MENA median value is 148 days, whereas that of non-MENA utilities is only 52 days. If these figures are representative, then it appears that MENA performs relatively poorly on this Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 47 indicator. The difference is even more stark for the best-performing utilities: the Q1 value for non-MENA utilities is 8.4 days. The ratios of accounts receivable to sales among MENA utilities are shown in figure 2.10. The large majority have values greater than 100 days. All utilities in MENA have values above the non-MENA median. Figure 2.10  Accounts Receivable to Sales for Distribution and Vertically Integrated Utilities Utilities in MENA (days), 2013 (or most recent year with data, 2009–12) Oman - RAECO Egypt, Arab Rep. - SCEDC West Bank - TUBAS Oman - DPC Morocco - RADEEMA Saudi Arabia - SEC Bahrain - EWA Djibouti - EDD Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - EEDC West Bank - NEDCO Morocco - ONEE Qatar - KAHRAMAA Jordan - JEPCO Oman - MEDC Morocco - REDAL Jordan - IDECO Oman - MJEC Jordan - EDCO Egypt, Arab Rep. - MEEDC Oman - MZEC Morocco - RADEEJ Tunisia - STEG Egypt, Arab Rep. - AEDC Morocco - LYDEC 0 50 100 150 200 250 300 350 400 Percent Distribution utilities Non-MENA median MENA median Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 48 Comparing the Region’s Performance with the Rest of the World Debt to equity. A high debt-to-equity ratio indicates an aggressive growth- financing approach to debt. Risks to this approach include the cost of additional interest expenses and any volatility if the debt is short to medium. If the cost of debt financing outweighs the returns generated by the additional capital, the financial load quickly becomes an issue, whether the utility is publicly or pri- vately owned. This is why a common rule of thumb is to cap the ratio at 2. Table 2.12 presents the values of the ratio for MENA and non-MENA utili- ties. With no a priori reason to argue that the debt-to-equity ratio should be different across utility types, no distinction is made in this table among VIUs, DUs, GUs, and TUs. The MENA median value is 357 percent whereas the non- MENA median, based on a small sample, is only 91 percent. The very high MENA value suggests an excessive reliance on debt financing. Table 2.12  Ratio of Debt to Equity for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: best Quartile 3: worst Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 14 62 91 443 MENA 47 207 357 767 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Data are for all utility types: vertically integrated, distribution, generation, and transmission. Figure 2.11 shows the spread of MENA values by utility type. The distribution is uneven. Several utilities, in particular GUs in Egypt and Oman, have debt-to- equity ratios over 10:1. Understanding the reasons behind these surprisingly high values would require detailed investigation, though the case studies of Egypt and Oman in part II of this book offer some insights. In 2013, GUs suffered from an average debt-to-equity ratio of the order of 20:1. Corporate governance should be improved to ensure these utilities’ restruc- turing, with the aim of long-term sustainability. In the meantime, two concrete actions could be taken: first, raise these utilities’ equity by converting public debt into equity; and second, reform tariffs (see, for example, the many suggestions made by Egypt’s regulator, ERA). Current assets to current liabilities (percent). The ratio of current assets to current liabilities measures the extent to which short-term assets (cash, cash equivalents, marketable securities, and receivables) are readily available to pay off short-term liabilities (payables, current portion of term debt, accrued expenses, and taxes). Generally, the higher the ratio, the better. The values for all types of MENA and non-MENA utilities are shown in table 2.13. The median Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 49 Figure 2.11  Ratio of Debt to Equity across Utility Types in MENA (%), 2013 (or most recent year with data, 2009–12) Egypt, Arab Rep. - EDEPC Morocco - ONEE Egypt, Arab Rep. - WDEPC Egypt, Arab Rep. - MDEPC Oman - BPDP Jordan - EDCO Oman - SPP Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - UEEPC Jordan - IDECO Jordan - SEPCO Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - CEDC Egypt, Arab Rep. - NDEDC Jordan - QEPCO Tunisia - STEG Jordan - JEPCO Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - MEEDC Algeria - SONELGAZ Saudi Arabia - SEC Oman - SSPWC Jordan - CEGCO Jordan - AES PSC Oman - RAECO Oman - DPC Oman - ABPC Oman - ASPC Jordan - AAEPC Morocco - LYDEC West Bank - JDECO Oman - APBS Djibouti - EDD Oman - OETC Oman - MZEC Oman - MEDC Jordan - NEPCO Oman - MJEC Oman - AKPP West Bank - NEDCO Qatar - KAHRAMAA Oman - UPC Bahrain - EWA Morocco - RADEEJ Morocco - RADEEMA 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 Percent Distribution utilities MENA median Non-MENA median Generation utilities Transmission utilities Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 50 Comparing the Region’s Performance with the Rest of the World non-MENA value (88 percent) and the median MENA value (84 percent) are similar, and both raise concerns because they are below 100 percent. The Q1 and Q3 values are also similar both inside and outside MENA. Table 2.13  Ratio of Current Assets to Current Liabilities for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 19 62 88 119 MENA 53 63 84 121 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Figure 2.12 plots the ratio of current assets to current liabilities across all types of utilities in MENA. Several GUs—the Wadi Al Jizzi Power Company (WAJPCO) in Oman, Qatrana Electric Power Company (QEPCO) in Jordan, Al-Ghubra Power & Desalination Company (GPDCO) in Oman, and Amman East Power Plant (AES PSC) in Jordan—boast a ratio of 3:1, indicating a strong financial position. On the other hand, a majority of utilities (33 out of 53) falls below the 1:1 threshold, and nine have a ratio of less than 1:2, indicating a weak financial position. Return on assets (percent). ROA is a measure of profitability. The higher the value, the better the performance. However, comparisons across markets need to recognize that riskier markets will require higher returns. Table 2.14 presents the results for MENA and non-MENA utilities. The sample sizes are very small for all the non-MENA categories, but some comparisons can be made. Across all utilities, the median ROA for the non-MENA group is 1 percent, whereas for MENA it is 3 percent. Although the same pattern is observed for Q1, in Q3 MENA utilities appear to have lower ROA than non- MENA ones. All these values are low and suggest generally weak financial performance. Figure 2.13 illustrates the ratios of individual VIUs in MENA. The value of Lebanon’s VIU, EdL, which has an ROA value of −150 percent, is not repre- sented. The Jordan Electric Power Company (JEPCO), and Morroco’s Régie Table 2.14  Return on Assets for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 12 −2 1 9 MENA 49 0 3 6 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 51 Figure 2.12  Ratio of Current Assets to Current Liabilities: Selected Utilities of All Types, MENA (%), 2013 (or most recent year with data, 2009–12) Oman - WAJPCO Jordan - QEPCO Oman - GPDCO Jordan - AES PSC West Bank - NEDCO Djibouti - EDD Qatar - KAHRAMAA Oman - SSPWC Oman - ARPP Algeria - SONELGAZ Oman - RAECO West Bank - JDECO Jordan - AAEPC Oman - APBS Oman - SPP Jordan - SEPCO Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - EEDC Jordan - EDCO Egypt, Arab Rep. - NDEDC Jordan - CEGCO Morocco - REDAL Tunisia - STEG Saudi Arabia - SEC Egypt, Arab Rep. - MEEDC Bahrain - EWA Jordan - IDECO Egypt, Arab Rep. - SCEDC Jordan - JEPCO Oman - AKPP Egypt, Arab Rep. - AEDC Yemen, Rep. - PEC Morocco - LYDEC Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - MDEPC Egypt, Arab Rep. - WDEPC Egypt, Arab Rep. - CEDC Morocco - RADEEJ Morocco - ONEE Egypt, Arab Rep. - UEEPC Oman - ABPC Oman - ASPC Egypt, Arab Rep. - EETC Oman - DPC Oman - MEDC Oman - MJEC Oman - BPDP Oman - UPC Egypt, Arab Rep. - EDEPC Oman - MZEC Lebanon - EDL Egypt, Arab Rep. - CEPC 0 100 200 300 400 500 600 Percent Distribution utilities MENA median Non-MENA median Generation utilities Transmission utilities Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 52 Comparing the Region’s Performance with the Rest of the World Autonome de Distribution d’Eau et d’Électricité de Meknès (RADEM) and Régie Autonome Intercommunale de Distribution d’Eau et d'Électricité de Safi (RADEES) have the highest ROAs, with values in excess of 10 percent. At the low end, JDECO (West Bank) stands out with a very large negative return. This needs to be investigated further to understand any special factors. AMENDIS TET (Morocco), STEG (Tunisia), Socièté Nationale de l’Electricité et du Gaz (SONELGAZ) (Algeria), and Office National de l’Electricité et de l’Eau Potable (ONEE) (Morocco) also have negative ROAs. It appears that VIUs tend to be at the lower end of the distribution, with almost all of them below the MENA median. Return on equity (ROE) (percent). This indicator measures the return on shareholders’ investments. Table 2.15 compares the values for MENA and non- MENA utilities. Again there are few observations for the non-MENA economies. The median MENA value of 6 percent is well above the non-MENA value of 0 percent, and there are similar differences in performance between the Q3 and Q1 utilities. Figure 2.14 plots the values for the individual MENA utilities. The value of Morocco’s VIU, ONEE, which has a ROE value of −127 percent, is not repre- sented. The top seven performers achieved an ROE of 10 percent or better, Table 2.15  Return on Equity for MENA and Non-MENA Utilities, 2013 (or most recent year with data, 2009–12) Quartile 1: worst Quartile 3: best Region Number of utilities performers (%) Median (%) performers (%) Non-MENA 13 −4 0 6 MENA 46 0 6 16 Source: World Bank calculations. Note: MENA = Middle East and North Africa. whereas four utilities had negative values. The markedly negative values for STEG (Tunisia) and JDECO (West Bank) require further research. GUs appear to be at the upper end of the ROE DUs, and VIUs, once again, are at the lower end. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 53 Figure 2.13  Return on Assets: Selected Utilities of All Types, MENA (%), 2013 (or most recent year with data, 2009–12) Jordan - JEPCO Morocco - RADEM Morocco - RADEES Jordan - CEGCO Oman - AKPP Oman - WAJPCO Oman - MEDC Oman - MJEC Oman - APBS Oman - OETC West Bank - TUBAS Oman - MZEC Morocco - RADEEL Jordan - IDECO Jordan - QEPCO Oman - UPC Jordan - EDCO Jordan - SEPCO West Bank - NEDCO Oman - BPDP Oman - RAECO Oman - SSPWC Oman - SPP Morocco - AMENDIS TAN Egypt, Arab Rep. - SCEDC Morocco - REDAL Egypt, Arab Rep. - CEDC Saudi Arabia - SEC Oman - GPDCO Bahrain - EWA Egypt, Arab Rep. - UEEPC Egypt, Arab Rep. - NDEDC Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - AEDC Oman - PPC Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - MDEPC Egypt, Arab Rep. - CEPC Egypt, Arab Rep. - EDEPC Egypt, Arab Rep. - WDEPC Morocco - AMENDIS TET Algeria - SONELGAZ Tunisia - STEG Morocco - ONEE West Bank - JDECO –30 –20 –10 0 10 20 30 Percent Distribution utilities MENA median Non-MENA median Generation utilities Transmission utilities Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. The value of EdL (Lebanon) which has a value of −150% is not represented on this graph for a matter of scale and representation. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 54 Comparing the Region’s Performance with the Rest of the World Figure 2.14  Return on Equity for Selected Utilities of All Types in MENA (%), 2013 (or most recent year with data, 2009–12) Jordan - AES PSC Jordan - QEPCO Oman - APBS Morocco - RADEM Jordan - CEGCO Oman - OETC Jordan - IDECO Morocco - LYDEC Jordan - SEPCO Oman - MEDC Jordan - EDCO Morocco - RADEES Oman - AKPP Oman - MZEC Oman - MJEC Oman - SSPWC Jordan - JEPCO Oman - RAECO Morocco - REDAL Egypt, Arab Rep. - SCEDC Egypt, Arab Rep. - CEDC Oman - UPC Morocco - RADEEL Saudi Arabia - SEC West Bank - NEDCO Egypt, Arab Rep. - UEEPC Morocco - AMENDIS TAN Oman - WAJPCO Bahrain - EWA Egypt, Arab Rep. - NDEDC Egypt, Arab Rep. - NCEDC Egypt, Arab Rep. - CEPC Egypt, Arab Rep. - SDEDC Egypt, Arab Rep. - MDEPC Egypt, Arab Rep. - EDEPC Egypt, Arab Rep. - AEDC Oman - GPDCO Egypt, Arab Rep. - UEEDC Egypt, Arab Rep. - MEEDC Egypt, Arab Rep. - EEDC Egypt, Arab Rep. - WDEPC Morocco - AMENDIS TET Algeria - SONELGAZ West Bank - JDECO Tunisia - STEG –40 –20 0 20 40 Percent Distribution utilities MENA median Non-MENA median Generation utilities Transmission utilities Vertically integrated utilities Source: MENA Electricity Database and World Bank calculations. Note: MENA = Middle East and North Africa. The value of ONEE (Morocco) which has a value of −127% is not represented on this graph for a matter of scale and representation. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Comparing the Region’s Performance with the Rest of the World 55 Conclusion Calculating the median values of various performance indicators allows a com- parison between MENA and non-MENA electricity utilities. Despite the chal- lenges of data availability and comparability, it has been possible to indicate where the utilities of the MENA region fall behind—or exceed—the performance of comparable utilities elsewhere. When the values of indicators for individual utilities in MENA are plotted against the MENA and non-MENA medians, some valuable insights can be gained. This approach allows outlying values to be identified so that further research can be focused on these cases. It also highlights indicators for which there is relatively little variation across utilities in the MENA region and those for which there is a large gap between best and worst performers (even after excluding clear outliers). Policy makers concerned with the performance of indi- vidual utilities in their economies can use these tools to set realistic targets for improvement and can monitor progress toward specific objectives. One interesting feature of the comparisons within MENA is that there were no immediately obvious “best” or “worst” performers among utilities across all indicators. An approach to measuring performance across several indicators is presented in chapter 4. Notes 1. The non-MENA data were drawn from 38 vertically integrated utilities, 4 generation utilities, 135 distribution utilities, and 4 transmission utilities that are found in appendix B. The list of non-MENA utilities can also be found in appendix B. 2. All indicator values expressed in monetary terms are converted to U.S. dollars at sur- vey year exchange rates. For intercountry comparisons, valuations at purchasing power parity (PPP) could well affect the relative magnitudes of the MENA and non-MENA indicators. 3. The average collection period is computed as the number of days in that period divided by the accounts-receivable-to-sales ratio. A value of 6 means that the average client pays once every two months. This implies an average collection period of 60 days. Shortening the collection period reduces the working capital cycle and often eases access to bank loans for needed investments. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 3 A Dynamic Look at MENA Performance over Five Years Comparing performance indicators over time is of interest when assessing whether utilities are improving their efficiency or not. Where governments have introduced power sector reforms, policy makers might examine the effects using certain indicators. Changes are expected to be gradual rather than sudden and may take several years to see. Other factors, apart from government policy, may impact indicators. Local demand, the international economic climate (including oil price fluctuations, for example), and political issues can all influence utility performance. Data Challenges In the survey that informs the Middle East and North Africa (MENA) Electricity Database, utilities were asked for information on several years, from 2009 to 2014, not just the base year (2013). Given that less data were obtained for 2014 than for 2009–13, we have not included that year in the analysis summarized in most of this book. This chapter is an exception, given its focus on the dynamic aspects of performance. The large number of indicators included in the MENA Electricity Database (36 core indicators) and the large number of utilities sur- veyed (67) meant that considering data at a utility level would require nearly 3,000 trend calculations to be made. When the data series are so short, and with inevitable questions concerning data accuracy, such an exercise would not be sensible. An alternative is to consider constructing aggregates across utilities, indicator by indicator, and to carry out trend analysis on these aggregates for the few years of data available. Preliminary examination of the data revealed that the coverage of each indica- tor is only partial, even for the base year, and utilities supplied data for different years within the six-year span requested. Few provided information for all 2009–14. This meant that it was not possible to set up a standard comparison between indicators or between utilities. Shedding Light on Electricity Utilities in the Middle East and North Africa   57   http://dx.doi.org/10.1596/978-1-4648-1182-1 58 A Dynamic Look at MENA Performance over Five Years The first step in the analysis was to compare average indicator values for each year based on (a) only those utilities that provided data for at least five of the six years (see table 3.1) and (b) all utilities answering for that year, this number vary- ing from year to year (see table 3.2). This comparison was made for four groups of utilities: all utilities, vertically integrated utilities (VIUs), distribution utilities (DUs), and generation utilities (GUs). The median values were then compared within each group to avoid the impacts of extreme values caused by issues in data collection, which might exist for only one year. To carry out this analysis we started concentrating on a single indicator of substantial importance to performance: the ratio of current assets to current liabilities, which was collected for at least five years by 35 utilities (seven VIUs, 15 GUs, and 13 DUs). The median values for each year, by utility type, based on these respondents (the common sample) are shown in table 3.1. The median values, based on all respondents for each year (that is, more than 35 utilities for each year) are shown in table 3.2. Tables 3.1 and 3.2 show a significant decline in most values as of 2014, which is caused by the sudden change in the number of utilities reporting for that year. The smaller sample in 2014 consisted of utilities that tended to have the lowest values in the sample, thus pulling the average down. Thus, 2014 should be excluded from any analysis in which the number of reporting utilities is notably smaller than for the rest of the years. Using all available data points on each year for time series analysis would produce movements that are due largely to the inclusion or exclusion of some utilities (that is, those who provided information for only four years or fewer).1 In addition, one utility’s performance on a given Table 3.1  Median Values of Ratio of Current Assets to Current Liabilities for Utilities (%), 2009–14 Minimum of five observations Type of utility 2009 2010 2011 2012 2013 2014 All 95 100 92 82 84 47 Vertically integrated 125 105 119 75 89 43 Generation 107 123 127 118 113 47 Distribution 93 97 89 91 80 37 Source: World Bank calculations. Note: The median samples are based on a common sample. Table 3.2  Median Values of Ratio of Current Assets to Current Liabilities for Utilities (%), 2009–14 Type of utility 2009 2010 2011 2012 2013 2014 All 95 105 96 87 85 52 Vertically integrated 111 105 96 75 109 76 Generation 93 100 100 107 87 105 Distribution 89 98 86 83 81 39 Source: World Bank calculations. Note: Observations not necessarily available for all utilities for a given year. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Dynamic Look at MENA Performance over Five Years 59 indicator could fluctuate widely over time, possibly due to problems of data col- lection and recording. Indicator Trends with All Utilities Aggregated Our preliminary findings led us to begin with sets of common data and analyze the behavior of the median between 2009 and 2014, measured over all utilities in the common dataset for one indicator. With this dataset, tests were carried out to look for the existence of a significant trend by regressing the log of the indicator value in each year on a time trend variable (increasing by one unit each year). If the coef- ficient of the trend term was not significantly different from zero, then we con- cluded there was no trend in the indicator, so it effectively remained constant. Table 3.3 estimates the trend value and the probability of significance of the 25 indicators for which there were adequate data, meaning that the proportion of missing observations for 2014 relative to the number of utilities was small. Table 3.3  Estimated Trend of Indicators for Utilities, 2009−14 Number of utilities Estimated trend Indicator with common data value Probability Availability factor 6 0.004 0.34 Capacity factor 17 0.010 0.57 Load factor 15 −0.002 0.73 Percentage of meters replaced 11 −0.110 0.12 Network maintenance 12 −0.070 0.49 OPEX/employee 2 0.100 0.15 OPEX/connection 26 0.150 0.04* OPEX/km 29 −0.010 0.74 Residential connections/employee 13 −0.150 0.12 Sales/employee 24 −0.020 0.69 Revenue/employee 25 −0.010 0.69 Fuel/OPEX 15 0.010 0.84 Energy purchase/OPEX 35 0.010 0.29 Labor costs/OPEX 26 −0.060 0.01* Sales/OPEX 38 −0.020 0.01* Accounts receivable/sales 34 0.020 0.52 Debt/equity 38 0.050 0.11 Current assets/current liabilities 35 −0.040 0.08 Return on assets 39 −0.040 0.70 Return on equity 42 0.090 0.29 Total billing/connection 19 0.110 0.09 Collection rate 15 0.010 0.55 Prepaid meters installed (%) 6 0.250 0.10 Distribution losses 27 0.040 0.22 SAIFI 10 0.020 0.51 Source: World Bank calculations. Note: km = kilometer; OPEX = operating expenses; SAIFI = System Average Interruption Frequency Index. Significance level: * = 5%. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 60 A Dynamic Look at MENA Performance over Five Years Using a two-sided test, which would accept evidence for either an increasing or a decreasing trend based on five years of data, the probability should be less than 0.05 to reject the hypothesis of no trend in the data. Three indicators showed significant time trends during the period studied, when measured against aggregate data (median values): (a) operating expenses (OPEX) per connection, which had a 15 percent annual growth rate; (b) labor costs to OPEX, with a negative annual growth rate of 6 percent; and (c) sales to OPEX, with a negative annual growth rate of 2 percent. These figures suggest that OPEX increased significantly throughout the region between 2009 and 2014, although the number of connections increased only slowly, and sales and labor costs increased at a moderate rate. Because the main components of OPEX are fuel and labor costs, the very large increase in oil prices at the beginning of the period2 is likely to have influenced these trends. The increase in the value of OPEX might also be attributable to an increase in wages or to a renewed empha- sis on maintenance and repair. A detailed breakdown of OPEX would be needed to understand the reasons for these trends. The growth rates estimated for the highly aggregated data are nearly all zero, which suggests that there are few regionwide trends that can be identified with such a short run of data. Indicator Trends Disaggregated by Utility Type Some patterns may be observable when a more disaggregated approach is used. This is illustrated by the ratio of current assets to current liabilities. Table 3.4 contains the estimated growth rates for the indicator based on data from utilities with a common sample of all observations from 2009 to 2013. Median values are constructed for VIUs, DUs, and GUs. The table shows that the groups produced different values of the estimated trend, but only the generators produce a signifi- cant positive annual trend of 8 percent. Therefore, we note that trends can differ among types of utilities for reasons connected with their nature. The same analysis was undertaken for the remaining indicators from table 3.3.3 Four additional disaggregated trends were found to be significant for a particular type of utility (table 3.5). Table 3.4  Estimated Trends for Median Ratio of Current Assets to Current Liabilities, by Utility Type, 2009–13 Utility type Number of observations Trend value Probability All 35 −0.04 0.08 Vertically integrated 7 −0.10 0.14 Distribution 13 −0.03 0.11 Generation 15 0.08 0.01* Source: World Bank calculations. Note: The estimated trends were based on a common sample. Significance level: * = 5%. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Dynamic Look at MENA Performance over Five Years 61 Table 3.5  Estimated Trends, by Utility Type, 2009–13 Utility type Number of observations Trend value Probability a. Median capacity factor All 17 0.014 0.570 Vertically integrated 3 −0.390 0.040* Generation 14 0.012 0.570 b. Median OPEX per employee All 24 0.110 0.150 Vertically integrated 7 0.087 0.520 Distribution 7 −0.180 0.120 Generation 8 0.470 0.002* c. Median sales per employee All 24 −0.017 0.690 Vertically integrated 6 0.220 0.090 Distribution 7 −0.260 0.070 Generation 9 0.380 0.020* d. Return on assets All 39 −0.047 0.700 Vertically integrated 6 −0.009a 0.010* Distribution 18 −0.260 0.140 Generation 11 −0.018 0.790 Source: World Bank calculations. Note: The estimated trends were based on a common sample. OPEX = operating expenses. Significance level: * = 5%. a. The number is a linear trend because of negative values for vertically integrated utilities. The capacity factor of VIUs appears to be declining over time, but caution should be taken when interpreting this result given the very small sample size. For GUs, OPEX per employee grew 47 percent a year—probably driven by the very large increase in oil prices at the beginning of the period—and total sales (in monetary value) per employee grew by 38 percent a year. This could reflect the pass-through of oil price variations from generators to the VIUs or the transmission utilities (TUs) buying their electricity, translated into increased sales due to the increase in oil prices during the period of interest. Finally, the return on assets (ROA) of VIUs appears to have decreased slightly over time (by 0.9 percent). To further disaggregate data for trend analysis, we use the ratio of current assets to current liabilities for the large group of nine GUs in Oman that all provided full data for this indicator (2009–13). This group should be free of the large intercoun- try differences caused by differences in policies and economic conditions, allowing common trends to be identified. Table 3.6 presents the results of a trend analysis carried out for each utility and for the median of the group. The results in table 3.6 indicate that for only one of the GUs (the Al-Ghubra Power & Desalination Plant, GPDCO) was there a significant trend in the ratio of current assets to current liabilities: an estimated decline of 27 percent per year. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 62 A Dynamic Look at MENA Performance over Five Years Table 3.6  Estimated Trends for Ratio of Current Assets to Current Liabilities for Generation Utilities in Oman, 2009–13 Utility Trend value Probability ACWA Power Plant 0.130 0.21 Al-Ghubra Power & Desalination Plant −0.270 0.01* Al-Kamil Power Plant −0.070 0.20 Al-Rusail Power Plant 0.060 0.08 Barka Power and Desalination Plant −0.160 0.25 Sembcorp Salalah Power and Water Co. 0.150 0.74 Sohar Power Plant −0.001 0.99 United Power Company −0.210 0.09 Wadi Al-Jizzi Power Plant −0.100 0.41 All generation utilities 0.080 0.01* Source: World Bank calculations. Note: The estimated trends were based on a common sample. Significance level: * = 5%. Such a steep decline suggests that there must have been special circumstances behind the data, requiring further investigation. The other nine utilities have insignificant trends, but there is a significant ­ positive trend for the aggregate of all GUs in Oman. This must be affected by the rapid change in this indicator’s values for some of the generation utilities in Oman. Conclusion The scale of the survey, covering 67 utilities and 36 indicators, prohibited a trend analysis at the utility level. Furthermore, the small number of years for which data were collected (a sample of typically four years within the period 2009–13) implies that trends had to be very well marked to be statistically significant and that errors in data collection could negate apparent results. As an alternate to trend analysis utility by utility, aggregation across utilities was explored, in which the average for a group of utilities for each year was then sub- jected to trend analysis. It was shown that the averages should be based on the same utilities in each year (the common sample) because averages based on all data available for each year were very sensitive to gaps in data for some utilities in some years. Trend growth rates were fitted to each indicator for which there were ade- quate data, based on median values for those utilities in the common sample (that is, that provided information for all years). Only three indicators exhibited signifi- cant trends: OPEX per connection exhibited a large positive trend, whereas energy sales over OPEX and labor costs over OPEX exhibited negative trends. These were likely due to the dramatic increase in oil prices during the period. The level of aggregation used for these tests allowed a small-scale investiga- tion, but too much aggregation can conceal common trends between sub- groups. A single indicator (current assets to current liabilities) was used to explore the effects of disaggregation. Disaggregating to the level of utility type Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Dynamic Look at MENA Performance over Five Years 63 (vertically integrated, distribution, generation) within the MENA region revealed a significant positive trend for generators as a group, although the other groups did not have significant trends. This result adds support to the idea that trend analysis, where there are sufficient data, should be carried out for the different types of utilities. We then undertook the same exercise for all the other indicators, and found statistically significant results for the capacity factor, OPEX per employee, sales per employee, and ROA. Further disaggregation to the level of individual GUs in Oman was carried out for a set of nine such utilities. Conditions within Oman were similar for all utili- ties, and policies applied equally. One utility exhibited a significant negative trend, so large as to require further investigation. The other eight utilities showed no significant trends, and it was clear that the data series were generally too short to pick up trends in performance. Aggregation, aimed at smoothing out random shocks, also did not reveal much in the way of regionwide trends. This analysis of the performance of MENA utilities suggests that to identify underlying trends in performance (if any), substantially longer time series of data would be required. Also, analysis should be carried out at a utility level or for aggregates of utilities over all the years analyzed. Taking yearly averages over a varying number of utilities is likely to produce large swings in the aggregate due to its composition. Notes 1. A similar analysis was carried out using the means rather than medians of the current assets to current liabilities ratios. The same conclusions were reached—and indeed reinforced by noting that changing the size of the sample between years allowed cer- tain extreme observations to dominate the data and produce large fluctuations between years. 2. In 2009, the average price of Brent crude oil was $62 a barrel (bbl); in 2010, it was $80 per bbl; and in 2010, it was $111 per bbl. It fell to $108 per bbl in 2013 and then modestly to $99 per bbl in 2014, which is still 60 percent higher than at the beginning of this period. 3. Only results generating a significant trend for some category of utilities are presented here. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 4 A Multi-Indicator Approach to Analyzing Utility Performance Chapter 2 ranked 67 utilities across the Middle East and North Africa (MENA) on several indicators and identified utilities whose performance on a particular indicator was very strong or very weak. Understanding the reasons behind these performance levels can inform policies to be applied elsewhere. Two features of the analysis stand out. First, because of data gaps it was not always possible to compare utilities on more than one indicator. Second, even when a utility provided data on multiple indicators, its performance was not consistently high (or low) across them. The use of a single indicator to assess relative performance offers a somewhat limited view, dominated by the specific characteristics of the utilities surveyed. Some form of average performance measure is required to reveal overall strengths and weaknesses. To this end, we compared a set of distribution utilities (DUs) against one another, using the same indicators for every utility in the set. This avoided the challenges that necessarily arise when comparing different types of utilities and different indicators. Methodology A performance assessment of multiple indicators reflects a wider range of a utility’s characteristics and reduces the chance of results being decided by pecu- liar circumstances. This methodology is useful given that the quality of our data is not sufficient to use more sophisticated approaches (for example, stochastic frontier analysis or data envelopment analysis). The challenge of this multi-­ indicator approach is combining indicators measured across very different con- texts. An average rank score addresses this problem. For example, suppose there are 10 utilities, all of which reported data on their return on equity (ROE) and sales-to-employee ratio. The utility with the highest ROE would receive a score of 10 for that indicator, the next-best utility a score of 9, and so on. The sales-to- employee ratio would be treated similarly. A combined performance measure Shedding Light on Electricity Utilities in the Middle East and North Africa   65   http://dx.doi.org/10.1596/978-1-4648-1182-1 66 A Multi-Indicator Approach to Analyzing Utility Performance would then be the average rank value for each utility. This approach can be gen- eralized to include as many indicators as necessary. The average rank indicator allows indicators based on different measurement units to be combined, but in doing so becomes a purely relative measure of performance. It does not distinguish between large and small actual differences across successive observations, and all indicators are of equal importance. If new data become available for other utilities, they can easily be added into the ranking so new comparisons are possible. An important feature of rank-based indicators is that they are robust against all but very large measurement errors in the original data. It is interesting to note the extent to which a utility’s rank order is similar or not for different indicators. For example, if its ranking is similar for both the ROE and the sales-to-employee ratio, this indicates that the utility tends to be strong or weak across the board. Meanwhile, very different ranks suggest there is little evidence for calling a particular utility strong or weak across all dimensions. The degree of agreement across indicators can be measured using Kendall’s coefficient of concordance (W).1 The maximum value of the coeffi- cient of concordance is unity, and the minimum is zero. At the maximum value, any single indicator would give the same performance ranking as an average of all indicators. The nearer to unity the W coefficient, the less the need to take an average of several indicators to produce an overall ranking of performance. For cases where there are 5 or more utilities or more than 15 indicators, a test of the null hypothesis (that is, that there is no agreement between the rankings of the different indicators) can be carried out.2 A high W value indicates that some utilities are more interested in efficiency than others and that they tend to look for improvements in several aspects of performance. If they pursued all avenues toward efficiency with equal effort, but the degree of effort varied among the utilities, then the concordance would be unity, with each utility achieving the same rank for every one of a set of indica- tors (and different from those of all other utilities). However, if they pursue all avenues toward efficiency with different degrees of effort, the correlations between the ranks will drop and W will move toward zero. In an ideal situation, all utilities would pursue all avenues toward efficiency at the same time and with the same intensity so that W would quickly move toward unity. In practice, it may be easier to focus on just one or two areas of performance, and different utilities may prioritize different target indicators accordingly. Data Considerations The issues of data availability and inconsistent ranking can be illustrated using material presented in chapter 2. One DU, the Muscat Electricity Distribution Company (MEDC) in Oman, is used for this purpose, but similar results would have been obtained for other utilities. Table 4.1 gives the rank score for MEDC (over all DUs for which there were data) for all indicators used in the global comparison exercise. Ranking is from the worst performer (1) to the best (value equaling sample size). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Multi-Indicator Approach to Analyzing Utility Performance 67 Table 4.1  Ranked Performance of MEDC (Oman) on Various Indicators Indicator Performance rank Sample size (value of maximum rank) OPEX/connection 1 25 OPEX/kWh — — Connections/employee 17 19 Distribution losses 13 28 Sales/connection — — Billing/connection — — Collection rate — — Sales/OPEX 3 23 Sales/costs — — Accounts receivable/sales 11 18 Debt/equity 15 19 Current assets/current liabilities 3 20 Return on assets 13 24 Return on equity 15 24 Source: World Bank calculations. Note: kWh = kilowatt-hours; MEDC = Muscat Electricity Distribution Company; OPEX = operating expenses; — = not available. For several of the indicators used for the global comparison there are no data for MEDC, and it cannot be compared to other DUs on these dimen- sions of performance. For indicators where there are data, the performance is variable. For example, it is high for return on assets (ROA) and ROE, but low for the ratios of assets to liabilities and sales to operating expenses (OPEX). Judging its performance on any single indicator could result in a misleading picture of overall performance, so a measure across several dimensions may be preferred. The data gaps of the MENA Electricity Database represent significant obstacles to this approach, or any approach based on a number of individual indicators. The database covers a large range of indicators (36) for the 67 utili- ties, but for many a full set of data was unavailable. The data gaps are different between types of utilities so that the larger the number of indicators considered for the average rank score, the fewer utilities would have data for all the indica- tors. A balance has to be struck between comparing performance across a large number of utilities (and fewer indicators) and comparing performance across a wide range of indicators to provide a more balanced assessment (using fewer utilities). In addition to data availability, indicators were selected in such a way as to include at least one indicator per performance category (technical, com- mercial, and financial). To choose the number of indicators and the set of utilities to include, the data were separated into 12 vertically integrated utilities (VIUs), 23 genera- tion utilities (GUs), 29 DUs, and 3 transmission utilities (TUs). Because 3 is such a small number, we decided to exclude this last group from the average rank score exercise. For each of the other three groups, the utilities with data available on each indicator were identified, as well as those with data for Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 68 A Multi-Indicator Approach to Analyzing Utility Performance all the most populous indicators in a sequence. This provided a picture of the trade-off between the number of indicators and the number of utilities avail- able for the average rank score exercise. Distribution Utilities: Average Rank Score Table 4.2 shows the results of the multi-indicator approach, as applied to DUs. The OPEX per kilometer (km) indicator is the most widely available: 37 obser- vations over all types of utilities and 27 observations for DUs. The total energy volume sold per connection is available for 26 distribution utilities, but only 25 DUs have data on both this and OPEX per km. Add the ratio of energy sales to OPEX, and this limits the set of DUs to 22. As more indicators are added, the number of utilities with data on all drops steadily. We decided to include the five indicators shown in table 4.2.3 This group of indicators covers technical, commercial, and financial performance measures, which should help capture different aspects of performance.4 Table 4.3 lists the individual and average rank scores of each of the 17 DUs for which there were data on all five indicators. Table 4.2  Trade-Off between Number of Distribution Utilities and Number of Indicators Common to All MENA Utilities Distribution utilities Common to sequential set of Indicator All only distribution utilities OPEX/km 37 27 27 Total energy volume sold/connection 35 26 25 Energy sales/OPEX 32 23 22 Return on equity 46 24 18 Revenue/employee 34 26 17 Source: World Bank calculations. Note: km = kilometer; MENA = Middle East and North Africa; OPEX = operating expenses. Table 4.3  Ranks and Average Rank Score for Distribution Utilities, MENA Energy volume/ Energy sales/ Return on Revenue/ Average Utility OPEX/km connection OPEX equity employee rank Jordan - EDCO 5 16 14 15 12 12.4 Morocco - LYDEC 1 10 15 16 16 11.6 Jordan - JEPCO 3 17 11 13 14 11.6 Morocco - REDAL 2 9 16 12 17 11.2 West Bank - NEDCO 8 11 17 8 10 10.8 Morocco - RADEM 6 3 13 17 13 10.4 Egypt, Arab Rep. - SCEDC 10 14 7 11 9 10.2 Egypt, Arab Rep. - CEDC 12 15 5 10 7 9.8 Egypt, Arab Rep. - NDEDC 15 5 12 7 6 9.0 table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Multi-Indicator Approach to Analyzing Utility Performance 69 Table 4.3  Ranks and Average Rank Score for Distribution Utilities, MENA (continued) Energy volume/ Energy sales/ Return on Revenue/ Average Utility OPEX/km connection OPEX equity employee rank Egypt, Arab Rep. - NCEDC 11 12 8 6 8 9.0 Morocco - RADEES 4 2 9 14 15 8.8 Morocco - RADEEL 7 4 6 9 11 7.4 Egypt, Arab Rep. - UEEDC 16 6 4 3 5 6.8 Egypt, Arab Rep. - EEDC 14 13 3 1 3 6.8 Egypt, Arab Rep. - MEEDC 17 8 2 2 4 6.6 Egypt, Arab Rep. - SDEDC 13 1 10 5 2 6.2 Egypt, Arab Rep. - AEDC 9 7 1 4 1 4.4 Source: World Bank calculations. Note: km = kilometer; OPEX = operating expenses. Based on the average rank score, Jordan’s Electricity Distribution Company (EDCO) is the best-performing utility in the group, followed by Lyonnaise des Eaux de Casablanca (LYDEC) of Morocco and the Jordan Electric Power Company (JEPCO). EDCO and LYDEC perform very well on four indicators but poorly on OPEX/km, whereas the Alexandria Electricity Distribution Co. (AEDC) in the Arab Republic of Egypt performs very poorly on three indica- tors but is in the middle of the group for the other two. These examples illustrate the potential danger of relying on a single indicator to describe a utility’s performance. AEDC is the worst performer overall, followed by the South Delta Electricity Distribution Company (SDEDC) and the Middle Egypt Electricity Distribution Company (MEEDC), both also in Egypt. It is notable that the Egyptian utilities tend to perform poorly, with five out of nine utilities performing near the bottom of the set. This suggests there may be some common factors behind their performance, such as an idiosyncrasy in compiling the data or a common national policy that leads to poor perfor- mance. (The Egyptian case study in part II of this book provides further insights.) This finding might not have been identified through the use of a single indicator and points to the value of using several dimensions of perfor- mance at the same time. The spacing of the average rank values is also of interest. With 17 utilities, the maximum average rank score is 17 (one utility is best at everything) and the minimum is 1 (one utility is worst at everything). In practice, the values range from 12.4 to 4.4, and the gap in the average rank score between successive per- formers is about 0.5 points. EDCO is 0.8 points ahead of LYDEC, indicating clear superiority on the basis of the average range score criterion. At the other end of the distribution, AEDC is 1.8 points below the next-worst performer, indicating a markedly poor performance. Finally, across the set, we find no indication that the utilities are trying to improve efficiency simultaneously on a subgroup of indicators. The extent to which utilities rank similarly on a particular indicator is measured by the Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 70 A Multi-Indicator Approach to Analyzing Utility Performance coefficient of concordance. The average Spearman rank correlation between indicators is +0.008, so that the concordance is 0.20 and the probability of exceeding such a value (under the null hypothesis of no association between the different rank scores) is 0.42. This value far exceeds the conventional 5 percent used to reject the null hypothesis. Generation Utilities: Average Rank Score For GUs, the coverage of the indicators was much thinner. To include technical and financial indicators and a reasonable spread of utilities, we decided to retain only three indicators: the ratio of current assets to current liabilities, ROA, and the capacity factor. This set of indicators was available for 13 utilities. Table 4.4 presents the individual and average rank scores for the GUs that provided data on all three indicators. The best performing is the Qatrana Electric Power Company (QEPCO) in Jordan, followed by the Al-Kamil Power Plant (AKPP) and the ACWA Power Barka (APBS), both in Oman. The worst performers are in Egypt: the Cairo Electricity Production Company (CEPC) and the West Delta Electricity Production Company (WDEPC). The score gap between QEPCO (12.0) and the next-best performer (9.7) indicates a very large difference in performance between these utilities and suggests that QEPCO is well in advance of the other GUs in the set. Three Egyptian utilities (out of the four for which there are data) occupy the bottom three places in the average ranking, suggesting that policy has not focused on improving performance even toward levels seen elsewhere in the MENA region. Table 4.4  Ranks and Average Rank Score for Generation Utilities, MENA Current assets/ Utility current liabilities Return on assets Capacity factora Average rank Jordan - QEPCO 13 10 13 12.0 Oman - AKPP 7 12 10 9.7 Oman - APBS 11 11 7 9.7 Oman - SPP 10 7 11 9.3 Oman - GPDCO 12 6 5 7.7 Jordan - CEGCO 8 13 1 7.3 Egypt, Arab Rep. - UEEPC 4 5 12 7.0 Egypt, Arab Rep. - MDEPC 6 4 9 6.3 Jordan - SEPCO 9 8 2 6.3 Oman - UPC 3 9 3 5.0 Egypt, Arab Rep. - EDEPC 2 2 8 4.0 Egypt, Arab Rep. - WDEPC 5 1 4 3.3 Egypt, Arab Rep. - CEPC 1 3 6 3.3 Source: World Bank calculations. Note: MENA = Middle East and North Africa. a. For an interconnected system. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Multi-Indicator Approach to Analyzing Utility Performance 71 The average Spearman rank correlation between the series is +0.21, which implies a concordance value of 0.47. The probability of exceeding this value, with 3 indicators and 13 observations, is 0.15. Therefore, the hypothesis of zero concordance between the indicators is accepted: utilities as a group show no tendency to perform well or badly across all dimensions of performance. They appear to focus randomly on certain indicators of performance and to pay less attention to other indicators. Vertically Integrated Utilities: Average Rank Score Four indicators covering all dimensions of performance (technical, commercial, and financial) were chosen for the average rank score: OPEX per connection, current assets to current liabilities, total energy sold per connection, and distribution losses. With this set of indicators, 8 of the total 12 VIUs in the sample could be included. The results for the rankings and average rank score are shown in table 4.5. The best performance is that of the Saudi Electricity Company (SEC) in Saudi Arabia, followed by Algeria’s Socièté Nationale de l’Electricité et du Gaz (SONELGAZ). The worst performance is that of Electricité du Liban (EdL) in Lebanon, followed by the Public Electricity Corporation (PEC) in the Republic of Yemen. The gap in the average rank score between the two worst performers is notably large (2–3.5 points), indicating that EdL’s performance is particularly poor. The average Spearman rank correlation is −0.06, implying a W value of 0.20. The probability of observing this value is 0.57, supporting the hypothesis that VIUs as a group did not focus on the performance of particular indicators. Table 4.5  Ranks and Average Rank Score for Vertically Integrated Utilities, MENA Current assets/ Total energy volume Distribution Average rank Utility OPEX/connection current liabilities sold/connection losses score Saudi Arabia—SEC 4 5 8 8 6.3 Algeria—SONELGAZ 7 8 5 3 5.8 Oman—RAECO 1 7 6 7 5.3 Tunisia—STEG 5 6 2 6 4.8 Morocco—ONEE 6 3 4 5 4.5 Oman—DPC 3 2 7 4 4.0 Yemen, Rep.—PEC 8 4 1 1 3.5 Lebanon—EdL 2 1 3 2 2.0 Source: World Bank calculations. Note: MENA = Middle East and North Africa; OPEX = operating expenses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 72 A Multi-Indicator Approach to Analyzing Utility Performance Conclusion The average rank score provides a method of identifying the better-performing utilities among a group that share a common set of data, and for which reliance on a single indicator could be misleading. In the case of MENA, the data gaps are substantial, which substantially reduces the number of utilities that could be compared. This effect was particularly notable for GUs: only 13 of the original 27 could be compared on a common basis. For all three utility types analyzed, the coefficient of concordance between the series was low, and the null hypothesis of no agreement in rankings between series was accepted. This suggests that relying on the ranking of a single indicator would produce very different results than the same exercise using any other single indicator. Combining scores is more likely to provide a reliable picture. Furthermore, low concordance values suggest that, generally, utilities were focusing on different subsets of indicators to improve performance. However, the clustering of poor performance scores for Egyptian DUs and GUs suggests that common policies are leading to poor performance within that country. The average rank scores also identified utilities with extremely good or extremely poor performance by comparing them to the next-best (or worst) utility. QEPCO (Jordan) was well in front of the other GUs analyzed, EdL (Lebanon) was well behind other VIUs, and AEDC (Egypt) was far behind other DUs. The ability of the method to highlight such cases could inform subsequent analysis, by indicating which policies are factors of success and which of failure. Notes 1. W can be calculated in alternative ways as shown in “Real Statistics Using Excel,” http://www.real-statistics.com/reliability/kendalls-w/. Taking the average of the Spearman rank correlations (the usual correlation formula applied to the ranked val- ues) of all pairs of indicator variables, denoted by r, m as the number of indicators, and k as the number of observations, then W = ( m − 1) × r + 1 . It can be shown that when m there is complete agreement between indicators (the ranking is the same for every indicator), then W reaches its maximum value of unity. When there is no agreement between indicators—differences in rank scores between indicators are large—the minimum value of W is zero. 2. Under the conditions m > 15 or k ≥ 5: m × (k − 1) × W ~ c 2(k − 1) when the null hypothesis of no agreement is true. 3. Most indicators show better performance as they increase and are ranked in ascending order (1 = worst, 17 = best), but for indicators such as OPEX/km, where smaller values are better, the ranking is in descending order. 4. OPEX/km and revenue per employee are technical indicators, energy volume per connection is a commercial indicator, and energy sales per OPEX and return on equity are financial indicators. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 5 Drivers of Utility Performance: Institutional and Contextual Characteristics The tremendous global heterogeneity of electricity sector structures may be one of the most striking stylized facts characterizing this sector.1 The indicators col- lected for this study show that the Middle East and North Africa (MENA) region is no exception. Even if utilities continue to be central to each of the organiza- tional models adopted in the region, these models differ across a number of institutional and contextual characteristics. Some of these differences have been credited with, or blamed for, differences in utilities’ performance. Policy choices, such as the unbundling of the sector, the introduction of private ownership, or the introduction of a separate regulatory authority, have been suggested as key steps in improving the overall performance of the electricity sector (Bacon and Besant-Jones 2001). However, the lack of overwhelming evidence for the bene- fits of power sector reform as a panacea for poorly performing power utilities is leading to a reevaluation of policy responses to this underperformance.2 At the same time, further evidence on the impact of various sector reform strategies can help inform the debate. The data collected for this analysis of sector perfor- mance in the MENA region provide the opportunity to contribute to this discussion. The specific institutional dimensions related to the data on performance indica- tors collected are as follows: (a) the degree of vertical integration and the special- ization of the utility (that is, the type of utility), (b) the size of the utility, (c) the nature of its primary ownership, and (d) the presence (or not) of a separate regu- latory agency. One contextual dimension characterizing the environment in which the utility operates is added to this list, namely (e) the economy’s overall level of income. Given that income levels have a highly negative correlation to energy imports in the economies of our study,3 the correlation between performance of utilities and income level should be similar to the one between performance of utilities and an economy’s net energy imports. This chapter relates these five dimensions to each of the performance indicators included in this study, in a first Shedding Light on Electricity Utilities in the Middle East and North Africa   73   http://dx.doi.org/10.1596/978-1-4648-1182-1 74 Drivers of Utility Performance: Institutional and Contextual Characteristics attempt to test for any connections (for example, to see if public and private utilities performed differently on a particular indicator). Table 5.1 lists the set of utilities in the MENA Electricity Database, catego- rized by the institutional and contextual dimensions used for this analysis. The study divides utilities into four classes: vertically integrated utilities (VIUs), generation utilities (GUs), transmission utilities (TUs), and distribution utilities (DUs). The number of VIUs (12) is much smaller than the number of GUs (23) or DUs (29), and there are but a handful of TUs (3). The majority of DUs are found in the Arab Republic of Egypt (9) and Morocco (11), and the majority of GUs in Oman (12) and Egypt (5). Table 5.1  Breakdown of Sample Utilities by Size, Ownership, Presence of a Separate Regulator, and Income, MENA, 2013 (or most recent year with data, 2009–12) Vertically Distribution Generation Transmission Categories Measure integrated utility utility utility utility Size Big 5 8 8 3 Medium 4 10 6 0 Small 3 11 9 0 Ownership Public 11 21 10 3 Private 1 8 13 0 Presence of separate Present 6 18 23 3 regulatory agency Absent 6 11 0 0 Income High 5 3 12 1 Upper-middle 4 3 6 1 Lower-middle 3 23 5 1 Source: World Bank calculations. Note: MENA = Middle East and North Africa. Table 5.1 shows that the sample is relatively well distributed across sizes, because it includes 24 big, 20 medium-sized, and 23 small utilities. Some biases are more peculiar, such as the fact that there are no large private utilities in our sample: this points to a major difference between the MENA region and other regions of the world. The most obvious economy-related bias is that 14 of the 24 big utilities are Egyptian. The table also shows that with respect to ownership, the sample is not balanced. All TUs are public, so we cannot assess the impact of ownership. well ­ Similarly, all the big utilities are public because the big utilities include VIUs and TUs as well as other specialized utilities. However, the many DUs and GUs spread throughout the region have both public and private ownership. There are 50 utilities operating in economies with a sector regulator, leaving 17 utilities without a sector regulator. However, none of the GUs (or TUs) are subject to a regulator, limiting the evidence of any impact on these utility types. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 75 The high-income country (HIC) group includes 21 utilities, the upper-middle- income country (UMIC) group includes 14 utilities, and the lower-­ middle- income country (LMIC) group includes 32 utilities. Although the sample sizes of TUs and VIUs are small, they are spread evenly across income levels. By con- trast, GUs are heavily concentrated in HICs, and DUs utilities in LMICs. Potential Determinants of Utility Performance Type of utility, organizational structure, and performance. The literature has emphasized unbundling vertically integrated power utilities as one step toward improving performance, and the horizontal unbundling of generation and distri- bution as a further performance-enhancing step. Unbundling generation from transmission and distribution (T&D) allows for multiple GUs, financed by pri- vate capital, and the introduction of some form of competition. These steps are expected to improve performance by reducing costs and increasing efficiency. Similarly, unbundling distribution allows the introduction of multiple utilities and private ownership and the possibility of competition, which again are expected to improve performance.4 However, for small utilities, vertical and hori- zontal separation and the introduction of multiple entities reduces the average utility size and may result in the loss of economies of scale and scope. In the MENA region, although several economies have experienced unbundling, none has yet introduced competition between generators or between distributors. Hence any structural changes could not rely on gains from competition. Indeed, if scale economies are important, as is likely true for generation and transmission elements, then reducing the size of the average power company in an economy might be expected to worsen performance. However, there is a counterbalancing factor made possible by unbundling the functionally different components of a VIU. Managers with limited experience may find it simpler to concentrate on the key functions of generation or of distribution rather than to balance the conflict- ing interests of a VIU. In this case performance could be higher where unbun- dling has been introduced. Structure may then have a positive relation to performance even in the absence of competition or of private ownership. It is to be expected that this effect is weaker before the introduction of private capital prompts an intensified search for higher profits and lower costs and weaker still than it would be amid competition between utilities. Size and performance. In a sector in which the existence of economies of scale and scope has been the working assumption, any significant performance differ- ences according to size deserve a close look. Size is notably varied across subsectors and contexts. Among other things, dif- ferences reflect policy makers’ efforts to address climate change concerns and attract private investors to finance at least part of a utility’s investment require- ments, and eventually to introduce competition. These sources of pressure have had a significant impact on the way optimal structure is being discussed in the literature. Indeed, many observers argue for unbundling the sector to make the most of the latest renewable technologies, which would impact the optimal Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 76 Drivers of Utility Performance: Institutional and Contextual Characteristics size of T&D. In a nutshell, the case for fragmenting the sector into smaller units seems to be growing. Therefore, we need to understand the extent to which cur- rent preferences leave room for improvement—and on which dimensions. Significant performance differences, especially in terms of costs, may explain some of the reluctance to restructure, as long as the new technologies or market structures cannot guarantee improvement. This chapter contributes to the dis- cussion by clarifying differences in performance according to size. It provides a baseline on which to anchor the growing case for a redesign of the sector so as to make the most of renewable resources. In summary, horizontal unbundling implies a reduction in the size of GUs and DUs. Until competition between utilities of the same type is introduced, the loss of scale due to unbundling may actually lower performance.5 One challenge in trying to assess the relevance of size is its subjective nature. In the context of this study, the following definitions have been adopted. The sizes of VIUs and DUs are defined by number of connections. A utility with fewer than 250,000 connections is small; utilities between 250,001 and 2 million are medium; utilities above 2 million are big. Because GUs do not have direct customers, the total installed capacity of power plants is the key determinant of utility size. For this purpose, GUs with installed capacity below 500 megawatts (MW) are considered small, those with installed capacity greater than 1 gigawatt (GW) are big, and anything between is medium. Finally, for TUs, the amount of energy transmitted determines size. Those transmitting less than 5 terawatt-hours (TWh) are small; those transmitting between 5 TWh and 10 TWh are medium; and those transmitting more than 10 TWh are big. Ownership and performance. For almost 30 years, the debate on the relative effectiveness of the public and private operation of electricity utilities has been raging. It has yet to be settled. The experiences have been so diverse that there is no possibility of a definitive answer. Arguments for the benefits of private partici- pation and ownership stress the pressure from new owners to maximize profits through efficiency and pricing strategies. Where prices are controlled, as in the MENA region, one expected effect is a reduction of costs. A further benefit of allowing private ownership into the sector is that it pro- vides a source of finance and thus lightens the government’s financial burden, which is sure to grow heavier as demand for power increases. Also, the discipline of market financing is more likely to avoid the adoption of suboptimal projects. Governments may support projects for political rather than economic reasons, without paying attention to the costs of doing so. A further argument for encour- aging the entry of private sector investment is that it sets an example of good management that publicly owned utilities may be encouraged to emulate. Other aspects of performance may become secondary, provided that they are not seen as interfering with the return on investment. This chapter summarizes some basic, stylized results of an exercise in comparing the performance of elec- tricity utilities in MENA depending on their ownership (that is, public or private). The discussion focuses on correlations rather than causality. In most cases Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 77 represented in the MENA Electricity Database—if not all—a self-selection bias prevails and explains why some economies have gone one way and some another in terms of ownership. Presence of a separate regulatory agency and performance. Creating inde- pendent regulatory institutions has been a standard component of electricity sector reforms for almost 20 years (see, for example, Jasmab and others 2015). Some MENA economies have jumped on the bandwagon, others not.6 According to Cambini and Franzi (2013), in a review of the regulatory gover- nance of Mediterranean economies—including many in the MENA region—this decision has mattered to the implementation of key policy decisions. Their research emphasizes the impact of separate regulatory agencies on the ability of MENA economies to attract investment to diversify energy sources. Cambini and Franzi do not, however, examine the impact on other more tech- nical and specific performance indicators at the utility level and focus instead on the impact at the economy level. This chapter provides additional insights on the impact of the decision to restructure regulatory governance in the MENA region by comparing (a) the performance of MENA electricity utilities supervised by separate regulatory agencies with (b) the performance of utilities operating in economies where regulation is still under the control of the sector ministry. To establish a possible link, we assess the correlation between any difference in performance across comparable utilities in the MENA region and the choice of regulatory gover- nance at the very broad level, as allowed by the limited data on the detailed nature of this governance in the region. It is a weak test that does not establish causality between institutions and performance, but it manages to produce MENA-specific information in a region in which little related data have been collected. The focus is on the broad signal offered by the institutional unbundling of the regulatory responsibility at the utility level. It does not get into the quality of the signal. As highlighted by Cambini and Franzi (2013) for their sample, the specific design of an institution may have a significant effect on the strength of the signal sent by its creation. The data collected here do not allow the inter- nalization of these important dimensions. For instance, we do not consider the extent to which the separate regulatory agencies are financially or politically autonomous or the relevance of staff skills or the menu of mandates assigned to regulators and matching regulatory instruments. Despite this limitation, the research proves useful in assessing the extent to which a simple increase in the transparency of the regulatory function, allowed by the creation of a sepa- rate institution, made some difference—no matter what the quality of this institution was and how much of a difference it made. Indeed, the chapter shows that the impact of differences in regulatory governance is not binary, let alone simple (but this was to be expected, based on earlier assessments of inter- national experience). This work suggests that the impact of the introduction of a separate regulatory agency is difficult to anticipate. It varies significantly across economies and Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 78 Drivers of Utility Performance: Institutional and Contextual Characteristics regions. In other words, context matters. In a survey of institutional reforms in the energy sector that includes a detailed assessment of the importance of regula- tory institutions, complemented by additional analytical evidence, Vagliasindi and Besant-Jones (2013) show that the impact of introducing an independent regulator depends on a wide range of factors, including system size, development level, and demand composition. It also depends on the performance indicators being analyzed. For instance, an independent regulator may send a strong signal to investors without doing much to affect actual investment levels, and it may either increase or decrease prices, already refining some of the insights of Cambini and Franzi (2013). Economy income level and performance. The performance of utilities may relate to the income level of the economy. For instance, demand for energy increases with income per capita, and this may change the composition of the demand base of the utilities. Growth usually comes with a stronger industrial sector, which tends to be more energy intensive. We also know that, in general, higher income levels are correlated with stronger institutions. This, in turn, may have an impact on the incentives utilities have to make stronger efforts to per- form (that is, by reducing the risk of moral hazard in the management of the sector and among its various actors). It may also lead to access to more-experienced and better-equipped utilities (that is, reducing the risk of adverse selection by increasing the scope for competition in the sector, which is often associated with more cost-effective technical solutions). To inform the discussion of this possible evolution, the sample has been divided into three groups: HICs, UMICs, and LMICs.7 Evidence of significant differences in performance according to income level indicate that simple comparisons of performance across economies, without tak- ing into account differences in their income level, may be misleading. There may be other contextual factors that correlate with differences in utility performance and have not been explored in the context of this study on MENA utilities. Summary of Results Studies of the impacts of power sector reform have concentrated on a time-series approach—that is, for a particular economy or utility, the performance accord- ing to a number of indicators is compared prior to the introduction of the reform and for a number of years post reform. Jones, Tandon, and Vogelsang (1990) developed a method of comparing the historical and predicted future course of an industry with a counterfactual in which the industry remained unprivatized. Galal and others (1994) applied this method to two DUs in Chile, and Newbery and Pollitt (1997) described how to evaluate the restruc- turing and privatization of the U.K. electricity supply industry using this approach. In the latter case, great attention was paid to identifying those changes taking place that were due to external forces (for example, changes in European regulations) and those changes brought about by the act of privatization. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 79 This approach focuses on one utility or one economy at a time and requires detailed knowledge of and data on the sector for a number of years before and after the policy change under evaluation. A crucial step in this type of analysis is the determination of how much performance would have changed in the absence of reform. Bacon and Besant-Jones (2001) quoted values for changes in performance on several indicators (energy sales, energy losses, employment, customers/employee, and net receivables) since privatization for four South American DUs. Treating such changes as entirely due to the effects of privatization is equivalent to assuming that without privatization there would have been no change in any of these indicators. For energy sales, certainly, this was an unrealis- tic assumption. In the present study, the availability of data drawn from a large number of utilities exhibiting different characteristics provides the opportunity to test for the effects of various reform strategies in a different way. If the average perfor- mance of all public utilities on various indicators is poorer than that of the aver- age for private utilities on the same indicators, then this supports the argument that privatization can help improve performance. In making such comparisons it is recognized that there are many individual factors that contribute to perfor- mance on a particular indicator, so that differences between public and private would not be due solely to their ownership status. A significant difference between performance levels across the two ownership types supports the argu- ment that ownership matters. If the difference is not significant, this indicates that ownership does not in itself outweigh all the other factors determining performance on this particular indicator. But this does not prove that ownership has no impact on performance. This chapter presents the results of an attempt to identify correlations, if any, between utility performance and five factors (type of utility, size, ownership, existence of a separate regulator, and income level of economy). A limitation of this exercise is that we only have cross-sectional and not time-series data, so no causality can be inferred. For each of the 36 performance indicators, the average for all relevant utilities over available observations is constructed. Next, for each of the five institutional and contextual factors, the averages for the same indica- tor are calculated and statistical tests of equality are carried out. For example, as a test of the importance of sector structure, data on the load factor from VIUs and DUs are tested to see whether the averages are the same for both utility types. Next, the mean load factor for utilities under a regulator is compared to the mean where there is no regulator, and so on. Results for the 36 performance indicators are grouped into the categories indicated in appendix A. Appendix D provides a brief account of the methodology used here. For each indicator, table 5.2 specifies the classes of utilities to be included in the analysis, the total number of observations included in the test, and the overall mean for this indicator. It presents the probabilities of the tests for equality of means across five institutional and contextual factors (utility type, size, national income, ownership, and the presence of a seperate regulator). These are the probabilities of obtaining a difference at least as large as that observed if the null Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 80 Table 5.2  Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities Separate Utility regulatory Classes of utilities included Indicator Category Number Mean type Size Income Ownership agency present VIU vs. DU Load factor System and 23 0.56 0.80 0.25 0.96 0.07* 0.63 VIU vs. GU Capacity factor operational 20 0.54 0.07* 0.61 0.12 0.43 S efficiency VIU vs. GU Availability factor 11 0.93 0.50 0.71 0.04** 0.50 0.50 VIU vs. TU vs. DU Network maintenance 10 0.02 0.79 0.41 0.85 0.52 0.20 VIU vs. DU Share of meters replaced (%) 9 0.02 0.41 0.40 0.70 0.29 0.82 VIU vs. TU Transmission losses Losses efficiency 3 0.03 0.86 S S S S VIU vs. DU Distribution losses 37 0.13 0.001** 0.76 0.52 0.63 0.69 VIU vs. DU Technical losses 18 0.075 0.0003** 0.37 0.32 0.22 0.14 VIU vs. DU Nontechnical losses 18 0.049 0.0003** 0.88 0.32 0.13 0.48 VIU vs. GU vs. TU vs. DU OPEX/employee Cost efficiency 48 274,000 n.a. 0.0001** 0.003** 0.006** 0.39 VIU vs. DU OPEX/connection 36 723 n.a. 0.16 0.0001** 0.99 0.80 VIU vs. DU OPEX/kWh sold 36 0.11 n.a. 0.002** 0.51 0.41 0.0001** VIU vs. TU vs. DU OPEX/km 37 24,381.0 n.a. 0.006** 0.95 0.02** 0.001** VIU vs. DU Residential connections/employee Labor efficiency 24 238 n.a. 0.09* 0.54 0.15 0.02** VIU vs. DU Energy sales/employee 31 170,000 n.a. 0.03** 0.48 0.0007** 0.005** VIU vs. DU Total revenues/employee 34 212,000 n.a. 0.1* 0.70 0.004** 0.001** VIU vs. GU Cost fuels/OPEX Cost structure 22 0.65 0.12 0.16 0.47 0.97 0.02** VIU vs GU Energy purchases + fuels/OPEX 8 0.77 S 0.05** 0.23 S 0.70 VIU vs. GU vs. DU Labor cost/OPEX 35 0.13 0.22 0.03** 0.02** 0.13 0.29 VIU vs. DU Energy sales/OPEX Cost recovery 32 0.95 0.42 0.49 0.07* 0.83 0.15 VIU vs. DU Energy sales/costs 19 0.82 0.11 0.13 0.03** 0.54 0.48 table continues next page Table 5.2  Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities (continued) Separate Utility regulatory Classes of utilities included Indicator Category Number Mean type Size Income Ownership agency present VIU vs. DU Accounts receivable Balance sheet 26 161 0.11 0.22 0.06* 0.84 0.63 VIU vs. GU vs. TU vs. DU Debt/equity 47 7.08 0.24 0.05** 0.04** 0.62 0.67 VIU vs. GU vs. TU vs. DU Assets/liabilities 53 1.17 0.32 0.0005** 0.31 0.56 0.84 VIU vs. GU vs. TU vs. DU Return on assets Profitability 49 0.3% 0.39 0.07* 0.22 0.05* 0.40 VIU vs. GU vs. TU vs. DU Return on equity 46 4.6% 0.009** 0.10 0.15 0.03** 0.12 VIU vs. DU Total energy volume/connection Consumption 35 6.4 0.002** 0.36 0.001** 98.0 0.21 VIU vs. DU Residential energy volume/ and billing connection 23 4.0 0.01** 0.72 0.0001** 0.62 0.51 VIU vs. DU Total billing/connection 27 297 0.17 0.005** 0.0001** 0.037** 0.09* VIU vs. DU Residential billing/connection 22 258 0.59 0.0001** 0.007** 0.37 0.34 VIU vs. DU Collection rate 15 88% 0.03** 0.003** 0.86 0.51 0.08* VIU vs. DU Share of installed meters (%) Metering 15 96% 0.32 0.33 0.02** 0.72 0.75 VIU vs. TU vs. DU SAIFI Customer 15 1.6 0.02** 0.70 0.06* 0.37 0.69 management VIU vs. TU vs. DU SAIDI 12 28.6 0.46 0.35 0.72 0.49 0.57 and service VIU vs. TU vs. DU CAIDI quality 9 52 0.21 0.46 S S 0.20 VIU vs. TU vs. DU Duration of interruptions 5 2.0 S 0.99 0.03** 0.32 0.03** Source: World Bank calculations. Note: Significant results are shaded in light red; performance indicators for which more than one factor gave significant results in one-at-a time testing are shaded in green; tests that are inappropriate are shaded in blue. CAIDI = Customer Average Interruption Duration Index; DU = distribution utility; GU = generation utility; km = kilometer; kWh = kilowatt-hour; MENA = Middle East and North Africa; n.a. = not applicable (tests are inappropriate); OPEX = operating expenses; S = singular dataset so estimation is not possible; SAIDI = System Average Interruption Duration Index; SAIFI = System Average Interruption Frequency Index; TU = transmission utility; VIU = vertically integrated utility. Significance level: * = 10 percent, ** = 5 percent. 81 82 Drivers of Utility Performance: Institutional and Contextual Characteristics hypothesis of equality of the two means were correct. If the probability is less than 5 percent we concluded that there is a significant difference between the means.8 Significant results are shaded in brown, and those tests that are inap- propriate (for example, testing the effect of structure on operating expenses [OPEX]/employee) are shaded in blue. Those performance indicators for which more than one factor gave significant results in one-at-a-time testing are shaded in green. Table 5.2 reveals that of the 36 performance indicators analyzed, as many as 25 have at least one factor that shows significant differences, and that there are 14 cases in which more than one factor is found to be significant in one-at- a-time testing. The substantial number of indicators for which there are signifi- cant results, even in the absence of detailed modeling of the situation, provides support for arguments that sector reform may be able to improve sector performance. To focus on the relevance of these factors to performance, table 5.3 indicates which factors were most commonly related to performance, both by absolute number and as a percentage of indicators that could be tested for this effect. Three indicators (type of utility, size of utility, and the income level of the economy) were significant in around 30 percent of the cases, whereas ownership and the presence of a regulatory agency were significant in about 20 percent of the cases. Income was the factor found to be most often significantly related to performance indicators. In a region with wide variation in incomes from the LMIC to HIC levels, this serves as an important reminder that comparing perfor- mance across economies, without allowing for the effects of income level, could lead to a misjudgment as to the policy intervention required. Income is an impor- tant contextual factor: it has to be taken into account when designing sector reform policies, but does not point to any particular policy choice. The four policy variables (utility type, size, ownership, and regulation) are significant often enough to suggest that based on the MENA utilities, with their wide range of individual circumstances, there is evidence to support the use of reform strategies that use vertical and horizontal unbundling, introduce private ownership, and create a regulatory body. The tests used to arrive at these conclu- sions support only broad approaches to policy—they do not distinguish, for example, between different types of regulatory bodies with different degrees of independence from the government. The contextual variable, income, is also Table 5.3  Number of Indicators with a Significant Relation to Each Factor, MENA Utilities Separate regulatory Type of utility Size Income Ownership agency present Number of significant results 8 11 12 6 7 % of significant resultsa 30 29 35 18 21 Source: World Bank calculations. a. Percentage of significant results equals number of significant results relative to number of applicable indicators (that is total number minus not applicable and singular cases). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 83 significant in several cases, making the point that the effects of policies may be dependent on the level of income in the economy concerned. The results summarized in table 5.2 are from introducing one factor at a time into the tests for differences. A total of 14 indicators showed significant results for more than one driver. These cases were analyzed to see the effects of intro- ducing more than one factor at the same time into the tests for differences. In 10 cases it was found that more than one factor is significant in testing for the simul- taneous effects of several factors, and in 4 cases there was no support for the significance of more than one factor. These results suggest that a more detailed examination of performance—that is, introducing more contextual factors and refining the specification of the institutional factors—could provide further insights into the determinants of performance. We then focus on indicator type. Grouping indicators into 11 categories and calculating the percentage of significant results by category yields the results in table 5.4. Certain categories of indicators show few significant links to the five factors, whereas others show a large number of significant links. For example, only 4 percent of the tests of system and operational efficiency are significant, compared with 56 percent for cost efficiency. These results suggest that the reform factors may be most often correlated with certain types of indicators. Further insights are obtained by noting, for each driver, where there were significant results for a substantial proportion of the indicators within a given category. Table 5.5 shows that the significant results for each driver are concen- trated within two or three indicator categories. For example, utility type has a substantial proportion of significant links to the losses efficiency, profitability, and consumption and billing categories, and no links at all to the systems and opera- tional efficiency, cost structure, cost recovery, balance sheet, and metering categories. These results suggest that the effects of reform would not be felt across all indica- tors but are likely to be concentrated in certain aspects of performance. Table 5.4  Number and Percentage of Significant Results, by Indicator Category Number of Absolute number of Percentage of Indicator category indicators significant results significant results System and operational efficiency 5 1 4 Losses efficiency 4 3 14 Cost efficiency 4 9 56 Labor efficiency 3 6 50 Cost structure 3 3 23 Cost recovery 2 1 10 Balance sheet 3 3 20 Profitability 2 2 20 Consumption and billing 5 11 44 Metering 1 1 20 Customer management and service quality 4 3 18 Source: World Bank calculations. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 84 Drivers of Utility Performance: Institutional and Contextual Characteristics Table 5.5  Categories of Indicators Whose Drivers of Performance Show Significant Results for a Substantial Proportion of the Indicators in that Category Driver of performance Categories with significant results Type of utility Losses efficiency, profitability, consumption and billing Size Cost efficiency, balance sheet, consumption and billing Ownership Cost efficiency, labor efficiency Regulation Cost efficiency, labor efficiency Income Cost efficiency, consumption and billing, metering Source: World Bank calculations. Statistically Significant Differences between Subgroups of Characteristics Quasi-Fiscal Deficits and Drivers of Performance The quasi-fiscal deficits (QFDs) described in chapter 1 measure performance through a combination of factors. This suggests that a general test of the relation between performance and the five drivers of performance can be made by relat- ing total QFD, and each of its components, to the drivers. For this exercise, it is sensible to focus on utilities, and so the QFD (and each of its components) as a share of utility revenue is used. We have data on 9 VIUs9 and 17 DUs. These groups are tested separately. (Further, the VIUs selected are all publicly owned so no test of ownership can be carried out for them.) On the one hand, for VIUs, no performance driver is significant for the total QFD or for its components (underpricing, T&D losses, collection losses, and overstaffing). For the DUs, no test is significant for the total QFD, for T&D losses, and for collection.10 On the other hand, there are also some significant results for DUs. Regarding overstaffing, the big utilities had a significantly (1 percent proba- bility) higher ratio of employees to total revenue (25 percent) than did medium (7 percent) or small utilities (8 percent). Utilities in LMICs had a significantly (probability 2 percent) higher share (21 percent) than in UMICs (2 percent) and in HICs (1 percent). Private utilities had a significantly (probability 3 percent) lower share (4 percent) than public utilities (19 percent), although regulation was not significant for overstaffing. Regarding underpric- ing, there is weak evidence (probability 10 percent) of a difference in the average share of revenue between private utilities (33 percent) and public utilities (64 percent). The correlations between the QFD components and the drivers of perfor- mance provided clear evidence of links to overstaffing in DUs, suggesting that private ownership is associated with a smaller degree of overstaffing. The other components of the QFD were not correlated with the drivers. This cannot be taken as a rejection of the relevance of reform strategies for the power sector, but rather indicates that a more fully specified analysis of the links to performance would be needed before such a determination could be made. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 85 Utility Type and Drivers of Performance For some indicators, differences across utility types are expected. OPEX covers a wider range of functions for a VIU than for a DU serving the same number of customers, because it has to incur costs for generation and transmission activities. Hence it would be meaningless to test for equality of utility type for any indica- tor incorporating OPEX. The broad picture of the impacts of reform indicate that utility type is signifi- cantly linked to losses efficiency, profitability, and consumption and billing indi- cators, and not to system and technical efficiency, cost structure, cost recovery, and balance sheet indicators. Distribution losses averaged over all utilities at 13 percent, which is similar to the range of values found in the non-MENA group. However, the data show that DUs have a significantly better performance level (10 percent) than VIUs (20 percent). This suggests that DUs are better able to focus on their primary busi- ness than VIUs, whose problems are more widespread. This provides support for those arguing that unbundling can stimulate cost reductions. Technical losses are made up of nonvariable technical losses and variable technical losses. These were significantly lower for DUs (7 percent) than for VIUs (10 percent). Technical losses may reflect a relatively low load factor, because consumption (and therefore load) is less even throughout the day. Nontechnical losses were on average 4.9 percent, and VIUs (10 percent) showed much larger losses than DUs (3.6 percent), indicating a significant dif- ference between the performance of these two groups. Return on equity (ROE) stands at 4.6 percent. The highest ROE is observed for GUs (11 percent), followed by DUs (7.0 percent), and VIUs (−23.0 percent). The Omani TU’s ROE stands at 20 percent, but this is not representative of all TUs in the region. Significance tests indicate that ROE for GUs and DUs is significantly higher than for VIUs and lower than for TUs. VIUs in the MENA region perform poorly in terms of ROE compared with those outside MENA. Risk perception would be high in any environment in which domestic tariffs need to be subsidized or depend on politically sensitive cross-subsidies, as is the case for most VIUs in MENA. Return on assets (ROA) and ROE, on average, would not pass the common hurdle rates considered by investors and lenders. This is even true for the hurdle rates adopted by most international organizations, whether they want to support public or private projects. This exposes economies to increased risk of projects ending up being packaged to meet these hurdle rates rather than to address the broader investment challenges of the sector. The average total energy volume sold per connection (and year) is generally quite reasonable by international standards, accounting for the income level and the industrial and service structure of the region. The VIUs (12.6 mega- watt-hours [MWh]) sell a significantly greater amount per connection than the DUs (4.2 MWh). This probably reflects the nature of the customers served by these different types of utilities, rather than reflecting an inherent superiority of VIUs. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 86 Drivers of Utility Performance: Institutional and Contextual Characteristics The residential energy volume sold per connection is also consistent with the international best practice in economies with similar income characteristics. The differences between VIUs (8.0 MWh) and DUs (2.6 MWh) are again significant. The collection rate is the ratio of the revenue collected to the total electricity billed. The higher the ratio, the higher the effectiveness of the utility in bill col- lection. DUs have a significantly higher collection rate (89 percent) than VIUs (69 percent), probably because their narrower focus frees them to focus more on collection. The average System Average Interruption Frequency Index (SAIFI) illus- trates, at best, a reasonable performance. The SAIFI for DUs (1.23) is signifi- cantly lower than that for VIUs (3.18). This group of indicators, for which there are significant differences in perfor- mance, offers a coherent picture: utilities that have only a distribution function are able to concentrate on reducing losses and improving collection and perform better than VIUs with respect to these indicators. Size and Drivers of Performance Size was found to be significantly linked to cost efficiency, balance sheet, and consumption and billing indicators. It was not linked to system and operational efficiency, losses efficiency, cost recovery, profitability, and customer manage- ment and service quality indicators—these are areas where the degree of govern- ment support is unlikely to impact performance but where management quality can have a significant effect. OPEX per employee differs quite significantly by type of utility. Values vary, from $376,000 for GUs, $190,000 for DUs, $216,000 for VIUs, to $58,000 for TUs. This result is in line with what practitioners expect. The hypothesis that OPEX per employee is constant across sizes is rejected. For the MENA region, the value for this indicator is significantly smaller for big utilities ($99,000) than for medium ($410,000) or small ($236,000) utili- ties, although the latter two were not significantly different. This result supports the notion of economies of scale being important at higher levels of ­ operation. This needs to be put in context. The average number of employees is almost seven times higher among big utilities. Some big utilities, such as Socièté Nationale de l’Electricité et du Gaz in Algeria, employ almost 20,000 people, whereas small utilities in Djibouti, for example, have about 1,000 employees. When comparing utilities in the MENA region by size, large utilities have a significantly lower OPEX per kilowatt-hour (kWh) ($0.04) than medium ($0.13) or small ($0.15) utilities. This is consistent with the suggestion that economies of scale are still strong in some economies of the region. In comparisons of OPEX per kilometer (km) by size of utility, the value for big utilities ($12,753/km) was significantly lower than for medium ($35,179/ km) and small ($27,896/km) utilities. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 87 Size also shows a significant difference for the share of labor cost in total OPEX, between the big (17 percent), medium (10 percent), and small (12 percent) utilities. A test of the relation between the debt-to-equity ratio and the size of the utility indicated that the ratio for big utilities (1,143 percent) was signifi- cantly higher than for medium (499 percent) and small (330 percent) utilities. Even though they are the least leveraged in the region, small ­ utilities are still very highly leveraged by international Organization of ­ Economic Co-operation and Development (OECD) standards in which leverage is usually less than 100 percent. Even if the current low levels of interest should provide a good margin to rely on debt finance, given the risk premia and the long-term nature of the financial commitment often indexed to price changes, the current MENA approach appears to be risky. The prospects for high leverage are, however, probably better for smaller utilities if their cost-recovery performance continues to be solid. For the larger utilities, costly government financing or guarantees continue to be the main option to stay highly leveraged. The ratio of current assets to current liabilities also differs according to size. Big utilities have a ratio of 79 percent, whereas medium are at 84 percent and small at 200 percent; and the difference between the large and medium and the small subgroups is statistically significant. In the MENA region, the larger the utility the less likely it is to be able to pay off its short-term liabilities. This reinforces the conclusion that the MENA region’s smaller utilities are better managed financially than the larger ones. The average total billing per connection is lower for DUs ($268) than for VIUs ($392), but this difference is not statistically significant. Total billing per connec- tion is significantly related to the size of the utilities. Big utilities ($155) have lower billing per connection than do medium ($404) or small ($381) utilities. Similar results are found for residential billing per connection. The collection rate is the ratio of the revenue collected to the total electricity billed. The higher the ratio, the higher the effectiveness of the utility in bill col- lection. The collection rate is significantly related to size, with big utilities at 91 percent, medium at 96 percent, and small at 65 percent. This result is some- what unexpected and may be due to the small sample size and problems in measuring this variable. Size is significant for a group of indicators that include OPEX as a compo- nent. For OPEX/employee, OPEX/kWh, and OPEX/km, large utilities have the lowest value and are the most efficient. Where OPEX is in the denominator, as for labor costs/OPEX the large utilities have the highest value, indicating that OPEX rises slower than labor costs as utility size increases. However, care has to be taken in interpreting these results. One of the major components of OPEX is fuel costs and the pricing of fuel across the region is by no means uniform. Large utilities may be concentrated in countries where the largest energy subsidies are available. The test for a size effect on the ratio of fuel costs did not Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 88 Drivers of Utility Performance: Institutional and Contextual Characteristics reveal significant differences, and further detailed analysis to understand these results could yield useful insights. This interlinking with government sup- port is probably reflected in the relationship between size and the debt/equity and current assets/current liabilities ratios. Big utilities have the highest debt/ equity ratios and the lowest current assets/current liabilities, both of which indi- cate a weak financial position, probably made possible by government support. These significant results for the relation of size to performance do not add support to the existence of technical economies of scale but are very important in indicating how in MENA government policy may have supported the larger utilities allowing them to support more adverse financial performance. Ownership and Drivers of Performance The state of ownership of a utility—private or public—has been one of the areas of discussion with respect to improving utility performance. Private ownership introduces the profit motive and incentives for improving performance. Privatization alone is recognized to run the danger of creating ­ private sector monopolies where profits are increased but at the expense of the consumer by allowing prices to rise while decreasing costs. Where com- petition can also be created then the dangers are less, but the complex market structures and institutions required to permit full competition do not ­ exist within MENA, or in many other countries. Accordingly increasing reli- ance is placed on government control possibly through the creation of a ­regulatory body. Ownership was found to be significantly correlated with cost efficiency, labor efficiency, and the ROE indicators. It was not correlated with system and opera- tional efficiency, losses efficiency, cost structure, cost recovery, and balance sheet indicators. Public utilities have significantly lower OPEX per employee ($180,000) than private utilities ($417,000). This is in a region in which utilities’ OPEX are largely dominated by the costs of fuel and labor. Although OPEX of public utilities is on average three times larger than that of private utilities in the MENA region, private utilities in the sample have almost six times fewer staff than public utilities. This would more than offset the differential in OPEX between the two categories. This result strongly supports the view that private ownership can result in the reduction of costly overstaffing. OPEX/km, too, is significantly lower for public utilities ($20,166/km) than for private ones ($37,290/km). There is a significant difference between sales per employee across private ($341,000) and public ($126,000) utilities. The public utilities are hampered by their much higher levels of employment. Private utilities had significantly higher revenues per employee ($360,000) than public utilities ($167,000). As is the case with sales per employee, this result supports the view that privatization does improve efficiency by reducing employment. The ROE among publicly owned utilities (0 percent) is well below that of privately owned (15.5 percent), indicating that public utilities are able to Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 89 withstand a low ROE because of government support. The magnitude of this difference suggests that the cost of this support must be substantial. For total billing per connection there is a significant difference between public ($259) and private ($464) utilities, easily explained by differences in the absolute number of connections. Public utilities in MENA have one-and-a-half times the value of total sales than private utilities, but almost four times the number of connections. It is notable that ownership type was not significant for any indicators of system and operational efficiency, losses efficiency, or customer management and service quality. These are activities where private management might have been expected to improve performance by introducing better proce- dures and more modern technology. The substantial improvements in private sector performance appear to come from small staff size. This probably reflects the use of state-owned enterprises (SOEs) as a means of absorbing some of the otherwise unemployed labor force. Regulation and Performance One of the main arguments for introducing a regulatory authority is to exercise some control over how VIUs and DUs set prices. This control, in turn, is expected to reduce costs as utilities look to maintain or increase profit margins. Improvements in system and operation efficiency, losses efficiency, and consumption and billing are expected under regulatory control. Table 5.5 indicates that regulation is significantly correlated with cost effi- ciency and labor efficiency indicators, but not with system and technical effi- ciency, losses efficiency, cost recovery, balance sheet, profitability, and consumption and billing indicators. Utilities operating where there is a separate regulatory agency have a signifi- cantly lower OPEX/kWh ($0.07) than utilities operating without one ($0.15). Similarly, for OPEX/km, utilities with a separate regulator have significantly lower values ($16,944/km) than those without ($36,469/km). Utilities with a separate regulator have far fewer residential connections per employee (205) than those without such a regulator (472). It is implausible that regulation would lead to a reduction in connections per employee (and an increase in employees per connection), so this result is likely circumstantial. The same goes for energy sales per employee: utilities with a separate regulator present had sales of $117,000 and with no regulator had sales of $279,000 per employee. Also, for total revenues per employee, utilities operating with a sepa- rate regulator had lower revenue per employee ($132,000) than those with none ($327,000). The average share of cost of fuel, lubricant, gas, and coal in total OPEX is surprisingly high in an oil- and gas-producing region, at 66 percent, and utilities operating with a separate regulatory agency have a significantly higher average share (70 percent) than utilities operating with none (45 percent). The presence or absence of a regulatory agency is not correlated with any indicator of system and operational efficiency, losses efficiency, balance sheet, Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 90 Drivers of Utility Performance: Institutional and Contextual Characteristics profitability, or consumption and billing. The correlations that do exist are prob- ably circumstantial. Income and Drivers of Performance Income is not a direct policy instrument, but may help explain government attitudes toward the power sector. Richer countries are better able to provide financial support (directly or indirectly) and considerations of political econ- omy within such countries may lead governments to provide such support. Lower tariffs meant to benefit consumers, or fuel input prices below interna- tional equivalents can play an important role in the performance of power utilities. Table 5.5 indicated that income was significantly correlated with cost effi- ciency, consumption, and billing and metering indicators. It was not correlated with losses efficiency, labor efficiency, and profitability. The availability factor is high, which suggests that the plants of the region can provide energy to the grid most of the time. However, although the availability factor is similar in HICs (92 percent) and in UMICs (98 percent), this difference is statistically significant (there are no observations on this indicator for LMICs). This suggests that in the HICs maintenance is less effective and the plants may be older (this conclusion may be influenced by the fact that most plants with low availability levels are in Oman, where weather conditions are extreme, with implications for peak load factors). Regarding OPEX/employee, the LMICs spend the least ($159,000/employee), UMICs spend in the middle ($293,000), and HICs spend the most ($400,000), and the differences between these subgroups are significant. These differences could reflect employee performance levels or could be explained by other factors. For example, per capita labor and some other input costs increase with income level. OPEX/connection is also significantly related to income. The mean value for HICs was $1,993, whereas for UMICs it was $839, and for LMICs, $394. Differences across income groups are also significant when considering the share of labor costs in total OPEX. LMICs have a share of 16 percent, UMICs have a share of 6 percent, whereas HICs have a share of 11 percent. The lower share seen in HICs might have structural reasons: four out of five HICs in the sample have a vertically integrated market. There is significant difference across income levels in the energy to sales ratio, with the highest ratio in the LMICs (0.91) and the lowest ratio (0.56) in the HICs. Again, the government support offered to utilities in HICs is likely to explain this. There was also a significant difference in the debt equity ratio between HICs (376 percent) and LMICs (1,065 percent). It appears that higher levels of eco- nomic development can lead to more acceptable levels of risk. The income level does appear to have an important effect on energy sales per connection (as would be expected). The average for HICs Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 91 (28.8 MWh) is significantly greater than for UMICs (5.7 MWh) and LMICs (3.9 MWh). For instance, HICs have high billing per connection, generally related to high energy per capita. Air conditioning is significant, particularly in economies in the Gulf Cooperation Council such as Bahrain and Saudi Arabia. The energy billed by LMICs is twice that of HICs, and LMICs also have five times the number of connections. Many utilities operate between $100 and $300 per connection. Because at least 50 per- cent of utilities in LMICs have an average billing of $300 per connection, these sales data provide further evidence of the impact of price controls and subsidies in LMICs. LMICs also have the lowest average residential billing per connection ($168), yet residential users represent a very large market segment, particularly in economies where self-generation in the industrial sector continues to be a common solution. Residential sector sales account for a large percentage of billing in most economies in the MENA region. The residential energy volume per connection shows a significant difference across HICs (17.9 MWh), UMICs (4.1 MWh), and LMICs (2.3 MWh). The reasons are the same as for total sales per connection. Income levels are also significant for total billing per connection—HICs ($925) have larger values than UMICs ($419) and LMICs ($245); and for residential billing per connection (HICs are at $478, UMICs at $313, and LMICs at $168). The percentage of installed meters is important: to measure consumption and manage demand when required, metering of all consumption points is needed. A large majority of customers have a meter (96 percent). If this is representative across the region, it indicates strong performance by international standards. Notably, the percentage in HICs (57 percent) is significantly below that in UMICs and LMICs (100 percent); this probably reflects government policy toward consumers in some HICs in the sample. Information on the duration of interruptions indicates that values are signifi- cantly higher in UMICs than in LMICs. However, the sample of utilities answer- ing this question is extremely small, so little import should be attached to this result. The results obtained suggest that income is likely to affect certain perfor- mance indicators in two ways. Higher-income countries may decide to support consumers by reducing tariffs via some form of subsidy. This then affects perfor- mance on indicators that relate to revenues or costs. Also, higher-income coun- tries tend to have greater demand for power, often met by larger utilities. Where there are genuine scale effects in supply costs, then performance will tend to be better in higher-income countries. The presence of a relation between income and performance in MENA indicates that care should be taken when comparing performance across utili- ties without taking national income levels into account, and also in trying to understand the nature of the relationship between economy income and utility performance. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 92 Drivers of Utility Performance: Institutional and Contextual Characteristics Conclusion The primary objective of chapter 5 was to use evidence on utility performance from the MENA Electricity Database to explore whether cross-sectional (inter- utility) differences in performance are correlated with key institutional and contextual variables. A number of important conclusions can be drawn from the results described above: 1. The tests carried out, despite lack of data for certain indicators, reveal a sub- stantial number of cases where performance indicators are correlated with one (or more) of the drivers—25 of the 36 indicators had some significant link with a driver of performance, and 14 of these had significant correlations with more than one driver. 2. About 30 percent of indicators had significant correlations to utility type, util- ity size, and national income; about 20 percent had the same to ownership (public or private) and the presence of a separate regulator. 3. Approximately 50 percent of results across three categories of indicators (cost efficiency, labor efficiency, and consumption and billing) were significant; the same was true for only 4 percent of results for system and loss efficiency. These results provide evidence of links between drivers and performance overall, without taking individual circumstances into consideration. The bunch- ing of significant results by indicator category suggests that certain areas of utility performance are more affected by policies linked to utility type, size, ownership, and regulation, whereas other areas of performance show few links to these policy-related drivers. Organizational Structure Unbundling the power sector has been said to permit more focused manage- ment and to increase the possibilities of competitive behavior once a market is liberalized. Utility type was significant for only a few indicators,11 but for these the results were coherent and highly significant. DUs performed much better than VIUs in distribution losses (10 percent versus 20 percent, respectively) and in technical and nontechnical losses, analyzed separately. The differences were also significant for collection rate (89 percent versus 69 percent, respectively) and for SAIFI (1.23 versus 3.18, respectively). The ROE for DUs was 7.0 per- cent, whereas that for VIUs was −23 percent. Energy sales per connection were significantly higher for VIUs (12.6 MWh) than for DUs (4.2 MWh), and there was a similar result for residential sales per connection. Testing for size differ- ences revealed that size was not significant for these variables, so that the differ- ences between VIUs and DUs could not simply be assigned to scale. More likely, DUs were established in areas where sales per connection tended to be lower (especially for nonresidential customers). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 93 The results suggest that DUs were better able to focus on business with end consumers and thus able to focus efficiency drives in a meaningful fashion. VIUs are more broadly focused, and their role as a national provider means that they may be required to pursue goals such as increasing employment, keeping consumer tariffs low through cross-subsidies, and keeping nonpaying customers connected. Size In considering power sector reform strategies, options such as unbundling VIUs and introducing more than one utility of the same type to better focus on core business and introduce a form of competition will reduce average utility size. Traditional analysis of the power sector has emphasized the role of economies of scale when discussing long-run pricing strategies. The factor of size can pull in two directions, so there is interest in considering the importance of size in the MENA context. The size factor was significant across about one-third of indicators, and it was notably insignificant for system and operational efficiency and losses efficiency indicators. Values for OPEX/employee, OPEX/kWh, and OPEX/km indicate the potential importance of scale economies. The group of big utilities had values significantly lower than the medium and small utilities (and the differences between these two groups were not significant). A similar pattern was found for energy sales per employee and total revenues per employee. The debt-to-equity ratio was much higher for big (1143 percent) than for medium (499 percent) and small utilities (330 percent), whereas the big utilities had a lower assets-to-liabilities ratio (79 percent) compared with medium utili- ties (84 percent) and small utilities (200 percent). Total billing per connection was ($155) for big utilities, whereas the values for medium ($404) and small ($381) utilities were significantly higher. These results suggest that size relates to performance in two ways. Economies of scale lead to lower OPEX per normal- ized scale factor (employee number, kilowatt-hour, kilometer), but that larger utilities also maintain very high debt-to-equity ratios, low assets-to-liabilities ratios, and low billing rates per connection, probably through the use of govern- ment finance and guarantees. Ownership Introducing private ownership is one of the main policy tools for improving the performance of power utilities. The direct introduction of the profit motive can be expected to prompt utilities to reduce costs as well as increase sales. One of the most direct ways to reduce costs may be to tackle overstaffing. Although public and private performance differed on only a few indicators, it was clear that private plants utilize less labor to achieve the same production. OPEX/employee was significantly higher where there was private owner- ship ($417,000) as opposed to public ownership ($18,000). OPEX itself was three times higher for the public utilities, but employment was six times as high—suggesting considerable overstaffing. Similarly, significant differences Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 94 Drivers of Utility Performance: Institutional and Contextual Characteristics were found for OPEX/km, total revenues per employee, and sales per employee. Total billing per connection was significantly higher for private ($464) than for public ($259) utilities. The ROE was also significantly higher for private (15.5 percent) than for public (0 percent) utilities, which is consistent with private utilities placing more emphasis on the profit motive. A major difficulty with the use of this indicator in cross-section studies is the selection bias. Governments do not privatize a ran- dom selection of some or all of the utilities in the sector. Rather, they may select those that are already well performing, on the basis that these will be easier to privatize and are the ones in least need of continuing government support. This selection pattern produces a positive correlation between performance and ownership—but, notably, it is not a causal link. Presence of a Separate Regulatory Agency How the presence of a separate regulatory agency affects utility performance depends on what is being regulated. If the primary focus is the tariff level, then governments that wish to set tariffs low for political reasons, through the use of subsidies, will not introduce a regulator. This produces a positive correlation between the presence of a separate regulator and certain indicators (for example, energy sales to OPEX). Thus, the presence of a separate regulator has to be ana- lyzed in context. A group of indicators was found to have significantly lower values for utilities operating in the presence of a separate regulator than for those without one. These indicators included OPEX/kWh ($0.07 versus $0.15, respectively), OPEX/km ($16,944/km versus $36,469/km, respectively), residential connec- tions per employee (205 versus 472, respectively), sales per employee ($117,000 versus $279,000, respectively), and revenues per employee ($132,000 versus $327,000, respectively). There is no apparent reason why the presence of a sepa- rate regulatory agency should produce such differences, and it is more likely that context is the deciding factor. Income The income level of the economy in which a utility is situated may well influence the utility’s performance independent of size and structure. Two effects may be involved: (a) at higher per capita incomes, the consumption of electricity per household increases steadily, with an income elasticity of around unity and (b) in economies with high income levels, governments may be more willing to subsi- dize utilities to keep consumer tariffs low. In the present study, the income factor was significant for one-third of the indicators. OPEX per employee increased significantly with income: the value for LMICs ($159,000) was significantly lower than for UMICs ($293,000) and HICs ($400,000), and OPEX per connection showed a similar pattern. Labor costs composed a larger share of OPEX in LMICs (16 percent) than in HICs (11 percent). An increase in OPEX per connection as an economy’s income level rises to that of an HIC is to be expected: much of the demand increase as Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Drivers of Utility Performance: Institutional and Contextual Characteristics 95 incomes rise is due to the same households purchasing more, rather than to low-income households deciding to connect to the grid (thereby increasing the number of connections but decreasing the average consumption per connection). The fact that energy volumes sold per connection increase sharply with income (HICs at 28.8 MWh and LMICs at 3.9 MWh) is more evidence of the impact of income on demand. A rise in OPEX/employee is consistent with the existence of higher wages per employee at higher income levels, and labor costs as a share of OPEX could fall for the same reason. Energy sales as a percentage of costs declined from 91 percent in LMICs to 56 percent in HICs, probably because of the willingness of HICs in the region to charge lower tariffs and subsidize the utilities. The ratio of debt to equity was significantly higher in LMICs (1,065 percent) than in HICs (376 percent), sug- gesting that more-developed economies were able to work with more acceptable levels of risk. Notes 1. See Jamasb, Nepal, and Timilsina (2015) for a broader review and Vagliasindi and Besant-Jones (2013) for a detailed analysis of organizational structures in low-income countries in the power sector. 2. Vagliasindi and Besant-Jones (2013) show that unbundling can deliver performance improvements, but not for all indicators. They emphasize that unbundling works best when part of broader reforms (for example, regulatory reforms and increased compe- tition in generation and distribution) and for large systems in countries with a certain threshold of development (as measured by per capita income). They also find that partial unbundling is not effective. 3. Using 2013 World Development Indicators data, the correlation between net energy imports (as a percentage of energy use) and gross domestic product per capita (in purchasing power parity) is −0.77, which indicates a high correlation between these two variables. 4. Bacon and Besant-Jones (2001) describe the traditional approach to power sector reform, whereas Eberhard and Gratwick (2011) discuss how this approach has since evolved. Vagliasindi and Besant-Jones (2013) provide a more recent evaluation of approaches to power sector reform. 5. The need to tailor reform strategy to the size of the power system was argued by Bacon (1995). 6. Eight economies out of the 14 considered in this study have an electricity regulator, whether it is independent or not. 7. HICs are those with a gross national income (GNI) per capita above $12,736; UMICs have a GNI between $4,126 and $12,735; and LMICs, a GNI in between $1,046 and $4,125 (see World Bank Country and Lending Groups Database at https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank​ -country-and-lending-groups; and World Bank Open Data at http://data​ .worldbank.org). 8. A value greater than 5 percent indicates that there is no difference between the means. Probability values of 10 percent or less are noted as indicating weak support for significant differences between the subgroups. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 96 Drivers of Utility Performance: Institutional and Contextual Characteristics 9. Excluding Iraq, for which the values were so large as to dominate any common relationships. 10. A weak effect—showing private distributors with a higher ratio of collection losses to total revenues than public utilities—was due to a single observation on the West Bank’s Tubas distributor, where collection losses were 144 percent of total revenue. No significant difference was found when this observation was omitted. 11. It should be noted that testing for structural differences was not meaningful for seven indicators. References Bacon, R. 1995. “Appropriate Restructuring Strategies for the Power Generation Sector: The Case of Small Systems.” Industry and Energy Department Occasional Paper 3, World Bank, Washington, DC. Bacon, R., and J. Besant-Jones. 2001. “Global Electric Power Reform, Privatization and Liberalization of the Electric Power Industry in Developing Countries.” Annual Review of Energy and Environment 26 (1): 331–59. Also as: Energy and Mining Sector Board Discussion Paper 2, World Bank, Washington, DC. Cambini, C., and D. Franzi. 2013. “Independent Regulatory Agencies and Rules Harmonization for the Electricity Sector and Renewables in the Mediterranean Region.” Energy Policy 60 (September): 179–91. Eberhard, A., and K. N. Gratwick. 2011. “IPPs in Sub-Saharan Africa: Determinants of Success.” Energy Policy 39: 5541–49. Galal, A., L. Jones, P. Tandon, and I. Vogelsang. 1994. Welfare Consequences of Selling Public Enterprises: An Empirical Analysis. New York: Oxford University Press. Jamasb, T., R. Nepal, G. R. Timilsina. 2015. “A Quarter Century Effort Yet to Come of Age: A Survey of Power Sector Reforms in Developing Countries.” Policy Research Working Paper WPS 7330, World Bank Group, Washington, DC. Jones, L., P. Tandon, and I. Vogelsang. 1990. Selling Public Enterprises: A Cost-Benefit Methodology. Cambridge, MA: MIT Press. Newbery, D., and M. G. Pollitt.1997. “The Restructuring and Privatization of the U.K. Electricity Supply—Was It Worth It?” Public Policy for the Private Sector, Note 124, World Bank, Washington, DC. Vagliasindi, M., and J. Besant-Jones. 2013. Power Market Structure: Revisiting Policy Options. Directions in Development. Washington, DC: World Bank. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 PA R T I I What Do the Country Case Studies Tell Us? These four case studies (of the Arab Republic of Egypt, Jordan, Morocco, and Oman) offer insights relevant to the Middle East and North Africa (MENA) region and beyond. The studies aim at providing not only an overview of each country’s power sector but also an analysis of utility performance to help iden- tify potential areas of improvement. The narrative and figures presented in these chapters focus on the year 2013 (as did part I). Although previous chapters compared utilities regionwide and across a range of performance indicators, chapters 6 to 9 compare utilities with one another and also with regional median values. The four countries chosen for the case studies have undertaken significant reforms of their electricity sectors over the past decades. These countries have a wide variety of characteristics and challenges representative of the 14 MENA economies of this study. In a region where the sector is mostly publicly owned and centralized under vertically integrated utilities (VIUs), Egypt, Jordan, Morocco, and Oman each have a story to tell, whether in relation to their dependence on fossil fuel imports, their population size and geographical spread, or the initial and organizational structure of their electricity sector. As illustrated in appendix B (table B.1), Egypt, Jordan, Morocco, and Oman had gone through some degree of unbundling in their electricity sectors in 2013. By then, private sector Shedding Light on Electricity Utilities in the Middle East and North Africa   97   http://dx.doi.org/10.1596/978-1-4648-1182-1 98 What Do the Country Case Studies Tell Us? involvement was well developed in Jordan and Oman. Egyptian utilities remained state owned with the exception of some independent power producers (IPPs) for generation. Morocco’s electricity sector structure involved a single VIU (Office National de l’Electricité et de l’Eau Potable, ONEE) with the electricity distribu- tion activities of most cities being delegated to 11 municipal entities, of which four are privately owned. A number of factors exogenous to the electricity sector have affected the performance of the region’s utilities. These factors include—but are not limited to—political instability (as in Egypt over the years 2011–14), disruptions in pri- mary fuel supply (as in Jordan, where the entire sector was reformed due to gas supply interruptions, resulting in a radical shift in the energy mix), and both the direct and indirect spillover effects of regional armed conflicts (an influx of dis- placed populations from the Levant in countries such as Jordan, for instance, have resulted in a stark increase in population and, consequently, demand). The case studies cover countries that have addressed, in different manners, the link between water and energy, which cannot be left unmentioned in the MENA region. Desalination plants are an integral part of the energy sector in the member countries of the Gulf Cooperation Council (GCC), supplying both municipalities and industries for the past two to three decades (Al Hashemi and others 2014). Also, several energy utilities are involved in water or sanitation activities. These two trends can be observed in Oman, where desalination ­ activities are common among several electricity generation utilities (GUs), and in Morocco, where the 11 distribution utilities (DUs) are also involved in water and sanitation activities. Also of interest is the introduction of renewable energies in the energy mix, in a region in which fossil fuels remain the dominant source of electricity, mostly due to their abundance and the conventional generation technologies and prac- tices that have been in place for several decades. In 2013 in Morocco, 31 percent of total installed capacity was from renewables (of which 7 percent was not hydropower). Oman, in contrast, depended entirely on thermal power genera- tion, with natural gas and diesel oil making up 98 percent and 2 percent of the energy mix, respectively (AER 2014). With several members of the MENA region benefiting from an abundance of solar and wind resources, the region’s potential has yet to be exploited and is lagging behind other world regions mostly because renewable energy sources are disregarded in policy design. Each of the four case studies here start out with a brief historical overview before detailing the main characteristics of the electricity sector’s three main activities: generation, transmission, and distribution. This overview is followed by a discussion of the relative performance of GUs and a discussion of DUs. The scope is limited by the availability of data. Yet this represents a good start at developing analysis that might, in turn, inform ways to address the major chal- lenges identified in this report. A synthesis of the evolution of the sector from 2014 until the writing of this book in 2017, which in some cases has gone through important reforms, is briefly presented before the concluding section. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 What Do the Country Case Studies Tell Us? 99 References AER (Authority for Electricity Regulation). 2014. Annual Report 2014. Muscat, Oman: AER. Al Hashemi, R., S. Zarreen, A. Al Raisi, F. A. Al Marzooqi, and S. W. Hasan. 2014. “A Review of Desalination Trends in the Gulf Cooperation Council Countries.” International Interdisciplinary Journal of Scientific Research 1 (2): 72–96. http://www.iijsr.org/data​ /frontImages/gallery/Vol._1_No._2/6.pdf. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 6 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt The electricity sector in the Arab Republic of Egypt is led by the Ministry of Electricity and Renewable Energy (MoERE), established in the early 1960s, with the main mandate of securing electricity supply at the national level. A reform in the early 2000s resulted in the corporatization of the power sector into an Egyptian joint stock (holding) company: the Egyptian Electricity Holding Company (EEHC). Following this reform, a legal unbundling of gen- eration, ­transmission, and distribution assets took place: six generation utilities (GUs), nine distribution utilities (DUs), and the Egyptian Electricity Transmission Company (EETC) were created, all of which are 100 percent owned by EEHC. Additionally, three private GUs were established under 20-year build-own- operate-transfer (BOOT) contracts with EETC, which since 1996 has operated thermal power plants with a combined installed capacity representing 6.4 per- cent of Egypt’s total installed capacity of 32 gigawatts (GW).1 Figure 6.1 shows the current structure of the Egyptian electricity sector. Since 2001, the sector’s regulation has been mandated to the Electric Utility and Consumer Protection Regulatory Agency (EgyptERA), which regulates, supervises, and controls electric-power-related activities, including generation, transmission, distribution, and consumption. EgyptERA’s mission is to ensure electricity supply, quality, and access at equitable prices, while considering envi- ronmental issues. In the current electricity market structure, EETC acts as single buyer and is the only utility licensed for extra high voltage (EHV) and high voltage (HV) electric- ity transmission. The EETC purchases electrical energy from the six GUs, the three private ones, and a small independent power producer (IPP) as well as from the New and Renewable Energy Authority (NREA). It then sells the electricity to the nine DUs and to approximately 100 EHV and HV consumers. In addition, the EETC conducts energy sales and exports with neighboring economies over the existing interconnections. Shedding Light on Electricity Utilities in the Middle East and North Africa   101   http://dx.doi.org/10.1596/978-1-4648-1182-1 102 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt Figure 6.1  Electricity Sector Organization, Arab Republic of Egypt Ministry of Electricity and Renewable Energy New and Nuclear Egyptian Hydro Atomic Nuclear Renewable Materials Electricity Power Energy Power Energy Authority Holding Plant Authority Plant Authority Company Executive Authority Authority Distribution companies Egyptian Public generation companies • North Cairo Electricity • Cairo Transmission • South Cairo • East Delta Company • Alexandria • West Delta • Canal • Middle Egypt • North Delta • Upper Egypt • South Delta • Hydro Power • Behaira Private generation companies • Middle Egypt • Suez Gulf • Upper Egypt • Port Said • Sidi Krir Electricity Generation The GUs produce electricity, which is sold to EETC, and are responsible for the management, operation, construction, rehabilitation, and overhauling of power plants. As shown in table 6.1, out of 32,015 megawatts (MW) of installed capacity, thermal power plants represent 89 percent, while hydropower and renewable energy (wind and solar) represent 9 percent and 2 percent, respectively. Figure 6.2 indicates that the technology most often used is steam (43 percent), followed by combined cycle (35 percent) and gas (11 percent). Table 6.1  Generation Mix, Arab Republic of Egypt, 2013 Generation type Amount Hydropower generation (MW) 2,800 Thermal power generation (MW) 26,480 New and renewable energy (wind and solar) (MW) 687 Private sector BOOTs (thermal) (MW) 2,048 Total installed capacity (MW) 32,015 Total generated energy (GWh) 168,050 Source: EEHC 2013/14. Note: BOOT = build-own-operate-transfer; GWh = gigawatt-hours; MW = megawatts. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 103 Figure 6.2  Share of Technology Type in Generating Electricity, Arab Republic of Egypt, 2013 Percent Renewable, 2 Hydro, Gas, 9 11 Combined, 35 Stream, 43 Source: EEHC 2013/14. Electricity Transmission EETC is the single public entity responsible for managing, operating, and main- taining the electric transmission grid on EHV and HV levels across Egypt. Table 6.2 includes basic data on the transmission lines and substations of the transmission utility (TU). Egypt’s geographical position allows for electricity exchanges to take place through existing regional interconnections, namely with Libya and Jordan. In 2013, Egypt exported 460 gigawatt-hours (GWh), almost eight times more than the amount it imported (61 GWh) (EEHC 2013/14). An electrical inter- connection between Egypt and Saudi Arabia is currently under i ­mplementation, and the possibility of Egypt–Sudan and Egypt–Ethiopia–Sudan connections is under study.2 Table 6.2  Electricity Transmission Data, Arab Republic of Egypt, 2013 Transmission Amount Total transmission lines and cables (132 kV, 220 kV, 500 kV) km 44,213 High voltage (66 kV and 33 kV) substation capacity MVA 99,635 Source: EEHC 2013/14. Note: km = kilometers; kV = kilovolts; MVA = megavolt ampere. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 104 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt Electricity Distribution The nine DUs responsible for the distribution and sale of electric energy pur- chased from EETC sold a total volume of 120,826 GWh to 30.6 million cus- tomers in 2013. Of this energy volume, 51.3 percent was distributed to the residential sector (see figure 6.3), which represents 73 percent of all medium- and low-voltage customers. The DUs are also responsible for managing, operat- ing, and maintaining the medium- and low-voltage grid, as well as for preparing, for instance, forecasts of customer demand. The share of private DUs does not exceed 1 percent of the market. Table 6.3 includes basic data for the distribu- tion lines and substations. Figure 6.3  Energy Sold from Distribution Utilities by Sector (medium- and low-voltage consumers), Arab Republic of Egypt, 2013 Percent Other, 9.7 Industry, Public lighting, 14.3 4.7 Agriculture, Commercial, 4.6 4.1 Gov. and public utilities, 11.3 Residential, 51.3 Source: EEHC 2013/14. Table 6.3  Electricity Distribution Data, Arab Republic of Egypt, 2013 Transmission Amounts Distribution transmission lines length (km) 425,611 Distribution substation capacity (MVA) 64,956 Customers Number of consumers (millions) 30.6 Source: EEHC 2013/14. Note: km = kilometers; MVA = megavolt ampere. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 105 Comparison of Egyptian Generation Utilities The unbundled nature of Egypt’s electricity sector makes it possible to conduct a comparative performance assessment of both GUs and DUs. Table 6.4 com- pares five public Egyptian GUs against one another through a set of character- istic and performance indicators.3 Comparisons are also made with Middle East and North Africa (MENA) median values when available.4 The first set of indicators listed in table 6.4 characterizes the type and size of GUs: all of which are big, at above 1 GW. Egyptian GUs have high installed capacities ranging from 3.4 GW for the Upper Egypt Electricity Production Company (UEEPC) to 6.2 GW for the Cairo Electricity Production Company (CEPC). The utilities mainly operate thermal power plants (including gas, steam, and combined cycles) and tend to be highly staffed (the West Delta Electricity Production Company [WDEPC] has the largest number of employ- ees, at 8,577). The capacity factor indicates how much of the plants’ potential capacity was used during the year. This factor indicates that units were working between 57 percent and 70 percent of their full capacity: although Cairo, East Delta, and West Delta Electricity Production Companies have capacity factors similar to the MENA median (58 percent), GUs in Middle Delta and Upper Egypt have higher figures. The availability factor—that is the percentage of a total year that plants were in service—is similar, and ranges from 79 percent to 87 percent, with the exception of UEEPC, which has an availability factor of 91 percent. This indica- tor depends on generation outages, whether they are caused by failure, mainte- nance, or the availability of fuel. All Egyptian GUs have values of operating expenses (OPEX) per employee lower than the MENA median ($297,000); Cairo and Upper Egypt GUs’ values are about half this median, whereas the other three GUs have significantly lower values (and a higher number of employees). Given that all these GUs operate with similar types of technologies and fuel use, a lower ratio could imply over- staffing and therefore greater labor inefficiencies. The cost structure indicators show that most OPEX is for fuel and lubricant (ranging from 79 percent to 88 percent), rather than labor. The share of fuel in OPEX is lowest for the Middle Delta Electricity Production Company (MDEPC) (79 percent) and could be a direct consequence of the generation technology used by this utility, which is essentially combined cycle. This utility’s relatively low fuel expenses could also explain why it has one of the highest capacity factors (65 percent). CEPC has among the highest OPEX values, yet the second-lowest labor cost as a share of total OPEX (8 percent). These figures are consistent with the fact that the three Delta GUs have more staff than the others. The cost-recovery indicators show that none of the GUs in Egypt recover their total OPEX or their total costs from sales, with the exception of MDEPC, which recovers its OPEX but not its total costs. The GU in Upper Egypt, on the other hand, has the lowest recovery rates both of OPEX (59 percent) and Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 106 Table 6.4  Comparing the Performance of Generation Utilities across Indicators, Arab Republic of Egypt, 2012/135 Indicator name Unit Cairo East Delta West Delta Middle Delta Upper Egypt Median MENA General Installed capacity GW 6.2 5.9 5.0 4.8 3.4 — Net generation TWh 31 31 25 27 21 — Employment employees, 5.4 7.0 8.6 6.2 3.2 — thousands Technology type % Gas (10), steam Gas (42), steam Steam (77), CC Steam (8), Steam (56), — (48), CC (41) (38), CC (20) (18), steam (5) CC (92) CC (44) OPEX $ millions 752 647 625 470 587 — Technical and Capacity factor % 58 60 57 65 70 58 operational Availability factor % 82 87 84 79 91 93 OPEX/employee $ thousands 138 91 77 76 179 297 Financial (Cost Share of cost of fuel, lubricant in structure) total OPEX % 88 88 81 79 83 75 Share of labor cost in total OPEX % 8 10 15 12 5 12 Financial (Cost Energy sales/total OPEX % 91 97 76 139 59 109 recovery)a Energy sales/total costsb % 61 68 56 83 46 107 Financial (Balance Accounts receivable Days 412 222 603 274 571 40 sheet) Debt/equity % — 3,484 3,074 2,509 1,270 357 Current assets/current liabilities % 52 37 67 68 56 95 Financial Return on assets % 0.02 0.02 0.01 0.03 0.35 3.00 (Profitability) Return on equity % 0.6 0.3 0.1 0.4 3.0 7.0 Source: MENA Electricity Database. Note: CC = combined cycles; GW = gigawatts; MENA = Middle East and North Africa; OPEX = operating expenses; TWh = terawatt-hours; — = not available. a. The values of MENA medians above 100 percent are mainly driven by Omani generation utilities (12 of 23 used in this study), which have 193 percent and 112 percent median values, respectively, for the two cost-recovery indicators. b. Data from regulator. The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 107 total costs (46 percent). According to EgyptERA, in 2013, a total of $1.6 billion was provided to the five GUs in the form of government subsidies. When OPEX are not being recovered, this could indicate that electricity is being underpro- duced, resulting in insufficient sales. But Egyptian utilities have high availability and capacity factors. Another possibility could be the high cost of fuel for genera- tion, yet Egyptian fuel is in fact highly subsidized. The inability of GUs to recover their costs must therefore be from low tariffs. The accounts receivable of Egyptian GUs are very high, from six to 15 times higher than the MENA median. But the average number of days involved could hamper the recovery of OPEX from sales. In the management of day-to-day activities, these delays can cause cash shortfalls, causing, for example, deferrals of scheduled maintenance. In Egypt, compromising maintenance activities might not be an issue, because plant availability factors are already quite high. However, there seems to be a relation between accounts receivable delays and low OPEX recovery from sales. The debt-to-equity ratios of Egyptian GUs are extremely high, from 1,270 ­ percent to 3,484 percent. This is between four and 10 times higher than the MENA median of 357 percent. Unable to recover costs, utilities most likely make use of debt instruments to finance their activities, or at least to cover their operating costs. Not surprisingly, when it comes to mobilizing liquid assets to repay short-term debts, Egyptian GUs are also underperforming, at below 70 percent on average for all utilities. Based on the financial stance of these utilities, it comes as no surprise that the return on assets (ROA) and return on equity (ROE) figures are close to 0 percent. The median value for ROA and ROE in the MENA region is 3 percent and 7 percent, respectively, whereas the most profitable utility, UEEPC, has ROA and ROE values of 0.35 percent and 3 percent, respectively. Comparison of Egyptian Distribution Utilities Table 6.5 compares nine Egyptian DUs across a set of performance indicators. The right-hand column presents median values for MENA DUs, thus allowing for a broader comparison beyond Egypt. Egyptian DUs perform technically well when considering two indicators: the load factor and distribution losses. Both have values close to the regional MENA median except for the Canal Electricity Distribution Company (CEDC), with a low load factor of 38 percent. High load factors in general lead to lower distribution losses, yet this is not necessarily the case observed in Egypt. This could be a result of high nontechnical losses (due to theft and erroneous meter readings), which contribute about 25 percent on average to total distribution losses.6 If the losses could be improved by 1 percentage point, this would result in a savings of 1,626 GWh per year, equivalent to about $71 million.7 A potential area of improvement would therefore be the reduction of nontechnical distribution losses, which could be a way of increasing the ROE of the utilities. Currently these ROE values are very low, as shown in table 6.5. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 108 Table 6.5  Comparing the Performance of Distributors across Indicators, Arab Republic of Egypt, 2012/13 North South North South Middle Upper MENA Category Indicator name Unit Cairo Cairo Alexandria Canal Delta Delta El Behera Egypt Egypt Median Technical and Load factor % 62 64 61 38 69 60 62 69 68 60 operational Distribution losses % 10 8 11 6 9 10 10 11 8 10 OPEX/employee $ thousands/ 46 47 24 47 39 28 35 37 38 188 employee OPEX/connection $/connection 160 169 134 230 101 75 157 115 119 346 OPEX/kWh sold $/kWh 0.04 0.04 0.04 0.04 0.03 0.03 0.04 0.03 0.03 0.1 OPEX/km $ thousands/ 12 14 15 10 9 10 9 6 6 19.6 km Commercial Total billing/ $/connection 138 148 111 197 97 68 132 96 101 299 (Consumption connection and billing) Financial (Cost Share of labor cost in % 21 21 41 20 24 35 26 27 26 12 structure) total OPEX Financial (Cost- Energy sales/OPEX % 88 87 83 86 96 91 84 84 85 93 recovery) Energy sales/total % 84 82 — 80 86 83 75 75 73 88 costs Financial (Balance Accounts receivable days 188 293 82 62 256 79 186 117 182 121 sheet) Debt/equity % 850 1,282 — 685 677 523 527 501 571 523 Collection rate % 93 86 99 94 84 93 95 92 88 93 Current assets/ % 71 81 77 66 97 103 103 85 113 85 current liabilities Financial Return on assets % 0.19 2.6 0.18 1.87 0.3 0.23 0.04 0.06 0.06 3.00 (Profitability) Return on equity % 0.6 8.8 0.3 7.7 0.8 0.5 0.1 0.1 0.2 7 Source: World Bank calculations. Note: km = kilometer; kWh = kilowatt-hours; MENA = Middle East and North Africa; OPEX = operating expenses; — = not available. The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 109 The OPEX per employee across Egypt’s DUs is much lower than the MENA median of $188,000. This is because they have the largest number of employees in the region, ranging from 8,083 for the Upper Egypt Electricity Distribution Company (UEEDC) to 17,917 for the South Cairo Electricity Distribution Company (SCEDC). By comparison, the DU with the largest number of employees outside Egypt is Morocco’s Lyonnaise des Eaux de Casablanca (LYDEC), with 3,850, which covers electricity, water, and sanita- tion for the region of Casablanca. Meanwhile, OPEX per employee varies across DUs, ranging from $24,000 for the Alexandria Electricity Distribution Company (AEDC) to $47,000 for SCEDC. The high cost of overstaffing is evident in labor’s share of total OPEX—ranging from 20 percent for CEDC to 41 percent for AEDC. This is two to three times higher than the MENA median. Moreover, this is despite the relatively low cost of labor in Egypt. On the commercial front, OPEX per connection oscillates between $75 for SDEDC to $230 for CEDC. These are low values when compared to the MENA median. For total billing per connection, South Delta has the lowest value and Canal Electricity the highest. In the case of CEDC, a high OPEX per connection could indicate that interruptions in supply are regularly and promptly solved. This would lead to customers having a continuous supply of electricity and, hence, allow their consumption to be high. Another reason for this high OPEX per connection and high billing rate might be that the utility spends money and effort in tracking bills and ensures that collection is frequent. It is also interesting to observe that the OPEX to sell a unit of energy in Egyptian DUs is two to three times higher than the MENA median. All Egyptian DUs boast high collection rates; six out of nine utilities are close to or above the regional median value of 93 percent. Yet in almost all cases, collec- tion periods are long (at 62 days, CEDC’s is the shortest among Egypt’s DUs; North Cairo’s is the longest, at almost 6 months). This can be attributed to delayed collection cycles resulting from the time-consuming manual registration of read- ings and bills. The EEHC has been exploring the option of shifting to smart meters since 2013 as a way to reduce both nontechnical losses and the time involved in bill collection. High debt-to-equity ratios and low current ratios suggest that the Egyptian DUs are not financially independent and rely heavily upon debt- financing instruments and mechanisms. Although this is a common trend among DUs in the MENA region (the median value of debt-to-equity is 523 percent and for the current ratio, 85 percent), Egyptian utilities face finan- cial constraints largely due to tariff levels, which did not allow full costs to be recovered from sales in 2013. Revenues of DUs in Egypt also included elements beyond pure electricity sales, as was the case for CEDC, which also included subsidies for the electricity exported to Gaza. None of the Egyptian DUs recover their OPEX from sales, and at best recover only 86 percent of their total costs through sales. Profits, on the other hand, are positive when considering revenues from electricity sales as well as other Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 110 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt sources, but remain negative if only electricity sales are considered, explain- ing why the DUs in table 6.5 all have positive ROA and ROE. Evolution of Egypt’s Electricity Sector since 2014 The electricity sector in Egypt has gone through a number of changes since 2014 that are worth mentioning, given that this analysis is based on 2013 data. In 2014, Egypt embarked on an ambitious energy subsidy reform and laid out its plans to phase out subsidies within five years to reach 0.5 percent of gross domestic product (GDP) by 2019, with the remaining subsidies covering only liquefied petroleum gas (LPG) and electricity consumption of the poorest house- holds. The fiscal burden of Egypt’s energy subsidies had grown continuously over the two decades up to 2014: the budget share of energy subsidies increased from 9 percent to 22 percent between 1990 and 2014. Electricity prices have risen cumulatively over the past three years, by more than 85 percent across consumer categories, and fuel prices have been raised twice, ranging from an accumulated increase of 60 percent to 150 percent across different fuel products from 2015 to 2017. Three successive electricity tariff increases and two major petroleum price reforms since 2014 have reduced energy subsidies from almost 7 percent to around 2.6 percent of GDP between 2014 and 2017 (as projected). In the electricity sector, the process is led by EgyptERA, the electricity regu- lator. For the years 2018 and 2019, it is planned that the regulator will present Egypt’s Cabinet with (a) the current average electricity tariff charged to con- sumers; (b) an estimate of the average electricity tariff consistent with cost- recovery based on actual fuel costs, fuel mix, and foreign exchange costs applicable in each year; and (c) an estimate of the average electricity tariff consistent with electricity subsidy targets. Based on these inputs by EgyptERA, the Cabinet would decide the average tariff, and the board of EgyptERA would approve the associated tariff structure to be issued by a ministerial decree. This institutional process, which strengthens the position of the regulator beyond what it was, is underpinned by the new Electricity Law No. 87/2015 and sup- porting executive regulations and has been successfully piloted during the tariff revision for 2017, enabling the regulator to raise tariffs beyond the origi- nal five-year trajectory. The energy sector is being prioritized for governance reforms due to its higher institutional capacity. In the electricity sector, the MoERE has decided to set up a modern governance structure for new generation assets, with a separate company for each of the three 4.4 GW combined cycle gas plants under construction, and using international norms for staffing and skills. Other initiatives include (a) setting up an internal audit department in the EEHC for the first time; (b) publishing the methodology for determining electricity tariffs across consumer categories for the first time, based on Cabinet approval; (c) initiating a business planning framework for all sector entities; and (d) implementing the decision of the EgyptERA to conduct public hearings on key policy issues from 2018. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 111 Following the electricity shortages of summer 2014, the sector has advanced a major investment program aimed at improving the security of supply. However, significant inefficiencies remain in both the dispatch of the genera- tion plant and the operation of the transmission and distribution networks. The new Electricity Law No. 87/2015 (DPF 1 Prior Action 1.5) envisages a full modernization of the sector. Its provisions strengthen the authority and transparency of the regulator and provide for an eight-year transition toward a competitive market. A critical first step is the separation of the EETC from its current role as a subsidiary of the EEHC, to become a network operator independent of generation and distribution activities, improve transparency and accountability of state-owned entities, promote competition and private investment in the sector, and provide nondiscriminatory third-party access to the grid. Egypt has barely begun to develop its rich renewable energy resources, which include excellent conditions for commercially viable wind power as well as high- intensity direct solar radiation throughout its territory. Egypt’s early investments in renewable energy were government owned; however, its ambitious plans to dou- ble the share of its generation capacity coming from renewable sources to 20 percent by 2022—and thus reduce reliance on fossil fuels—call for a substantial scale-up in private investment. The new renewable energy law (no. 203/2014) reduces risks and improves the financial viability of investments in wind power and solar photovoltaics (PV), improving the climate for private sector investment. The law and its associated feed-in tariff regulations provide incentives for the first 4,300 MW (wind and solar PV) as well as a regulatory framework for further private investment through competitive bidding mechanisms for IPPs. Moreover, recent increases in grid electricity tariffs combined with declining costs of renewable energy are increasingly making renewable energy solutions more competitive from an end-user perspective. Conclusion In this chapter, we analyzed the performance of Egyptian electricity utilities in 2013. Egypt has some serious challenges to overcome in the electricity sector, and maximizing the efficiency of its utilities can help achieve this. The period 2011–14 witnessed two peaks of political instability, which makes the issue of satisfying the increasing electricity demand a highly sensitive one, in particular when three- quarters of all the electricity volume sold by the DUs is destined for the residen- tial sector. Although thermal power plants represent almost 90 percent of total capacity in the country, the sharp decline in oil and natural gas production— changing Egypt’s status from a net exporter to a net importer—makes electricity supply even more challenging. The five GUs studied in this chapter are big (3 GW to 6 GW of installed capacity) and rely on thermal power plants (gas, steam, and combined cycles). Although they perform reasonably well from a technical standpoint, Egyptian Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 112 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt GUs do not recover their OPEX (with the exception of the Middle Delta Electricity Distribution Company), although Cairo and East Delta are not far from doing so. GUs have financial performance indicators that are of real concern when compared to the rest of the region. In addition, overstaffing appears to be another area of concern, particularly in the three Delta GUs. The nine Egyptian DUs are big as defined by this study (that is, more than 2 million connections) except for the El-Behera Electricity Distribution Company (EEDC), which is of medium size. Again, overstaffing appears to be a key area of concern: for example, labor’s share of total OPEX is two to three times bigger than the MENA median. The OPEX per unit of energy sold is three to four times bigger than the median of the rest of the region. No Egyptian DU recovers its OPEX from sales, but all values are above 80 percent. Most balance sheet indicators show poor performance (though to a degree not nearly as concerning as that of GUs). Finally, the profitability of DUs is low: ROA and ROE tend to be low, with the notable exception of ROE for both South Cairo Electricity Distribution Company and CEDC, which are above the MENA median. EEHC’s expansion plan indicates that 3,000 MW of additional electricity generation capacity would need to be added every year to meet 2020’s forecasted demand. What this chapter has shown is that simply expanding supply will not be sufficient to improve the performance of the Egyptian electricity sector. The financial situation of Egyptian GUs is so delicate that financial restructuring will presumably be needed (for example, by utilities raising the companies’ equity through conversion of the public debt into equity). In addition, tariff reforms are required if the sector is to be financially viable. Improved efficiency of electricity operators and better corporate governance will inevitably be part of the solution to the sector’s challenges. Last but not least comes the issue of data collection and its quality. Most of the quantitative evidence on performance provided in this chapter is not available online, and required numerous exchanges with the regulator to check the validity of figures and establish a common understanding of the factors behind specific values. Even after these efforts, some values were left aside as they did not appear to be reasonable. The exercise of carrying out periodic performance assessments with the direct involvement of the GUs and DUs should be reinforced by the regulator. The multiplicity of genera- tion and distribution actors within the same economy provide a unique opportunity to benchmark performance across operators. But for these exer- cises to be of use, operators need to adopt international accounting standards and a cost accounting system (it remains unclear which Egyptian utilities have done so) and DUs should implement supervisory control and data acquisition (SCADA) (absent across Egypt’s DUs). Without these, the reli- ability of many of the financial and commercial indicators collected is deemed to be very low. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 The Urgent Need for Sector Reforms: The Case of the Arab Republic of Egypt 113 Notes 1. These three private generation utilities have not been included in this study. 2. Within the framework of the Eastern Africa Power Pool and Nile Basin Initiative Plans. 3. Other Egyptian generation utilities—namely the publicly owned Hydro Power Plants Electricity Production Company (2,800 MW) and private thermal generation—are not included in this analysis because data were not collected. 4. We had insufficient data for non-MENA generation utilities to make a meaningful comparison. 5. The indicators energy sales/total OPEX, energy sales/total costs, and accounts receiv- able are not applicable to generation utilities based on how these indicators were categorized for the purpose of the MENA Electricity Database. However, for com- parative purposes, their values are presented and discussed in this chapter but not in previous chapters of this book. 6. Based on a calculation of seven distribution utilities in Egypt. 7. Savings of Egyptian LE 486 million (EEHC 2014). References EEHC (Egyptian Electricity Holding Company). 2014. Annual Report 2013/2014. http:// www.moee.gov.eg/english_new/EEHC_Rep/REP-EN2013-2014.pdf. EgyptERA (Electric Utility and Consumer Protection Regulatory Agency). Cairo: EEHC. http://www​.egyptera.org. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 7 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan In 1994, the government of Jordan initiated an electricity restructuring and reform program that opened the sector to private sector involvement. Today the electricity sector is unbundled, and sector policy is set by the Ministry of Energy and Mineral Resources (MEMR). Generation utilities (GUs), whether public or private, sell their electricity to a single buyer, the National Electric Power Company (NEPCO), which is also the fuel purchaser bearing all costs and risks related to fuel price fluctuations. NEPCO also acts as a transmission system operator (TSO), manages the nation’s electricity transmission infrastructure, and sells electricity to the three main distribution utilities (DUs): the Jordan Electric Power Company (JEPCO), Irbid District Electricity Company (IDECO), and Electricity Distribution Company (EDCO). In addition to the Central Electricity Generation Company (CEGCO) and Samra Electric Power Generating Company (SEPCO), four independent power producers (IPPs) are also present: AES Jordan (the first IPP in Jordan), the Qatrana Electric Power Company (QEPCO), the Amman Asia Electric Power Company (AAEPC), and AES Levant. Figure 7.1 shows the organization of the Jordanian electricity sector as of 2014. The introduction of the General Electricity Law No. 64 in 2002 marked an important milestone. Soon after, the Electricity Regulatory Commission (ERC) was established as an autonomous regulatory body tasked with licensing the country’s electric utilities (generation, transmission, and distribution). In 2014, its mandate was expanded to include regulation of other forms of energy— namely nuclear and mining activities—and it is now known as the Energy and Minerals Regulatory Commission (EMRC). Another milestone was the introduc- tion of the Renewable Energy Law in 2012. According to this law, self-production is authorized by law, with the possibility of either net metering or selling excess electricity to the grid. Shedding Light on Electricity Utilities in the Middle East and North Africa   115   http://dx.doi.org/10.1596/978-1-4648-1182-1 116 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan Figure 7.1  Electricity Sector Organization, Jordan, 2014 Ministry of Energy and Mineral Resources (MEMR) Generation companies Energy and Minerals Regulatory Commission (EMRC) Central Electricity Generating Company Samra Electric Power Company Distribution companies Jordan Electric Power Company IPPs National 1. AES-Jordan PSC Electric 2. Qatrana Electric Power Company Power Irbid District Electricity Company 3. Amman Asia Electric Power Company Company 4. AES Levant (NEPCO) Electricity Distribution Company Source: World Bank. Note: IPP = independent power producer. Table 7.1  Generation Mix, Jordan, 2013/14 Generation type Amount (MW) Hydropower generation 12.0 Steam 791.0 Diesel 27.0 Gas 621.0 Combined cycle 1,737.0 Wind 1.4 Biogas 3.5 Total installed capacity (MW) 3,193.0 Total generated energy (GWh) 17,886.0 Source: NEPCO 2013. Note: GWh = gigawatt-hours; MW = megawatts. Electricity Generation In 2013, thermal power plants represented 99 percent of the installed capacity of GUs. Hydropower and renewable energy made up the remaining 1 percent. Table 7.1 presents Jordan’s installed capacity by technology. Until 2010, around 80 percent of the electricity generated was from natural gas imported through the Arab gas pipeline. Frequent interruptions in supply led to the use of costlier secondary fuels for generation, that is, diesel oil and heavy fuel oil (HFO). The GUs, however, are shielded from the risk of fuel costs by NEPCO, to which they also sell all the electricity produced. Figure 7.2 illustrates the percent- age share of each fuel from 2009 to 2013, clearly showing that the energy mix of Jordan shifted in these years, reaching more than 90 percent natural gas and more than 75 percent of HFO and diesel by 2013. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 117 Figure 7.2  Share of Fuel Type in Electricity Generation, Jordan, 2009–13 100 90 80 70 60 Percent 50 40 30 20 10 0 2009 2010 2011 2012 2013 Heavy fuel Natural gas Diesel Source: NEPCO 2012, 2013. Electricity Transmission NEPCO is the state-owned single buyer of all electricity produced in Jordan except for renewable energy sources that are directly connected to the distribu- tion network. NEPCO also has the role of system operator and is responsible for managing and operating the Jordanian electricity transmission grid, which consists of 132 kilovolts (kV) and 400 kV networks. Table 7.2 shows data related to the transmission network length and substation capacities. The transmission system it operates interconnects the power generation plants with the load centers. The total length of the transmission network is about 4,463 kilometers (km) of circuit and includes 400 kV tie lines with Syria. Table 7.2  Electricity Transmission Data, Jordan, 2013 Transmission 2013 Total transmission lines and cables (132 kV and above) (km) 4,463 High voltage (132 kV and 33 kV) substation capacity (MVA) 7,444 Source: NEPCO 2013. Note: km = kilometers; kV = kilovolts; MVA = megavolt ampere. Electricity Distribution Three DUs operate in Jordan, each covering a certain geographical region. Although JEPCO’s service area includes industrial areas as well as the capital city of Amman, the other two DUs cover mostly rural areas. The total amount of electricity sold in 2013 amounted to 13.8 terawatt-hours (TWh), of which 62 percent was sold by JEPCO, which has a customer base of over a million customers. Table 7.3 lists basic data for the distribution lines and substations. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 118 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan As in most of the economies in the Middle East and North Africa (MENA) region, most of Jordan’s electricity is used in the residential sector, as shown in figure 7.3. Table 7.3  Electricity Distribution Data, Jordan, 2013 Distribution Amount Distribution transmission lines, km 57,635a Distribution substation capacity, 400/132/33 MVA 3,760 Customers Number of consumers 1,744,000 Source: NEPCO 2013. Note: EDCO = Electricity Distribution Company; IDECO = Irbid District Electricity Company; JEPCO = Jordan Electric Power Company; km = kilometers; MVA: megavolt ampere. a. Sum of length of distribution network of EDCO, JEPCO, and IDECO, according to the MENA Electricity Database. Figure 7.3  Volume of Energy Distributed by Sector, Jordan, 2013 Percent Street lighting, 2 Agriculture, 15 Residential and government buildings, 43 Commercial, 15 Industrial, 25 Source: NEPCO 2013. Electricity Tariffs between Utilities The main responsibility of EMRC is to set tariffs ensuring that the prices charged by licensees are sufficient to finance their activities and allow them to earn sufficient return on their investments. As far as DUs are concerned, EMRC sets both the tariff between them and end consumers and the tariff that they Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 119 pay to NEPCO. As far as private GUs are concerned, the tariff that NEPCO pays them is set once the generator enters the market in accordance with a power purchase agreement (PPA). CEGCO’s tariff was determined at the time of its privatization. The other GUs—that is, the IPPs, in the current single-buyer model—compete for the market after NEPCO specifies the capacity and energy needs, location, and time when capacity is required. The winner is the generator with the lowest levelized price. For generator plants, the cost of fuel is passed through, and NEPCO provides the fuel (or pays its cost). The mecha- nism is different for the only state-owned GU, SEPCO, for which the tariff is decided by EMRC following a request by SEPCO and NEPCO. Comparison of Jordanian Generation Utilities Six GUs in Jordan are compared against one another in table 7.4, as well as to the MENA region median.1 Although most of the data defining the general character- istics of GUs is available, this is not always the case for the technical and opera- tional, commercial, and financial indicators. Most of the data gaps concern the IPPs, from which it was a challenge to obtain data. Jordanian GUs are thermal power plants running on HFO, natural gas, and diesel. The sizes of the six Jordanian GUs are heterogeneous: AES Levant, AES PSC, and QEPCO are considered small in this study (less than 500 megawatts, MW); AAEPC, medium (500 MW to 1 gigawatt, GW); and CEGCO and SEPCO, big (above 1 GW). Looking at the ratio of installed capacity to employees or the ratio of operating expenses (OPEX) to employees, CEGCO, AAEPC, and SEPCO appear to be overstaffed. The result for CEGCO could be explained by the fact that it is the oldest and largest GU in the country, and high staff numbers could be customary. Another reason could be the ownership structure of CEGCO: the private sector owns 51 percent of the utility’s shares, the public sector owns 49 percent (of which the Government of Jordan owns 40 percent, and the social security corpo- ration owns 9 percent). This argument is strengthened by the fact that the IPPs have a much lower staff number than publicly owned utilities: 47 for AES Levant, 51 for AES PSC, 75 for QEPCO, and 287 for AAEPC. In the case of AAEPC, the plant was not fully operational in 2013, and most of the employees numbered here might have been those outsourced during the precommissioning phase. According to the MENA Electricity Database (MED), AAEPC had 135 outsourced employees and 152 full-time employees in 2013. The utilities with the lowest capacity factors are the two IPPs that were not fully operational in 2013—AES Levant and AAEPC. The high capacity factors for AES PSC and QEPCO, on the other hand, could be explained by the fact that they are both private utilities and have to fulfill contractual obligations regarding sales to NEPCO under their PPAs. In the case of CEGCO, several old generating units have retired, and, due to regulatory constraints, the utility does not, in principle, have the right to add new units, which could explain the low capacity factor.2 Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 120 Table 7.4  Comparing the Performance of Generation Utilities across Indicators, Jordan and MENA Median, 20133 Category Indicator name Unit AES Levanta AAEPCa AES PSC CEGCO QEPCO SEPCO MENA median General Installed capacity MW 246 573 370 1,687 373 1,031 — Net generation TWh 0.6 0.3b 2.6 7.4 2.4 4.5 — Employment Employees 47 287 51 1,037 75 345 — Fuel mix n.a. HFO, natural HFO, natural Natural Natural gas, diesel, Natural gas, Natural gas, — gas gas, diesel gas HFO, small diesel diesel hydro, wind OPEX $, millions — 89 472 1,401 440 833 — Techbnical and Capacity factor % 28 6.3 80 50 75 50 58 Operational Availability factor % 99 — — 90 98 — 93 OPEX/employee $, millions/ — 0.3 9.3 1.3 5.9 2.4 0.3 employee Financial (Cost Share of cost of fuel, lubricant in structure) total OPEX % — 63 98 94 99 97 75 Share of labor cost in total OPEX % — 1.3 0.7 1.3 — 0.7 12.0 Financial (Cost Energy sales/total OPEX % — 20 68 99 75 101 109 recovery)c Energy sales/total costs % — 14 — 96 71 95 107 Financial Accounts receivable Days — — 62 98 91 50 40 (Balance Debt/equity % — 290 333 354 621 876 357 sheet) Current assets/current liabilities % — 123 287 95 488 113 95 Financial Return on assets % — — — 12 5 4 3 (Profitability) Return on equity % — — 36 21 25 17 7 Source: World Bank calculations. Note: AES Levant = AES Levant Holding BV Jordan PSC; AAEPC = Amman Asia Electric Power Company; AES PSC = Amman East Power Plant; CEGCO = Central Electricity Generation Company; HFO = heavy fuel oil; MENA = Middle East and North Africa; MW = megawatts; n.a. = not applicable; OPEX = operating expenses; QEPC = Qatrana Electric Power Company; SEPCO = Samra Electric Power Generating Company; TWh = terawatt-hours; — = not available. a. Fully operational only in 2014. b. Not fully in operation in 2013, which explains the low electricity output and capacity factor. c. The values of MENA medians above 100 percent are mainly driven by Omani generation utilities (12 out of 23 used in this study), which have high values, as shown by the median values of 193 percent and 112 percent for the two cost-recovery indicators. Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 121 Because Jordan is one of the only economies in the region not to have sub- sidies for fuel, OPEX values4 are much higher than those observed elsewhere in the region. In fact, the majority of GUs’ spending is on fuel, ranging from 95 percent to 99 percent. Consequently, the share of labor costs in OPEX is considerably smaller than the MENA median (12 percent). CEGCO has the largest number of employees, yet its OPEX per employee, estimated at $1.3 million, is seven times the MENA median value of $267,000 per employee. AES PSC has the highest OPEX per employee value in Jordan, at $9 million. In the case of AAEPC, which was not yet fully operational in 2013, this indicator might be misleading, particularly when considering the net generation figures in table 7.4, which show that AAEPC generated about 15 times less electricity than SEPCO. For an installed capacity of 1,687 MW and net generation of 7.4 TWh, CEGCO compares poorly with SEPCO, which generated 4.5 TWh (equivalent to 60 percent of CEGCO’s energy output), with three times fewer employees and almost two times less OPEX. The high fuel costs as a share of OPEX could be explained by analyzing the fuel mix of the two utilities. SEPCO produced three times more electricity than CEGCO from natural gas sources (representing 25 percent of SEPCO’s total generation). Natural gas is more efficient and could be the reason why SEPCO has lower OPEX even though the share of fuel costs in its OPEX is 70 percent. In the case of CEGCO, 72 percent of the electricity generated is from HFO, and because fuel costs con- stitute 94 percent of its OPEX, this explains why CEGCO also has the highest OPEX value among the GUs. Only one of the utilities in table 7.4 recovered their total OPEX from energy sales: SEPCO. AAEPC is a private GU and is therefore expected to recover its OPEX, yet it has the lowest OPEX recovery rate. This could be explained by the fact that AAEPC was not fully operational in 2013, as can be observed in the listed electricity output of 317 gigawatt-hours (GWh), which is low for an installed capacity of 573 MW (compared with the 2,591 GWh output of AES PSC’s 370 MW installed capacity). As far as full cost-recovery is concerned, although CEGCO and SEPCO are very close to fully recovering costs via energy sales, AES PSC and QEPCO are not quite there, and AAEPC is very far from cost-recovery (again, this could be because it was not fully operational in 2013). We now look at financial performance. The debt-to-equity ratio in Jordan is similar across GUs and in some cases lower than the MENA median, with the exception of the fully state-owned utility SEPCO (876 percent). SEPCO has been expanding its generation capacity since 2010 by adding new gener- ating units, which must have been financed mainly through debt rather than equity. The GUs are all profitable, with high return on equity (ROE) and return on asset (ROA) values. In particular, the IPPs’ ROE is twice the MENA median value of 7 percent. This is mainly a result of the provisions of the IPP contracts under which they operate and the PPAs with NEPCO that ensure that IPPs can Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 122 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan actually make a profit. In addition, NEPCO absorbs the risk of fuel price fluctua- tions and so protects GUs from it. Comparison of Jordanian Distribution Utilities Table 7.5 compares the three DUs operating in Jordan against one another, using several performance indicators. The MENA median values are also included where available. While EDCO is considered a small DU in this study (less than 250,000 connections), IDECO and JEPCO are considered medium (250,000 to 2 million connections). Among the three DUs, EDCO (34 per- cent) has the lowest load factor, whereas IDECO has the highest (56 percent). The load factor depends upon the amount of electricity distributed and the peak load, which both vary according to the consumption patterns and type of consumers the DU serves. Usually, the higher the load factor, the lower the distribution losses in the distribution system, and this is indeed the case for IDECO, whose distribution losses, at 11 percent, are the lowest of the three utilities. This value is similar to the MENA median value of 10 percent. The distribution losses of JEPCO are the highest, at 14 percent, because the utility distributes the largest amount of elec- tricity and across a much longer set of medium- and low-voltage networks than the other two (27,000 km for NEPCO against almost 19,000 km and 12,000 km for IDECO and EDCO, respectively).5 OPEX per employee figures for Jordanian DUs are much higher than the MENA median, most likely due to the high OPEX figures attributed to other costs (labor costs represent only about 6 or 7 percent of OPEX for each utility listed in table 7.5). EDCO covers the largest service area (68,359 square kilometers, km2) and also has the smallest number of connections, which could explain the high OPEX costs per connection ($1,483), as well as the high OPEX per kilowatt-hour sold ($0.23). For IDECO, the values of OPEX per connection and OPEX per kilowatt-hour are the closest to the MENA median value, and closer than those of EDCO and JEPCO. It is the most efficient utility in terms of operational per- formance, also having the lowest OPEX per kilometer among the three DUs. However, JEPCO seems to underperform in this category (although this might not necessarily be the case), mainly as a result of its relatively high OPEX, which could be linked to the larger amount of energy purchased to supply to its consumers. Total billing per connection is highest among customers serviced by EDCO ($1,528) and lowest for IDECO ($576). This could be the result of a difference in the tariffs applied by the two utilities. In the case of JEPCO, total billing per connection is high. This reflects high consumption in the capital, Amman, and also among the industrial consumers serviced by JEPCO, as compared with the more rural consumers serviced by the other two utilities (the agricultural sector represented only 5 percent of JEPCO’s sales, whereas it represented 12 percent and 11 percent of sales for IDECO and EDCO, respectively). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 123 Table 7.5  Comparing the Performance of Distributors across Indicators, Jordan and MENA Median, 2013 Category Indicator Name Unit EDCO IDECO JEPCO MENA median Technical and Operational Load factor % 34 56 51 60 Distribution losses % 12 11 14 10 OPEX/employee $, thousands/ 230 197 448 188 employee OPEX/connection $/ connection 1,483a 547 1,038a 346 OPEX/kWh sold $/kWh 0.23a 0.10 0.14 0.1 OPEX/km $, thousands /km 26 12 43 19.6 Commercial (Consumption Total billing/connection % 1,528a 576 936a 299 and billing) Financial (Cost structure) Share of labor cost in % 6 7 7 12 total OPEX Financial (Cost recovery) Energy sales/OPEX % 97 107 93 93 Energy sales/total costs % — 99 — 88 Financial (Balance sheet) Accounts receivable Days 117 120 122 121 Debt/equity % 1,476 981 576 523 Collection rate % — — 97 93 Current assets/current % 99 84 80 85 liabilities Financial (Profitability) Return on assets % 5 6 24 3 Return on equity % 16 20 12 7 Source: MENA Electricity Database. Note: EDCO = Electricity Distribution Company; IDECO = Irbid District Electricity Company; JEPCO = Jordan Electric Power Company; km = kilometers; kWh = kilowatt-hours; MENA = Middle East and North Africa; OPEX = operating expenses; — = not available. a. Outlier not used in calculations of MENA average and median values mentioned in earlier chapters of this book. The only utility that fully recovers OPEX from sales is IDECO, although the values for EDCO and JEPCO are not far from cost-recovery. All Jordanian DUs have high debt-to-equity ratios, suggesting a high level of financing through debt. (This is common in the MENA region, where the median value is 523 percent.) The current shares for the three utilities are below 100 per- cent, with the highest value belonging to EDCO (99 percent). IDECO’s low value could be attributed to the utility’s low contribution of cash to current assets. JEPCO has the lowest current ratio, suggesting that the current liabili- ties are extremely high. Although JEPCO does have high receivables, this is unlikely to be the cause of the low current ratio because JEPCO collects its receivables within 120 days (comparable to the MENA median value) and has a high collection rate of 97 percent. EDCO, JEPCO, and IDECO are profitable, as shown by the ROA and ROE indicators in table 7.5. Since the privatization of EDCO and IDECO in 2008, the revenues and profits of these utilities are controlled and regulated by the national regulator (EMRC), which reevaluates assets at the end of the concession period and at the time of obtaining a license. These licenses grant the companies a 10 percent profit on their regulatory asset base after the regulator reviews and approves their annual budgets, their projects, and anticipated electricity losses. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 124 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan Evolution of Jordan’s Electricity Sector since 2014 The electricity sector in Jordan has gone through a number of changes since 2014 that are worth mentioning, given that the analysis here is based on 2013 data. The government of Jordan faces the challenge of pursuing its reform agenda while also accommodating an influx of Syrian refugees. An estimated 1.3 million Syrian refugees are currently residing in Jordan—equivalent to over 20 percent of Jordan’s population before the start of the Syrian crisis in 2011. To mitigate sup- ply risks while keeping up with demand, the Jordan 2025 strategy, approved in 2015, set targets to (a) increase the share of local energy sources in the energy mix (from 2 percent in 2014 to 39 percent by 2025); (b) reduce the energy intensity of the economy; and (c) decrease the percentage of electricity transmission and distribution losses (from 17 percent in 2014 to 11 percent by 2025). The government’s reform program aims to lock in the achievements of energy sector reforms over recent years despite the additional strain of the Syrian crisis and further strengthen resilience to external shocks of fuel supply interruptions and price volatility. Key measures under the government’s multiyear reform program include (a) restoring the financial sustainability of the electricity sector, (b) diversifying gas import sources, (c) developing domestic energy resources, and (d) promoting energy efficiency. The government succeeded in restoring the financial sustainability of the elec- tricity sector by the end of 2015. The rising cost of fuels since 2010 had created a gross imbalance between costs and revenues for NEPCO. In 2013, the govern- ment adopted a five-year (2013–17) Electricity Tariff Adjustment Plan to restore the adequacy of NEPCO’s revenue base. A number of factors allowed NEPCO to reach cost-recovery in the final quarter of 2015: tariff increases, a decline in international oil prices after mid-2014, a switch from oil to cheaper natural gas dating from mid-2015, and the commissioning of the first large-scale renewable energy plant. The government is committed to locking in its reform achieve- ments through further tariff reforms with the aim of sustaining cost-recovery for NEPCO amid volatile energy import prices. Jordan’s private distribution sector comprises a number of bilateral perfor- mance agreements between the regulator and the DUs, with the aim of applying international best practices to achieve efficiency gains. Loss reduction targets for 2016 and 2017 were finalized by EMRC and agreed upon by the three DUs at the end of 2015, with a plan to agree on targets for 2018 and 2019 at the end of 2017. NEPCO has developed a holistic strategy for securing a supply of relatively clean fuel. Implementation of the strategy began in 2015. The main thrust of the strategy is the diversification of supply sources. Natural gas, most of which is imported in liquid form through the liquefied natural gas (LNG) terminal in Aqaba, is sourced through two multiyear LNG supply contracts and on the spot. These contracts allowed NEPCO to provide natural gas for 84 percent of power generation until mid-2016. However, all of Jordan’s long-term LNG imports Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 125 remain linked to the Brent oil price, which makes the country vulnerable to price shocks. In addition to LNG, Jordan is pursuing longer-term supply options— including piped gas from the Arab Republic of Egypt, Iraq, and the Eastern Mediterranean—to ensure a secure and clean fuel supply to its electricity sector in the long term. Renewable energy is procured from IPPs. A total of 30 IPP projects, totaling 1,374 MW, are now at various stages of development. PPAs for around 1,000 MW of capacity have been signed and around 240 MW are operational. This makes Jordan a leader in private-sector-owned renewable energy in the MENA region. Conclusion This chapter analyzed the performance of Jordanian electricity utilities in 2013. Jordan’s main challenge does not reside in energy access and infrastruc- ture development but is primarily in guaranteeing supply to meet increased demand and as a net energy-importing country. The challenge is even greater since the Syrian refugee influx. Because its main fuel imports for electricity generation were interrupted in 2010, the country has had to look for alterna- tive sources while depending on secondary options, such as relatively costly diesel and HFO.6 Of the six GUs studied in this chapter, half are considered small (AES Levant, AES PSC, and QEPCO); one, medium (AAEPC); and two, big (CEGCO and SEPCO). All are private except for SEPCO, and all rely on thermal production. CEGCO, AAEPC, and SEPCO have the worst ratio of employees to capacity, which could be an indication of overstaffing. Because Jordan is one of the only economies in the region not to have subsidies for fuel, its OPEX values are much higher than those observed in other economies in the region. This results in a high share of fuel costs, ranging from 95 percent to 98 percent of OPEX among GUs. Although CEGCO and SEPCO appear to almost fully recover their costs, QEPCO and AES PSC do not (and we did not have data for AES Levant). The debt-to-equity ratio in Jordan is mostly similar across GUs and in some cases lower than the MENA median. The GUs are all profitable, with high ROE and ROA values. In particular, the IPPs enjoy an ROE at least three times higher than the MENA median value of 5 percent. This is mainly a result of the provisions of the IPP contracts under which they operate and the PPAs with NEPCO that ensure that the IPPs can actually make a profit. Another reason is that NEPCO absorbs the fuel price fluctuation risk from GUs. The three Jordanian DUs are private. For the purposes of our study, two are considered of medium size (IDECO and JEPCO) and one as small (EDCO). IDECO appears to be the most efficient DU in terms of operational perfor- mance. The value of OPEX per employee across all DUs is very high, presum- ably because electricity purchase prices reflect the fact that the fuels used to produce it are not subsidized. The only utility that fully recovers OPEX from sales is IDECO, although the values for EDCO and JEPCO are not far from Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 126 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan cost-recovery. All Jordanian DUs have high debt-to-equity ratios, suggesting a high level of financing through debt. The three distribution utilities are notably profitable, as shown by the ROA and ROE indicators. Jordan’s main challenge is to provide affordable energy to end users while ensuring the profitability of mostly private generators and distributors in the wake of a significant energy transition. On the generation side, disruptions in the supply of natural gas have spiked interest in alternative sources of fuel, particularly renewable energy, which is high on the government’s agenda since the adoption of the 2012 Renewable Energy Law. Other options include the potential use of vast oil shale resources. On the supply side, the main issue is to strike the right balance so as to (a) reduce dependence on subsidies while mini- mizing the impact of tariff reforms on the poorest consumers on the one hand and (b) reduce the fiscal deficits of utilities operating in an already challenging environment on the other. Last but not least comes the issue of data collection and quality. Only part of the quantitative evidence on performance provided in this chapter is avail- able online, and collecting the rest required numerous exchanges with the regulator and utilities to check the validity of figures and establish a common understanding of the factors behind specific values. Even after these efforts, some values were left aside because they did not appear to be reasonable. The exercise of carrying out periodic performance assessments with the direct involvement of both GUs and DUs should be a central task of the regulator. Jordanian utilities are generally well equipped to collect reliable information. All utilities report having implemented supervisory control and data acquisition (SCADA), and adopted international accounting standards (IAS). But most GUs have not yet implemented a cost accounting system, so this is a pending task. Notes 1. Data for non-MENA generation utilities were insufficient to establish meaningful comparisons. 2. The total installed capacity in 2014 was 1,687 MW with available capacity of 1,267 MW whereas the installed capacity in 2009 was 1,747 MW with available capacity of 1,599 MW. 3. The indicators energy sales/total OPEX, energy sales/total costs, and accounts receiv- able are not applicable to generation utilities based on how these indicators were categorized for the purpose of the MENA Electricity Database. However, for com- parative purposes, their values are presented and discussed in this chapter but not in previous chapters of this book. 4. The cost of fuel was estimated for the Jordanian generation utilities based upon the aver- age cost of fuel per kilowatt-hour from the regulator and the amount of kilowatt-hours generated by each utility in 2013. This cost of fuel was then added to the operating costs to obtain the total OPEX as per the definition used in this study. 5. Figures from MED rounded to the nearest thousand. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Harvesting Results from a Restructuring of the Power Sector: The Case of Jordan 127 6. After 2013, the Jordanian government started buying liquefied gas in the Port of Aqaba and transporting it to generation plants as a solution to this problem, which has had positive effects on the sector. References EMRC (Energy and Minerals Regulatory Commission). 2013. Annual Report 2013. Amman, Jordan: EMRC. NEPCO (National Electric Power company). 2012. Annual Report 2012. Jordan: NEPCO. NEPCO (National Electric Power company). 2013. Annual Report 2013. Jordan: NEPCO. World Bank. 2017. MENA Electricity Database. World Bank, Washington, DC. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 8 Benefits and Challenges of Multiservice Providers: The Case of Morocco The Ministry of Energy, Mines, and Sustainable Development (MEMDD) is in charge of energy sector policy in Morocco. Office National de l’Electricité et de l’Eau Potable (ONEE),1 the national operator in charge of generation, transmis- sion, and distribution in large parts of the country, is under the control of MEMDD but is under the financial supervision of the Ministry of Economy and Finance. With respect to power distribution, the Ministry of Interior oversees public municipal operators (régies autonomes) as well as private distributors (socié- tés délégataires), which are elected by municipalities that grant public services concessions to the private sector. The setting and regulation of sector tariffs is the responsibility of an ad hoc interministerial committee (Commission des Prix) chaired by the Ministry of Governance and General Affairs (figure 8.1). As of 2013, Morocco did not have an independent regulator, although the establishment of one had been under consideration for some time.2 The electric- ity market was structured as a single-buyer model, in which ONEE acts as the sole buyer and supplier of bulk power.3 ONEE supplies power from its own generation plants, purchases it from licensed independent power producers (IPPs) or through its international interconnections, and sells it to other distribu- tion utilities and large industrial clients and through its own distribution grid. Electricity Generation ONEE is a state-owned vertically integrated utility (VIU), covering the generation (4,500 megawatts [MW]), transmission, and distribution of electricity. It has a monopoly on transmission operations and is the sole power supplier to distri- bution utilities. ONEE is also in charge of water and sanitation service delivery in large parts of the country. Whereas it is still the largest power producer of installed generation capacity (more than 63 percent), most electricity has been generated by IPPs since a 1994 amendment to the law governing ONEE’s Shedding Light on Electricity Utilities in the Middle East and North Africa   129   http://dx.doi.org/10.1596/978-1-4648-1182-1 130 Benefits and Challenges of Multiservice Providers: The Case of Morocco Figure 8.1  Electricity Sector Organization, Morocco ONEE own generation IPP generation Interconnections Autoproduction Capacity: 4,942 MW Capacity: 2,175 MW Production: 13,460 GWh Production: 12,738 GWh Imports: 5,551 GWh Production: 251 GWh (42%) (40%) (17.4%) (less than 1%) ONEE Single buyer Total demand 32 TWh ONEE (Direct supplier) Public and private utilities (58%) (42%) Distribution MV/LV EHV-HV direct customers MV/LV Customers (43%) (15%) 11739 GWh 11846 GWh 4212 GWh Source: Amegroud 2015. Note: EHV = extra high voltage; GWh = gigawatt-hours; HV= high voltage; LV = low voltage; MV = medium voltage; MW = megawatts; TWh = terawatt-hours. activities opened up generation to private operators.4 The remaining amount of electricity is imported from Spain and Algeria (5.5 terawatt-hours [TWh] in 2013) (Kharbat 2014). By the end of 2013, the total installed power capacity in Morocco was reported to be 7,994 MW, of which hydropower and renew- able energy (wind and solar) represented 32 percent, including hydro pumped storage (see table 8.1). Figure 8.2 illustrates the share of fuel types in generation. Thermal generation facilities are mostly used to produce electricity, with coal being the predominant fuel. Coal contributes significantly to the energy mix, with 31 percent of installed capacity; fossil-fuel-based power generation takes the lion’s share, at 68 percent of total installed capacity. The contribution of hydropower is 22 percent (includ- ing hydro pumped storage), accounting for the largest share of renewable energy generation. Morocco’s wind power production is the largest in the Middle East and North Africa (MENA) region, accounting for 10 percent of the nation’s power generation capacity. Four IPPs had a total installed capacity of 3,086 MW and supplied 52 percent of electricity in 2013.5 The largest of these by far is the Jorf Lasfar Energy Company (JLEC), owned and operated by the Abu Dhabi National Energy Company PJSC (TAQA). JLEC runs the largest coal-fired power plant in MENA. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 131 Table 8.1  Generation Mix, Morocco, 2013 Generation type (MW) Amount Hydropower generation 1,306 Pumped hydroelectric energy storage (PHES) 464 Steam power generation 3,145 Coal power generation (2,545 MW) HFO power generation (600 MW) Gas turbines power generation 1,230 Combined cycle power generation 850 Diesel power generation 202 Total thermal power generation 5,427 New and renewable energy (wind) 797 Total ONEE installed capacity (MW) 7,994 Total electricity generated (GWh)a 28,081,540 Source: ONEE 2014. Note: GWh = gigawatt-hours; HFO = heavy fuel oil; MW = megawatts; ONEE = Office National de l’Electricité et de l’Eau Potable. a. Does not account for imports from Spain. Figure 8.2  Generated Electricity in Morocco, by Technology Share, 2013 Percent Diesel, 3 Wind, CC, 10 11 Gas Hydro, turbines, 22 15 Steam (HFO), 7 Steam (coal), 32 Source: ONEE 2013. Note: CC = combined cycle; HFO = heavy fuel oil. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 132 Benefits and Challenges of Multiservice Providers: The Case of Morocco Electricity Transmission Transmission activity is carried out by ONEE. The state-owned monopoly is responsible for managing, operating, and maintaining the electric transmission grid and interconnections with neighboring economies. The Moroccan power system is interconnected with the Spanish and Algerian grids. Table 8.2 provides some figures on electricity transmission by ONEE. Electricity Distribution While ONEE is in charge of power distribution in most of Morocco’s cities and regions, there are 11 other electricity distribution entities, 7 public municipal utilities, and 4 private concession holders. With its 5.2 million clients, ONEE serves the largest number of consumers by far. The second-largest distribution utility is Lyonnaise des Eaux de Casablanca (LYDEC), which delivers electricity in Casablanca and Mohammedia to 0.9 million consumers. REDAL, which covers distribution in Rabat and Sale, has 0.5 million customers (MENA Electricity Database [MED]).6 The seven municipal public distributors, or régies autonomes de distribution, are RAEEF, in Fès; Régie Autonome de Distribution d’Eau d’Électricité et d’Assainissement liquide de la province de Kenitra (RAK), in Kenitra; RADEEL, in Larache; Régie Autonome de Distribution d’Eau et d’Électricité de Meknès (RADEM), in Meknès; Régie Autonome de Distribution d’Eau et d’Électricité de Marrakech (RADEEMA), in Marrakech; Régie Autonome de Distribution d’Eau, d’Électricité et d’Assainissement liquide des Provinces d’El Jadida et de Sidi Bennour (RADEEJ), in El Jadida; and Régie Autonome Intercommunale de Distribution d’Eau et d’Électricité de Safi (RADEES), in Safi. All these utilities are multiservice operators offering water and sanitation services as well. The private distributors manage concessions in Casablanca- Mohammedia (LYDEC, a privately owned utility with Engie as the main shareholder), in Tangier and Tetouan (Amendis, part of the French utility Veolia), and in Rabat-Sale (Redal, part of Veolia). Table 8.3 includes basic data on the distribution network and numbers of customers. Table 8.2  Electricity Transmission Data, Morocco, 2013 Transmission Amount Total transmission lines and cables (150 kV, 225 kV, 400 kV) km 22,995 High voltage (400 kV, 225 kV, and 60 kV/60 kV, 22 kV) substation capacity MVA 26,072 Sources: ONEE 2013 and MED. Note: km = kilometers; kV = kilovolts; MED = MENA Electricity Database; MENA = Middle East and North Africa; MVA = megavolt ampere; ONEE = Office National de l’Électricité et de l’Eau Potable. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 133 Figure 8.3 provides some insights on the heterogeneity of ONEE’s client base. For 2013, it shows the breakdown of energy sales per customer usage. Whereas 50 percent of sales were to distribution utilities, the residential sector accounted for 20 percent, the industrial sector for 12 percent, and the agricul- tural sector for 7 percent. Table 8.3  Electricity Distribution Data, Morocco, 2013 Transmission Amount Distribution transmission lines, medium voltage and low voltage (km) 243,568 Distribution substation capacity (MVA) 6,360 Customers (millions) Non-ONEE 2.9 ONEE 4.9 Total 7.8 Source: ONEE 2013. Note: km = kilometers; MVA = megavolt ampere; ONEE = Office National de l’Électricité et de l’Eau Potable. Figure 8.3  Share of Volume of Energy Distributed, by Sector, Morocco, 2013 Percent Industry, 12 Agriculture, 7 Distributors, 50 Tertiary, 7 Residential, 20 Public sector, 4 Source: ONEE 2013. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 134 Benefits and Challenges of Multiservice Providers: The Case of Morocco Comparison of Moroccan Generation Utilities Although the sample used for this study did not include the four Moroccan IPPs, this chapter provides information on one of them, JLEC.7 ONEE is also included in this section because it generates 42 percent of Morocco’s electricity consump- tion. However, the financial and technical indicators related to its generation activity should be considered with caution, because it is difficult to separate ONEE’s generation activity from the consolidated operating results (among other things, the utility provides water and sanitation services as well). Finally, and given the limited amount of data available for Moroccan generation utilities (GUs), we also look at an Egyptian GU (Upper Egypt Electricity Production Company [UEEPC]) as a comparator, in addition to the usual MENA median. Table 8.4 compares ONEE, JLEC, UEEPC, and the MENA median across several indicators. The Moroccan private GU has a capacity factor above the MENA median and similar to that of UEEPC. This could be a consequence of the power purchase agreement (PPA), whereby electricity purchase is guaranteed by Table 8.4  Comparing the Performance of Moroccan Generators across Indicators and against Egypt’s Upper Egypt Production Company and the MENA Median, 20138 Category Indicator name Unit ONEE JLEC Upper Egypt Median MENA General Installed capacity GW 4.9 2.0 3.4 — Net generation TWh 13.0 13.5 21.0 — Employment Employees 8,796 482 3,200 — Technical and Capacity factor % 31 75 70 58 Operational Availability factor % Between 75 and 80a 91 — 93 OPEX/employee $, thousands 284b 1,205 179 297 Financial (Cost Share of cost of fuel, % structure) lubricant in total OPEX 38.0 94.5 93.0 75.0 Share of labor cost in % total OPEX 10 5 5 12 Financial (Cost Energy sales/total OPEX % 118 153 — 109 recovery)c Energy sales/total costs % 87 129 — 107 Financial (Balance Accounts receivable Days 159b 45 — 40 sheet) Debt/equity % — 277 1,270 357 Current assets/current % liabilities 63b 247 56 95 Financial Return on assets % −4.4b 4.35 0.35 3.0 (Profitability) Return on equity % — 17.3 3.0 7.0 Source: World Bank calculations. Note: GW = gigawatts; JLEC = Jorf Lasfar Energy Company; MENA = Middle East and North Africa; ONEE = Office National de l’Électricité et de l’Eau Potable; OPEX = operating expenses; TWh = terawatt-hours; — = not available. a. Average of all the utilities under ONEE is between 75 percent and 80 percent, depending upon the year. b. Denotes values for ONEE, which are not disaggregated to the level of electricity generation. c. The values of MENA medians above 100 percent are mainly driven by Omani generation utilities (12 out of 23 used in this study), which have high median values, at 109 percent and 107 percent, respectively, for the two cost-recovery indicators. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 135 ONEE, and the utility is encouraged to maximize use of its generation capac- ity. The capacity factors varied between 31 percent in the case of ONEE’s power generation activity and 75 percent for IPPs such as JLEC and Energie Electrique Tahaddart (EET).9 These numbers show that private power pro- duction mainly serves the base load, and ONEE’s facilities operate as load followers and for peaking. The availability factor indicates the amount of time a power plant is available to generate electricity. This indicator varies greatly depending on type of fuel, plant design, and operations. It does not provide any indication of a plant’s con- version performance or utilization rates. Morocco’s private power producers generally boast high availability factors: 91 percent for JLEC and 93.4 percent in case of EET.10 The availability of ONEE’s thermal power plants range from an average of 53 percent for steam plants fueled by heavy fuel oil (HFO) to 98 percent for hydropower plants. The availability gaps observed in ONEE’s generation facilities can mainly be explained by the existence of repair and main- tenance issues. According to ONEE, its average availability factor ranges between 75 percent and 80 percent. There are significant differences in terms of technical performance between ONEE and the private GUs: for example, JLEC outperforms ONEE’s coal generation plants in terms of the heat rate. At the low end, one such coal plant, ­ Jerada, requires almost twice the quantity of coal to produce the same amount of power as JLEC.11 JLEC’s cost performance is strong and reflects its high availability and capacity factors. For GUs, operating expenses (OPEX) are associated with operating the power plant and generating electricity (fuel costs, maintenance, and administration). OPEX per employee values are mainly affected by the overall heat rate of the generating facility and its capacity factor, and also by its human resources management policy and the degree of reliance on out- sourcing and subcontracting. Even so, OPEX per employee for JLEC is rela- tively high ($1.2 million). The share of energy purchases and cost of fuel, lubricant, gas, and coal in total OPEX was 94.5 percent for JLEC. This is explained by the relatively high price of coal ($85 per ton)12 and the high capacity factor of the plant. Though an accurate estimation of ONEE’s generation activity OPEX is not available, we can say that JLEC’s share of energy purchases, while still dominant, is much smaller because of ONEE’s (a) aging generation facilities and hence high maintenance expenses, (b) larger number of employees and therefore relatively higher wage bill, and (c) relatively lower fuel bill as a result of a lower capacity factor. Table 8.4 shows that JLEC exhibits strong performance indicators and a very healthy financial profile compared to that of Upper Egypt. JLEC’s performance is to be viewed in light of its status as an IPP operating under a government- backed PPA, while Upper Egypt is a publicly owned utility. JLEC’s cost-recovery rate is comfortably high (153 percent for recovery of OPEX from sales), as would be expected from an IPP. This level reflects the strong profitability of the business. Morocco’s private power producers enjoy Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 136 Benefits and Challenges of Multiservice Providers: The Case of Morocco attractive contractual arrangements, and PPAs are designed to pass most market and institutional risks to ONEE. Cost-recovery indicators for this category of GUs are strong. The cost-recovery performance of ONEE’s generating activity is difficult to assess without detailed analysis of the overall costs of the VIU and its sales. This is further complicated by cross-subsidies between different types of clients and a complex tariff structure. ONEE’s electricity sales alone are not sufficient to cover total costs (that include depreciation and interest rates), as shown in the energy sales to total costs indicator, with a value of 87 percent in 2013. The average number of days for receivables from sales is 45 days for JLEC, which is higher than the regional median for GUs, at 40 days, while the number of days for receivables from sales is 159 days for ONEE. The debt-to-equity ratio for JLEC is 277 percent, which is relatively high but still much lower than the Egyptian GUs (1,270 percent for UEEPC). This is the typical ratio of an IPP, which often resorts to project financing to fund PPA- backed power generation infrastructures. In the case of the Arab Republic of Egypt, finance would be obtained in large part from the national budget in the form of direct subsidies allocated to the utility. At 17.3 percent, JLEC’s return on equity (ROE) was high, while its return on assets (ROA) was 4.35 percent. This compares positively with profitabil- ity performance indicators of GUs in the region, showing that the utility is highly profitable. ONEE, on the other hand, had a negative ROA in 2013 (−4.40 percent). To conclude, the impact of an imbalanced pricing structure has pushed ONEE to reduce or delay investments in maintenance and performance improvements, thereby increasing its focus on fulfilling its obligations under signed PPAs and electricity imports from Spain. Furthermore, the extensive use of IPPs with government-backed PPAs since 1994 has resulted in a situation where perform- ing assets are owned by new entrant private investors while a large proportion of risks are passed to ONEE (for example, fuel price, exchange rate). Distribution is also organized in such a way as to shield large distributors from market risks and the impact of volatile power generation costs. Comparison of Moroccan Distribution Utilities Table 8.5 presents indicators for the 11 distribution utilities in Morocco, as well as the MENA median values. Load factor values were obtained for all utilities; they ranged from a low of 45 percent for AMENDIS Tetouan to a high of 64 percent for RADEEJ. High load factors are representative of the load profile and indicate that the ratio of peak demand to average demand is relatively low. This in turn indicates that the industry’s stable consumption comprises a significant share of total demand. RADEEJ, LYDEC, AMENDIS Tanger, RAK, and REDAL had load factors higher than 56 percent, which is consistent with the industrial role across MENA (where the median value is 60 percent). Distribution losses in Morocco are lower Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Table 8.5  Comparing the Performance of Moroccan Distributors across Indicators and against the MENA Median, 2013 AMENDIS AMENDIS MENA Category Indicator name Unit Tanger Tetouan LYDEC RADEEL REDAL RAK RADEEMA RADEM RADEEJ RADEEF RADEES Median Technical and Load factor % 56 45* 58* 53* 56* 57* 51 48* 64 46* 52* 60 operational Distribution losses % 10 11 7 8 8 8 5 7 4 — 3 10 OPEX/employeea $ thousands 321 151 527 203 642 248 287 254 190 186 200 188 OPEX/connection $ 508 346 836 361 644 412 410 309 396 318 339 346 OPEX/kWh sold $ 0.12 0.15 0.20 0.12 0.17 0.12 0.10 0.11 0.10 0.11 0.13 0.1 OPEX/km $ thousands 36 31 96 19 49 20 32 21 21 37 32 19.6 Commercial Total billing/ $ thousands 473 299 520 — 442 306 466 301 436 312 302 299 (Consumption connection and billing Financial (Cost Share of labor cost % — — 12 — 14 — 8 — 12 — — 12 structure) in total OPEX Financial Energy sales/OPEX % — — 100 86 103 94 130 97 136 98 89 93 (Cost-recovery) Energy sales/total % — — 89 — 92 — — — 119 — — 88 costs Financial (Balance Accounts Days — — 76 — 121 — 205 — 106 — — 121 sheet) receivable Debt/equity % — — 279 — — — 41 — 66 — — 523 Collection rate % — — — — — — — — — — — 93 Current assets/ % — — 72 — 92 — — — 64 — — 85 current liabilities Financial Return on assets % 3 −1 — 6 2 — — 21 — — 14 3 (Profitability) Return on equity % 3 −2 18 7 10 — — 22 — — 16 7 Source: MENA Electricity Database except when marked with a “*” in which case obtained directly from utilities. Note: For LYDEC and REDAL, OPEX values are consolidated figures that include activities other than power distribution. OPEX = operating expenses; km = kilometers; kWh = kilowatt-hours; MENA = Middle East and North Africa; — = not available. a. Values reflect the estimated number of employees needed to support a utility’s electricity activities. 137 138 Benefits and Challenges of Multiservice Providers: The Case of Morocco than or equal to the MENA median of 10 percent, with the exception of AMENDIS Tetouan at 11 percent. Among municipal distributors, OPEX per employee varies from a minimum of $151,000 per employee for AMENDIS Tetouan, to a maximum of $642,000 per employee for REDAL. The number of employees provided by utilities was not disaggregated by function, and an estimate of those focused on electricity was used for the purposes of this study.13 The share of labor costs in total OPEX was low for the utilities that reported values, not exceeding 14 percent (REDAL). At 8 percent, RADEEMA reported the lowest value, while LYDEC and RADEEJ reported values of about 12 percent. This suggests that the OPEX are mainly made up of other costs such as the pur- chase costs of electricity from ONEE. Disregarding LYDEC and REDAL, the OPEX per connection was highest for AMENDIS Tanger ($508 per connection) and lowest for RADEM ($309 per connection). The separation of electricity and water within these municipal utilities is not as clear as within ONEE, and electricity services are often used to cross-subsidize the water services (as well as the heavy investments required in sanitation-related activities). Hence, it is common that tariffs are not related solely to energy consumption (for example, meter renting and maintenance, technical interventions, specific studies, or open- ing and closing accounts) and are fixed by each operator following munici- pal agreements. These fees can represent a substantial amount of the operator’s total revenues and are sometimes used by the operators to com- pensate for low national tariffs. RADEEJ has the lowest OPEX per kilowatt-hour sold ($0.10 per kilowatt-hours [kWh]). RADEEJ operates in a region in which the weight of industrial activities (medium-voltage clients) is significant, therefore explaining the low cost of main- tenance per energy sales. With regards to the OPEX per kilometer (km) among the 11 distribution utilities, it costs the most to maintain and operate 1 km of the existing distribu- tion network for LYDEC ($96,000 per km) while these costs are the lowest in the case of RADEEL ($19,190 per km). Considering the limited data available, it appears that as shown in table 8.5, LYDEC, REDAL, RADEEMA, and RADEEJ positively recover their operating costs from energy sales, whereas all the other utilities show values that are close but still below 100 percent. In terms of ROA, AMENDIS Tetouan showed negative values, while all Moroccan utilities showed low positive values, with the lowest shown by REDAL (2 percent) and the highest by RADEM (21 percent). The same trend for ROA is reflected in ROE, whereby AMENDIS Tetouan once again is the only utility with a negative value (−2 percent). In Morocco, RADEM reported a ROE of 22 percent, which is more than three times the MENA median. Although ONEE engages itself in distribution activities, it was excluded from this analysis because it does not publish usable commercial and financial Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 139 data on its distribution business. Some of the data published or provided by the two largest private distribution utilities, LYDEC and REDAL, are consoli- dated across all business activities (OPEX, head count, labor cost). The data used in the case of public municipal utilities (régies) were a combination of information directly collected from the utilities and data from the 2014 report on municipal public services, prepared by the Direction des Régies et Services Concédés. The discrepancies between Moroccan power distribution utilities mainly reflect the specific features of each utility. These include the following: • Type of customers. An important share of medium-voltage customers tends to push the load factor higher, as well as cost-recovery ratios, while keeping employment needs low (RADEEJ). • Economic activity. A thriving economy reflected by relatively high standards of living tends to have a positive impact on recovery and profitability indicators. Tariffs charged to households with high electricity consumption are generally more profitable (LYDEC, REDAL). • Geography. Operational costs and investments are generally higher for util- ities operating in extended geographical areas, covering scattered clients (RADEEL). • Local climate. Utilities operating in regions in the central part of the country are generally faced with higher operating and investment costs (RADEEMA, RADEEF). Evolution of Morocco’s Electricity Sector since 2014 The electricity sector in Morocco has gone through a number of changes since 2014 that are worth presenting, given that the analysis of this chapter is based on 2013 data. To alleviate ONEE’s poor financial state while simultaneously pushing the utility to improve its operational performance, in 2014 the government of Morocco and ONEE signed a framework contract (contrat programme) for the period 2014–2017. The goal of this financial restructuring plan was to help ONEE overcome a long-running precarious state of affairs. It focuses on tariff rate revisions, supplemented with an increase in capital and active help to collect former receivables from municipal utilities, public administrations, and municipalities. The plan also included a lump-sum payment to ONEE as a one-time flat subsidy for fuel oil used in electricity production to pave the way for a complete phase-out of all forms of oil subsidies. As a result, the net producer has gone from a deficit of about $285 million at the end of 2013 to a surplus of about $80 million at the end of 2016. In 2015 self-generation was further opened with two measures. First, the power sector’s legal framework for self-generation (above 300 MW) was fur- ther extended with the suspension of restrictions on capacity, type, and site of generation, provided that the total installed capacity be above 300 MW. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 140 Benefits and Challenges of Multiservice Providers: The Case of Morocco The new amendment recognizes the importance of providing more flexibility in terms of energy management to large industrial consumers such as in the mining sector. Second, for small self-generators, the scope of the renewable energy law was further widened to distribution grids. Private developers of renewable energy were allowed to connect their projects to the medium volt- age grid and were given access, albeit with some restrictions, to the end users. This step will be taken further to allow households and small businesses (and all other low-voltage clients) to install on-grid renewable distributed genera- tion equipment (such as rooftop solar kits), with the amendment of Law 13-09 in August 2015. In 2016, Morocco adopted a law introducing an independent energy regulator (Agence Nationale de Régulation de l’Energie—ANRE) and detailing its func- tions, missions, and organization. The role conferred on this new authority will be confined to policing the power generation regime introduced under the renewable energy law. The law also paves the way toward the separation of own- ership and operations of grids with generation and commercial activities. This cautious approach to the introduction of an independent regulation authority is considered to be more realistic than pursuing a body with wider prerogatives in an environment where the politics do not necessarily favor an active independent authority. Finally, six years after the creation of the Moroccan Agency for Solar Energy (MASEN), in 2016 the Moroccan government decided to extend the agency’s prerogatives to include the development and operation of all types of renewable energy facilities. The agency was renamed the Moroccan Agency for Sustainable Energy, and ONEE was required by law to transfer all its renewable energy assets to the new entity. This measure aims at emphasizing the role of renewable energy in future sector development and fast-tracking the implementation of the country’s targets in terms of the overall share of renewables in power production. Conclusion This chapter analyzed the performance of Moroccan electricity utilities in 2013. A singular characteristic of Morocco’s power sector is that most of its electricity service providers also provide water and sanitation services. The rationale was to use the relatively comfortable margins from electricity sales to subsidize water sales and finance investments in sanitation infrastructure networks and waste water treatment. ONEE is considered to be a big VIU in this study, with more than 2 million connections. Beyond being the single buyer in Morocco, it produces more than 40 percent of the electricity and distributes almost 60 percent of it. In addition to ONEE, in the generation segment we have included the biggest of the four Moroccan IPPs, JLEC. We find significant differences in terms of technical per- formance between ONEE and the private GUs: for example, JLEC’s heat rate Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 141 outperforms ONEE’s coal generation plants. JLEC’s cost performance is strong and reflects its high availability and capacity factors. In addition, it exhibits a very healthy financial profile compared to that of Upper Egypt. JLEC’s performance levels are to be viewed in light of its status as an IPP operating under a govern- ment-backed PPA, while Upper Egypt is a publicly owned utility. JLEC’s cost- recovery rate is comfortably high (153 percent for the recovery of OPEX from sales), as would be expected from an IPP. This level reflects the strong profit- ability of the business. Morocco’s private power producers enjoy attractive con- tractual arrangements, and PPAs are designed to pass most of the market and institutional risks to ONEE. Cost-recovery indicators for this category of GUs are strong. The cost-recovery performance of ONEE’s generating activity is difficult to assess without detailed analysis of the entity’s overall costs and sales. This is further complicated by cross-subsidies between different client types and by a complex tariff structure. This study also analyzed 11 distribution utilities, of which 5 were medium and 6 were small as defined by this study. Distribution utilities in Morocco are also involved in water and sanitation activities. They all purchase electricity from ONEE. All the small distribution utilities are public (municipal distributors) except for AMENDIS Tetouan, which is private, whereas three of the five medium utilities are private. The distribution utilities perform fairly well at the technical level, with values on technical performance indicators close to the MENA medi- ans. While labor costs of all distribution utilities make up 8 percent to 14 percent of OPEX, the ratio of OPEX to employees is higher than the MENA median for all except two utilities (RADEEF and AMENDIS Tetouan). All private utilities have strong profitability ratios, with the exception of AMENDIS Tetouan, which had a negative ROA and ROE. This can be explained by the fact that AMENDIS Tanger and Tetouan are actually a joint concession, and the private operator compensates Tetouan with the business in Tangier. Overall, data availability was the main obstacle to analyzing the financial performance of the municipal distributors. Last but not least come the issues of data collection and data quality. Only part of the quantitative evidence on performance provided in this chap- ter was publicly available for ONEE and JLEC. Very little is publicly available on the other Moroccan electricity utilities.14 Some values collected from utili- ties were left aside since they did not appear to be reasonable. There is space in Morocco to increase the performance information publicly available, which would also help the government carry out its regulatory tasks for operators. The issue of quality is of concern for public distribution utilities but not for private distribution utilities. The good news is that almost all Moroccan elec- tricity utilities have implemented supervisory control and data acquisition (SCADA). Meanwhile, only private distribution utilities have adopted inter- national accounting standards (IAS), and no public distribution utility reports data using cost accounting systems, while ONEE and private distribution utilities do use them. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 142 Benefits and Challenges of Multiservice Providers: The Case of Morocco Notes 1. Since 2012 and the adoption of law no. 40-09, the water and electricity activities of the former ONE (electricity) and ONEP (water) have been regrouped under one holding company, ONEE. 2. A 2015 law has since provided for the establishment of the ANRE (Agence Nationale de Régulation de l’Electricité), but its role is limited to the regulation of the renewable energy sector and associated transmission activities (tariffs and access conditions). 3. An incipient liberalized model has emerged under the framework of law no. 13-09. In 2014, these liberalized markets covered 2 percent of total electricity demand. In this framework, private large consumers have direct access to the transmission grid and can purchase part or all of their electricity from private renewable energy producers. 4. Decree No. 2-94-503 (September 1994). 5. None of these IPPs or generation utilities were included as part of the 67 utilities of this study due to lack of data availability. However, JLEC is included in this country case study. 6. See World Bank’s MENA Electricity Database. 7. The reason being that Taqqa Morocco shares floated on the Casablanca stock exchange in 2013, and it was therefore required to publish its financial statements on a regular basis. 8. The indicators energy sales/total OPEX, energy sales/total costs, and accounts receiv- able are not applicable to generation utilities based on how these indicators were categorized for the purpose of the MENA Electricity Database. However, for com- parative purposes, their values are presented and discussed in this chapter, but not in previous chapters of this book. 9. See Energie Electrique de Tahaddart website, http://eet.ma/decouvrir-notre-activite. 10. See EET’s website (http://www.eet.ma). The 384 MW combined-cycle gas turbine power plant in Tahaddart is owned by EET. Shareholders include the Moroccan ONEE, the Spanish Endesa, and Siemens. 11. While JLEC has a heat rate of 2,195 kilocalories per kilowatt-hour, the values for the Jerada and Mohammedia coal plants—both pertaining to ONEE—are 3,850 and 2,534, respectively. 12. See the Intercontinental Exchange Futures Database, https://www.quandl.com/data/ ICE/ATWK2013-Rotterdam-Coal-Futures-May-2013-ATWK2013. 13. It was estimated that electricity-related employees made up about one-third of all employees in Moroccan utilities (which carry out water, sanitation, and electricity activities). This was based upon actual figures for certain utilities (for example, RADEEMA, which has 350, 302, and 272 employees for a total of 921). 14. A yearly report on public distribution utilities prepared by the Ministry of Interior provides part of the information collected in this chapter. References Amegroud, T. 2015. “Morocco’s Power Sector Transition: Achievements and Potential.” Paper produced in the IAI-OCP Policy Center partnership, Istituto Affari Internazionali (IAI). http://www.iai.it/sites/default/files​/­iaiwp1505.pdf. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Benefits and Challenges of Multiservice Providers: The Case of Morocco 143 Kharbat, F. 2014. “Interconnection between Arab Countries.” Presentation at the “Enabling Renewable Energy in the Electricity Systems” workshop, April 16, Tunis. Arab Union of Electricity. http://www.medelec.org/media/1066/khabat-2014-04-16.pdf. LYDEC (Lyonnaise des Eaux de Casablanca). 2013. Annual Activity Report, 2013. Morocco: LYDEC. ———. 2014. Annual Activity Report, 2014. Morocco: LYDEC. Ministère de l’Intérieur. 2014. Direction des Régies et des Services Concédés—Activités. Morocco: Ministère de l’Intérieur. ONEE (Office National de l’Electricité et de l’Eau Potable). 2013. Annual Activity Report 2013. Morocco: ONEE. ———. 2014. Annual Activity Report 2014. Morocco: ONEE. TAQA (Abu Dhabi National Energy Company, PJSC). 2013. Annual Activity Reports 2013. Abu Dhabi: TAQA. ———. 2014. Annual Activity Reports 2014. Abu Dhabi: TAQA. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 9 A Remarkably Sophisticated Power Market: The Case of Oman In Oman, before 2004, the Ministry of Housing, Electricity, and Water (MHEW) was the sole body responsible for the purchase, transmission, distribution, and supply of electricity on the country’s main interconnected system (MIS) and on its rural systems. Then in 2004, the Law for the Regulation and Privatization of the Electricity and Related Water Sector (the Sector Law), promulgated by Royal Decree 78.2004, significantly altered the way the country’s electricity sector was organized, managed, and regulated. The electricity functions of the MHEW were transferred to nine newly established government-owned successor companies. The sector was thus vertically and horizontally unbundled. All nine of these successor companies remain government owned except for Al-Rusail Power Company, which was privatized in 2007. The Electricity Holding Company SAOC (EHC), which was established at the same time as the successor companies and is 100 percent owned by the Ministry of Finance, held 99.99 percent of shares of the successor companies, while the Ministry of Finance directly held 0.01 percent of the shares. Oman’s power systems are not fully interconnected. Oman’s MIS covers the northern part of the country. A smaller system known as the Dhofar Power Company (DPC) serves the Salalah region in the south. There is also a dedicated small system owned by Petroleum Development Oman (PDO)—the country’s most important state-owned oil producer—with a capacity of 1,500 megawatts (MW). The transmission grids of MIS and DPC are both connected to the PDO system with very limited transfer capacity. Other areas are served by the Rural Areas Electricity Company (RAECO). Since its establishment in 2005, the Authority for Electricity Regulation (AER) has undertaken numerous important projects to support the electricity sector in Oman. These have included reviewing and approving electricity and water-related bulk supply tariffs, as well as reporting on major developments in the electricity and water sectors.1 In 2013, the electricity sector in Oman was largely unbundled except for the rural system where RAECO is still a vertically integrated utility (VIU) responsible Shedding Light on Electricity Utilities in the Middle East and North Africa   145   http://dx.doi.org/10.1596/978-1-4648-1182-1 146 A Remarkably Sophisticated Power Market: The Case of Oman Figure 9.1  Electricity Sector Organization, Oman Authority of Ministry of Housing, Electricity and Electricity Holding Company SAOCO (EHC) Electricity Regulation Water (MHEW) (AER) Public generation companies Oman Distribution companies Oman Rural Areas • AI-Ghubra Power & Desalination Electricity • Mazoon Electricity Power and Electricity Plant Transmission Company (SAOC) Water Company • Al Batinah Power Company Company • Majan Electricity Procurement (SAOC) • AI Suwadi Power Company (SAOC) Company (SAOC) Company • Wadi AI-Jizzi Power Company • Muscat Electricity Distribution Company (SAOC) Private generation companies • ACWA Power Barka SAOG • Barka Power and Desalination Plant • AI-Kamil Power Plant • AI-Rusail Power Plant • Sohar Power Plant • Phoenix Power Company (SAOG) • Sembcorp Salalah Power Company • United Power Company Source: World Bank. for the generation, transmission, distribution, and supply of electricity to custom- ers in its (mostly rural) concession area. Figure 9.1 shows the structure of the electricity sector in Oman in 2013. The Oman Power and Water Procurement Company (OPWPC) is the single buyer of capacity and output from licensed production facilities and other enti- ties, whereas the Oman Electricity Transmission Company (OETC) is a monop- oly provider of transmission services to the MIS. The Rusail Power Company, Wadi Al Jizzi Power Company (WAJPCO), and Al Ghubrah Power and Desalination Company (GPDCO) are electricity genera- tors, while Mazoon Electricity Distribution Company (MZEC), Majan Electricity Company (MJEC), and Muscat Electricity Distribution Company (MEDC) each have monopoly rights to distribute and supply electricity within authorized areas stipulated in their respective licenses. Electricity Generation Electricity generation in Oman is covered by 12 generation utilities (GUs), two VIUs, and three distribution utilities (DUs). All the electricity produced is pur- chased by OPWPC. At the end of 2014, total installed capacity in Oman was 8,143 MW (table 9.1).2 Around 88 percent of this capacity (7,191 MW) was on the MIS, 9 percent (718 MW) was on the Dhofar Power Company system, and the remaining 3 ­percent was on the rural power system. Ninety-eight percent of the fuel used to produce power (and desalinate water) for the MIS is natural gas supplied by the Ministry of Oil and Gas (the other 2 percent is oil). Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Remarkably Sophisticated Power Market: The Case of Oman 147 Table 9.1  Generation Mix, Oman, 2013 Generation Type (MW) Amount Thermal power generation 8,143 Total installed capacity 8,143 Total generated energy (GWh) 28,343 Source: AER 2014. Note: GWh = gigawatt-hours; MW = megawatts. Electricity Transmission Transmission of electricity in Oman is ensured by OETC, which was established as a monopoly provider of transmission services to the MIS. In 2013, 2.7 percent of total net energy generated and the energy purchased was lost during transmission. Table 9.2 includes basic data for the transmission lines and substations of the transmission utility (TU). Table 9.2  Electricity Transmission Data, Oman, 2013 Transmission Amount Total transmission lines and cables (230 kV, 150 kV, 132 kV) km 4,405 High voltage substation capacity (MVA) 20,700 Source: AUE 2013. Note: km = kilometers; kV = kilovolts; MVA = megavolt ampere. Electricity Distribution Three DUs supply the MIS, namely MEDC, which in 2014 had the largest number of customers (261,480); MZEC (318,182 customers); and MJEC (174,592 custom- ers) (AER 2014: 72). Table 9.3 includes some information on the distribution net- work, and the total number of customers in Oman. There are significant differences between the DUs in the ­ volumes of elec- tricity supplied to different customers. Although more than 60 percent of the electricity MZEC supplies goes to residential customers, only 36.6 percent of MJEC’s and 48 percent of MEDC’s electricity supply goes to residential cus- tomers. In turn, more than 40 percent of the electricity that MJEC supplies goes to industrial customers, whereas only 6.3 percent of MEDC’s and only 1.6 percent of MZEC’s electricity supply goes to industrial customers (AER 2014, 73). Figure 9.2 provides insights on the heterogeneity of the utilities’ client base. It reports the percentage of energy sold in 2014 by customer type and use. The residential sector accounts for 75 percent of the customer base and also buys the largest share (at 35 percent of all energy distributed). The public sector accounts for 4 percent of the customer base, yet consumes 16 percent of the electricity distributed. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 148 A Remarkably Sophisticated Power Market: The Case of Oman Table 9.3  Electricity Distribution Data, Oman, 2013 Distribution Amount Distribution lines length (km) 40,552 Distribution substation capacity (MVA) 10,299 Consumers (millions) 0.81 Sources: AUE 2013; OPWPC 2014. Note: km = kilometers; MVA = megavolt ampere. Figure 9.2  Share of Energy Distributed, by Consumer Sector, Oman, 2013 Percent Government and public administration, 16 Public lighting, 3 Residential, 35 Industrial, 20 Agiculture, 2 Commercial, 24 Source: MENA Electricity Sector Assesment Report. Comparison of Generation Utilities in Oman Oman’s unbundled market allows the comparative assessment of generation and distribution. The partial indicators reported here are only a first step toward a full diagnostic since they need to be considered collectively and corrected for the specific supply and demand conditions faced by each individual utility. With this limitation in mind, these partial indicators compared the performance of the two utility types against one another, as well as with the Middle East and North Africa (MENA) median values. As seen from the general indicators in table 9.4, data gaps are significant for the utilities in Oman. The comparison is therefore limited to utilities for which data were available for each indicator. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Table 9.4  Comparing the Performance of Oman’s Generation Utilities across Indicators and against the MENA Median, 20133 MENA Category Indicator name Unit APBS ABPC ASPC GPDCO AKPP ARPP BPDP PPCa SSPWC SPP UPC WAJPCO median General Installed capacity MW 456 744 744 499 285 687b 672 200 489b 585 270 245 — Net generation GWh 2,372 2,655 2,083 2,514 1,672 3,459 3,029b — 1,860b 3,538 1,194 474 — Employment Employees 74 65 50 273 — 16 — — 80 76 48 — — Fuel mix Gas Gas Gas Gas Gas Gas Gas Gas Gas Gas Gas Gas — OPEX $ millions — 44.0 41.0 83.0 37.0 80.0 33.0 0.3 49.0 72.0 14.0 19.0 — Technical and Capacity factor % 59 41 32 58 67 — — — — 69 51 — 58 operational Availability factor % 93 96 90 85 89 — — — — 93 91 — 93 OPEX/employee $ thousands — 676 816 304 — — — — 615 942 289 — 297 Financial (Cost Share of cost of fuel, structure) lubricant in total OPEX % — 59 61 75 78 77 52 — 51 68 — 47 75 Share of labor cost in total OPEX % — — — 13 — — — — — — — 25 12 Financial Energy sales/total OPEX % — 256 276 127 146 110 314 — 246 173 214 108 109 (Cost- Energy sales/total costs % recovery) — 135 136 107 120 — 123 — 86 112 108 97 107 Financial Accounts receivableb Days 36 18 19 46 24 5 33 — 44 30 54 58 40 (Balance Debt/equity % 249 303 294 — 94 — 1,857 — 357 1,399 72 — 357 sheet) Current assets/current liabilities % 121 54 53 443 79 156 42 — 179 118 38 504 95 Financial Return on assets % 8 — — 1 9 — 3 0.1 3 3 5 8 3 (Profitability) Return on equity % 24.0 — — 0.2 15.0 — — — 13.0 — 7.0 2.0 7.0 Source: World Bank calculations. Note: ABPC=Al Batinah Power Company; AKPP = Al-Kamil Power Plant; APBS = ACWA Power Barka; ARPP = Al-Rusail Power Plant; ASPC = Al Suwadi Power Company; BPDP = Barka Power and Desalination Plant; GPDCO = Al-Ghubra Power and Desalination Company; GWh = gigawatt-hours; MENA = Middle East and North Africa; MW = megawatts; OPEX = operating expenses; PPC= Phoenix Power Company; SPP = Sohar Power Plant; SSPWC = Sembcorp Slalah Power and Water Company; UPC = United Power Company; WAJPCO = Wadi Al-Jizzi Power Company; — = not available. a. Commercial operation started on December 11, 2014. b. Values obtained from annual reports and not present in the MENA Electricity Database. 149 150 A Remarkably Sophisticated Power Market: The Case of Oman All the GUs for which data were available have installed capacities above 200 MW. All utilities in Oman are either small or medium (except for Phoenix Power Company), and five of the GUs are publicly owned whereas the rest are private. Several generation plants in Oman are also involved in desalination and other water sector activities. In terms of electricity generated, the lowest value recorded was for WAJPCO (474 gigawatt-hours, GWh), whereas the highest, shown in table 9.4, was for Sohar Power Plant (SPP) (3,538 GWh). The number of employ- ees in Oman’s GUs remains low. The utilities are thermal power plants, using natural gas. The capacity factors vary between 32 percent and 69 percent. These values are comparable to the normal range observed for thermal units, although they are on the lower bound. Yet some utilities in Oman perform better than the MENA median of 58 percent, such as SPP, with the highest capacity factor among the observations in table 9.4. GUs’ availability factor was relatively high, ranging from 85 percent (GPDCO) to 96 percent (ABPC). This is because all electricity generation plants on the MIS use gas turbines to generate electricity. The availability fac- tor—that is, the ratio of the in-service time period to the total year—depends on generation outages, whether due to failure or maintenance. It also depends on the availability of fuel, yet doesn’t indicate whether the units are working at full or partial capacity. The differences between certain indicator values can be explained by the fact that some generation plants are also involved in desalination activities. Most of these plants are private entities that have small staff numbers compared to other utilities in the region. Taking into consideration the generation technology, fuel types, staffing characteristics, and general site layout can help clarify the differ- ences between the GUs. For example, some plants in Oman are gas turbine units used for peaking only. This is characteristic of a high-income country in the MENA region. Operating expenses (OPEX) per number of employees is higher among GUs in Oman when compared to the MENA median. This is mainly driven by the low employment levels in Oman, as shown in table 9.4. The cost of gas accounts for a significant part of the OPEX, representing as much as 78 percent (Al-Kamil Power Plant [AKPP]). In general, the share of fuel cost in total OPEX of utilities in Oman is similar to, or lower than, the MENA median of 75 percent. All utilities recover their total OPEX from sales of energy. The Barka Power and Desalination Plant (BPDP) recovers total OPEX at 314 percent, which is the highest value among the utilities. Performance for this indicator is generally much higher among GUs in Oman than the MENA median of 109 percent. This could be a result of the attractive purchasing prices charged by the GUs, as per their power purchase agreements (PPAs) with the single buyer. A similar trend is observed for the recovery of total costs from the sales of energy (except for the Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Remarkably Sophisticated Power Market: The Case of Oman 151 Sembcorp Salalah Power and Water Company [SSPWC], which has a cost- recovery of 86 percent). With regard to the time lapse between accounts receivable and sales, values range from as low as 18 days (ABPC) to as high as 58 days (WAJPCO), which is not far from the MENA median of 40 days. These values allow the utilities to have a constant flow of cash at hand. The debt-to-equity ratio varies from 94 percent as observed for AKPP to 1,857 percent for BPDP. A debt-to-equity ratio that is too low could imply that the utility has a large amount of cash on hand and is, therefore, not necessarily managing its equity in the most efficient way, whereas a very high ratio would imply that the utility has a high level of debt and depends on debt to finance its projects and operations. Ideally, a ratio value oscillating around 100 percent would show that the utility is capable of managing its equity while at the same time using debt as a strategic financing tool. For half the utilities, the ratio of current assets to current liabilities was 100 percent or more. In the other half, ABPC, the Al Suwadi Power Company (ASPC), AKPP, BPDP, and the United Power Company (UPC) had values in the range of 38 percent to 79 percent. For these utilities, current liabilities are not liquid. The return on assets (ROA) for generators in 2013 ranged from 0.1 percent (Phoenix Power Company [PPC]) to 9 percent (AKPP), whereas the return on equity (ROE) ranged from 0.2 percent (GPDCO) to 24 percent (ACWA Power Barka [APBS]). The ROA for PPC, as shown in table 9.4, is low mainly because it only became operational in 2014. The profitability of generators in Oman is derived from their availability and reliability. Changes in the demand and supply landscape do not affect profits because a pass-through cost exists for the GUs. Comparison of Distribution Utilities in Oman Table 9.5 compares the three DUs MJEC, MZEC, and MEDC. The MENA median values are also included for comparison. The load factor of the three DUs ranged from 44 percent to 71 percent, largely reflecting differences in their customers base. Commercial and industrial customers tend to have a higher load factor than residential customers. Because commercial and industrial customers account for 56 percent of MJEC’s total supply, compared to 36 percent in the MIS overall, MJEC has a relatively higher load factor. Distribution losses, on the other hand, are relatively high among the Omani DUs, ranging from 9 percent to 13 percent. OPEX per employee is of the same order of magnitude for MZEC and MEDC ($226,672 and $174,580, respectively). Values of this indicator are higher than the MENA median ($188,000). Labor costs’ share of OPEX in Omani DUs remains low (5 percent for MEDC and 6 percent for MJEC), show- ing that the OPEX is most probably made up of other costs such as electricity purchase costs. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 152 A Remarkably Sophisticated Power Market: The Case of Oman Table 9.5  Comparing the Performance of Oman’s Distributors across Indicators and against the MENA Median, 2013 Category Indicator name Unit MJEC MZEC MEDC MENA Median Technical and operational Load factor % 71 44 55 60 Distribution losses % 13 11 9 10 OPEX/employee $ thousands/ employee 227 175 — 188 OPEX/connection $/connection — 1,150 1,698 346 OPEX/kWh sold $/kWh 0.05 — — 0.1 OPEX/km $, thousands/km — 14 42 19.6 Financial (Cost structure) Share of labor cost % in total OPEX 6 — 5 12 Financial (Cost-recovery) Energy sales/OPEX % 69 61 80 93 Financial (Balance sheet) Accounts receivable Days 119 110 122 121 Debt/equity % 109 148 147 523 Collection ratea % 79 77 74 93 Current assets/ % current liabilities 43 18 46 85 Financial (Profitability) Return on assets % 8 6 8 3 Return on equity % 14 14 16 7 Source: World Bank calculations. Note: km = kilometers; kWh = kilowatt-hours; MEDC = Muscat Electricity Distribution Company; MJEC = Majan Electricity Company; MZEC = Mazoon Electricity Distribution Company; MENA = Middle East and North Africa; OPEX = operating expenses; — = not available. a. Values obtained from calculations in appendix C and not present in the MENA Electricity Database. Accounts are received within 119 days for MJEC, 110 days for MZEC, and 122 days for MEDC. With regard to cost-recovery performance indicators, values of energy sales to total OPEX were available for all three DUs, with MEDC having the highest total OPEX recovery from sales, at 80 percent. However, no ­ values were available for total cost recovery from sales of electricity and for total billing per connection. Current assets to current liabilities ratio for the three utilities were 43 percent, 18 percent, and 46 percent, respectively. MZEC has the lowest liquidity ratio. All three DUs have low values, which suggests that they are unable to repay their current liabilities (as could be the case with the lower value observed for MZEC) or that they are managing their current assets in a strategic manner (as could be the case for MEDC, because the ratio is closer to 50 percent). Finally, in terms of profitability all three DUs showed positive results. MEDC and MJEC have similar ROA—8 percent each. MEDC also has the highest ROE, at 16 percent, followed by MJEC and MZEC, at 14 percent each. Evolution of Oman’s Electricity Sector since 2014 The electricity sector in Oman has gone through a number of changes since 2014 that are worth mentioning, given that the analysis of this chapter is based on 2013 data. The system peak demand in MIS grew by 9 percent annually between 2009 and 2014, reaching 5,122 MW in 2014, and was forecast to grow at about Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Remarkably Sophisticated Power Market: The Case of Oman 153 the same rate to 9,530 MW in 2021. Energy requirements were expected to grow from 25 terawatt-hours (TWh) to 47.1 TWh during the seven-year period 2015–21. Fuel use was expected to increase by 4 percent per year. Natural gas consumption as fuel was expected to increase by 4 percent a year. There are planned projects to interconnect the MIS and DPC grids and to extend the grid to some of the rural areas gradually. As Oman is dependent on hydrocarbon exports, which account for around two-thirds of total export earnings, it remains vulnerable to fluctuations in oil prices. Oil and gas revenue fell by 40 percent in 2015 due to lower oil prices, despite higher output (natural gas production rose by 5 percent in 2015). In particular, significant energy subsidy reforms have taken place over the past few years. Since January 2015, Oman doubled gas tariffs for industrial producers and the power industry to $3.0/million British thermal units. In the absence of a pass-through, the subsidy level was expected to increase 46 percent in the MIS, which triggered AER’s decision to increase electricity tariffs for commercial and industrial users at the end of 2016. Taking into account the likely international prices of liquefied natural gas (LNG) for Oman’s market (the Asia-Pacific) and traded diesel prices, a recent International Renewable Energy Agency report calcu- lated that the total volume of subsidies for power consumption in 2012 would amount to some $2.63 billion compared to the country’s LNG export receipts of about $4 billion. The electricity sector has been restructured and regulatory reforms have been successfully implemented along with the transparent calcula- tion of subsidies by AER. Much needs to be done in restructuring retail electricity tariffs and adjusting them upward. In 2014, the government subsidy was at 38 ­ percent of the economic cost. The subsidies in the much smaller systems of DPC and RAECO in 2014 were 44 percent and 78 percent, respectively, of their economic costs. The government is working to gradually adjust the electricity tar- iffs to phase out the subsidy, but the impact mitigation mechanism is yet to be determined. More recently, effective January 2017, electricity tariffs were also increased for commercial and industrial customers. Oman is currently developing and implementing a competitive wholesale market for electricity that will provide a route to market for generators and the creation of an electricity spot market. Generators not under a PPA contract will need to adapt to operating in such an environment. Demand-side response is expected to play an increasing role in the mid to long term in Oman. By changing the profile of demand, and increasing its flexi- bility, the demand-side response can reduce the need for investment in genera- tion and network capacity. Within this context, the new tariff will reflect the actual cost of supplying electricity and would provide a relatively small number of customers with strong incentives to reduce demand at peak times. This prom- ises significant benefits in terms of reducing overall peak demand and the requirement for future investments. Finally, stand-alone reverse osmosis will play a significant part in water pro- duction in the future. This has the potential to reduce the linkage between water and power production. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 154 A Remarkably Sophisticated Power Market: The Case of Oman Conclusion This chapter analyzed the performance of Omani electricity utilities in 2013. Oman’s unbundling of the electricity sector has set the country on a path to overall reform in the power sector. The main challenges observed across utilities in Oman are related to covering their operating costs, which makes them heavily dependent upon state subsidies. Because revenues from the sale of electricity do not cover the total economic cost of supply, the Ministry of Finance provides an electricity subsidy to licensed suppliers on an annual basis. On the demand side, growing income per capita, continued government investment in infrastructure projects, and a growing population are all expected to contribute to a continued high growth in electricity demand in the sultanate. Of the 12 GUs studied in this chapter, 8 are private (1 big, 4 small, and 3 medium) and 4 are public (2 small and 2 medium). With a fuel mix exclu- sively based upon natural gas, the GUs in Oman use some of the most effi- cient technologies in the region for fossil-fuel-based electricity generation. Fuel costs’ share of total OPEX is close to the MENA median of 75 percent. The number of employees in GUs remains very low in comparison to other MENA economies, which results in high OPEX per employee (from $289,000 to $942,000). All utilities recover their OPEX from the sales of energy, which could be a result of OPWPC purchasing all the electricity from the GUs based upon PPAs. OPEX recovery values range from one to three times the MENA median. This is also reflected in the profitability indicators, which show positive performance for ROA and ROE across all the GUs for which data were available. All three DUs in Oman are public: two medium and one small (MEDC). In terms of technical performance, there is still room for improvement, particularly for MJEC and MZEC, which have distribution losses higher than the MENA median value of 10 percent. The three utilities seem to perform well financially, with ROA and ROE values more than twice the MENA median in most cases. This seems to contradict the low performance reported on indicators such as accounts receivable, collection rate, and recovery of OPEX from sales, which would be expected to indicate poor overall financial performance. This could be explained by the fact that DUs in Oman benefit from transfers from other sources, such as the government, to help maintain their positive financial performance. Unbundling and private sector involvement have contributed to the overall improvement of utility performance on several levels, though more so in genera- tion activities than in distribution-related activities. Any analysis of Oman’s power sector must consider the role that many electricity utilities play in desali- nation and water-related activities. On the supply side, the country still depends exclusively on gas and to a much lesser degree on diesel as sources of energy (about 2 percent of the energy mix), whereas the development of renewable energy—in spite of the abundant solar potential—has yet to be considered. However, the presence of subsidies, as in most parts of the MENA region, does not make renewable energy an attractive economic alternative for the generation Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 A Remarkably Sophisticated Power Market: The Case of Oman 155 of electricity, for which a review of tariffs and subsidies would need to take place. Not only would this encourage renewable energy development, but also it would allow for the country’s energy sector to cope with increasing demand while mini- mizing the risks and impacts of higher fiscal burdens. Last but not least come the issues of data collection and data quality. Across power sectors in MENA, Oman’s may have the most transparent reporting, and the AER collaborated closely with us on this study. In terms of data quality, all DUs and the TU report having implemented supervisory control and data acquisition (SCADA), and a majority of utilities have adopted international accounting standards (IAS). As of 2013, more than half the GUs, all the DUs, and the two VIUs had yet to implement cost-accounting systems. Notes 1. In addition, the Public Authority for Electricity and Water (PAEW) is the regulator for the water sector in Oman. Created in 2007, its role in the electricity sector is limited to policy overview. In addition to this, PAEW is also a direct water service provider. 2. Calculated from data on individual capacity of different plants (AER 2014, 9). 3. The indicators energy sales/total OPEX, energy sales/total costs, and accounts receiv- able are not applicable to generation utilities based on how these indicators were categorized for the purpose of the MENA Electricity Database. However, for com- parative purposes, their values are presented and discussed in this chapter but not in previous chapters of this book. References AER (Authority for Electricity Regulation). 2014. Annual Report 2014. http://www​ .aer-oman.org/pdfs/Annual%20Report%202014%20-%20Eng.pdf. AUE (Arab Union of Electricity). 2013. Annual Statistical Bulletin 2013. Amman, Jordan: AUE. OPWPC (Oman Power and Water Procurement Company). 2013. Annual Report 2013. http://www.omanpwp.com/PDF/OPWP%20Annual%20Report%202013%20eng.pdf. ———. 2014. Annual Report 2014. http://www.omanpwp.com/PDF/01-AR-2014​ -OPWP-Eng%20(12).pdf. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 CHAPTER 10 Synopses of the Case Studies Introduction The four case studies presented (of the Arab Republic of Egypt, Jordan, Morocco, and Oman) offer insights relevant to the Middle East and North Africa (MENA) region and beyond. The studies aimed at providing not only an overview of each country’s power sector but also an analysis of utility performance to help identify potential areas of improvement. This chapter presents the key findings of each case study. Arab Republic of Egypt: An Urgent Need for Sector Reforms The Egypt case study should be put in the context of 2013—the year of the data used for this analysis—recognizing that the Egyptian power sector has gone through some important changes since then. The technical and operational performance of Egypt’s generation utilities (GUs) is consistent with the Middle East and North Africa (MENA) median, but the country’s commercial and financial performance is much worse. Capacity fac- tors range from 58 percent to 70 percent of their full capacity, and availability factors range from 79 percent to 91 percent, which is consistent or better than the MENA median. But operational expenses (OPEX) per employee are lower, and costs are high because of high fuel costs and excessive staffing. Cost-recovery and accounts receivable indicators point to a major dependence on subsidies. Accounts receivable of Egyptian GUs are from 6 to 15 times higher than the MENA median, and none of the GUs in Egypt recover their total OPEX or their total costs from sales, except the Middle Delta Electricity Production Company, which recovers its OPEX, yet not its total costs. This poor commercial performance explains the high fiscal cost of the sec- tor: $1.6 billion in subsidies in 2013. For most of the GUs, the low cost- recov- ery reflects low selling tariffs, combined with the high cost of fuel, as well as in some cases low production and hence low sales. The need to borrow to finance business explains the very high debt-to-equity ratios of Egyptian GUs, between 4 and 10 times higher than the MENA median of 357 percent. It also Shedding Light on Electricity Utilities in the Middle East and North Africa   157   http://dx.doi.org/10.1596/978-1-4648-1182-1 158 Synopses of the Case Studies explains the low current ratio, which is below 70 percent for all GUs. The outcome of the poor commercial and financial management of the GUs is a return on assets (ROA) and a return on equity (ROE) close to 0 percent in most cases, well below the 3 percent and 7 percent median value for ROA and ROE in MENA. Meanwhile, Egypt’s DUs have significant room for improvement on the tech- nical and financial front but do reasonably well on the commercial dimensions of performance—in spite of the complex political and social context in which they need to operate. They also perform relatively well technically compared to their peers, although not on all dimensions. Their load factor and distribution losses are close to the regional MENA median. Their high nontechnical losses (owing to theft and erroneous meter readings) are, however, quite high and explain about 25 percent of total distribution losses. OPEX per employee is much lower than the MENA median, reflecting the largest number of employees in the region. The share of labor costs in total OPEX is two to three times higher than the MENA and non-MENA medians. The high costs and the social context help explain why average tariffs were below costs in 2013 generally, further fueling the subsidy cost of the sector already noted for GUs. Some of Egypt’s DUs get additional revenue. For instance, the Canal Electricity Distribution Company gets subsidies for the electricity exported to Gaza. Egypt’s DUs compensate for the low cost-recovery rates and some of the excess costs resulting from their technical performance with a solid commercial performance in a difficult social context. They manage to enjoy high billing levels and high collection rates, close to or above the regional median. The only significant issue is the high receivable period (for example, almost six months for the North Cairo Electricity Distribution Company). This can be attributed to delayed collec- tion cycles resulting from time-consuming manual registration of readings and billing. Since 2013, the Egyptian Electricity Holding Company has been exploring the option of shifting to smart meters as a potential solution to reducing nontechni- cal losses and the time taken for bill collection. Low cost-recovery is compensated not only by subsidies but also by borrowing. This explains the high debt-to-equity ratio and the low ratio of current assets to current liabilities. And this also explains the low ROE and ROA, except for Canal Electricity Distribution Company and South Cairo Electricity Distribution Company, which benefit from additional rev- enue sources. Delaying important investment and management decisions because a country is going through difficult political and social times may simply lead to more political and social tension. While Egypt’s reforms have allowed its utilities to achieve a technical and operational performance largely consistent with or often better than MENA’s median performance, they have not been able to attract the investment needed to meet growing demand. Moreover, the sector has relied extensively on subsidies to finance its operational expenditures, a practice that is unlikely to be sustainable as the country adjusts its fiscal balance. Delaying investment decisions further could exacerbate political tensions by increasing the risk of consumer rationing. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Synopses of the Case Studies 159 Simply adding more installed capacity is not the only way to address grow- ing demand. The country has been slow to cut costs and improve its commer- cial performance. Improving labor efficiency and cost-recovery should be on the agenda eventually, even if doing both jointly may be too difficult amid current tensions. Additional options include a redesign of tariff regulations to increase the scope for cross-subsidies and improve the social targeting of elec- tricity pricing. Jordan: Harvesting Results from a Restructuring of the Power Sector The Jordan case study should be put in the context of 2013—the year of the data used for this analysis—recognizing that the Jordanian power sector has gone through some important changes since then. Overall, Jordan’s GUs do quite well on most performance dimensions for which data are available, with cost-recovery rates a notable outlier. They do not do very well on monitoring and transparency, because there are significant data gaps for some of the utilities. But the data are solid enough to be able to provide a strong diagnostic. At the technical and operational levels, the king- dom’s GUs stand out by the high dispersion of their performance. The capac- ity factor is well below the MENA median for four of them, including two independent power producers (IPPs) that were not fully in operation in 2013, and well above for two of them (both are private—Amman East Power Plant and Qatrana Electric Power Company—with strong contractual service obligations). The availability factor is only obtainable for half the utilities but is consistent or better than the MENA median. At the operational level, Jordan’s DUs are overstaffed, although the IPPs less so. Because Jordan is one of the few countries in the region that does not provide fuel subsidies, OPEX values1 are much higher than values observed elsewhere in the region. This results in high shares of fuel costs for the GUs, ranging from 95 percent to 98 percent. Consequently, the share of labor costs in OPEX is considerably smaller than the MENA median, and OPEX per employee is much larger than the MENA median. At the commercial level, only one utility recovered its total OPEX or total costs. The differences for the rest are essentially covered by subsidies. At the finan- cial level, Jordan’s GUs align with MENA values. The debt-to-equity ratio is simi- lar across GUs and in some cases lower than the MENA median, with the exception of the fully state-owned utility Samra Electric Power Generating Company (SEPCO) at 876 percent. SEPCO has been expanding its generation capacity since 2010 by adding new generating units, of which most have been financed mainly through debt rather than equity. The GUs are all profitable, with high ROE and ROA values—in particular the IPPs—reaching at least three times the median MENA ROE of 5 percent. This is mainly a result of the provisions of the IPP contracts under which they operate, which ensure their profits. Jordan’s DUs are profitable. Compared with their MENA peers, they perform better on billing and cost-recovery but underperform on most other dimensions. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 160 Synopses of the Case Studies At the technical level, Jordan’s three DUs have lower load factors and somewhat higher losses than the median MENA values. At the operational level, OPEX per employee figures for Jordanian utilities are much higher than the MENA and non-MENA medians. This is most likely due to the high OPEX figures attributed to other costs, because share of labor costs represents only about 6 percent or 7 percent of OPEX in each of the utilities. Differences in operational performance are influenced by the differences in the geographical areas covered by the utilities. Total billing per connection is much higher than the MENA median. Differences across DUs in Jordan are explained by differences in the customer base. For instance, Jordan Electric Power Company benefits from higher con- sumption in the capital, Amman, and among its industrial consumers. The other utilities have more rural clients. Despite their good billing performance, only the Irbid District Electricity Company (IDECO) recovers its OPEX from sales. As in other countries of the region, all Jordanian DUs have high debt-to-equity ratios, suggesting a high level of financing through debt. Considering Jordan’s DUs’ underperformance on various indicators, it may be surprising that the bottom line is so positive and that Jordan’s DUs enjoy high ROEs and ROAs relative to the regional medians. But this is partially linked to the fact that the licenses granted to the two privatized utilities, Electricity Distribution Company and IDECO, guarantee a 10 percent profit on their regulatory asset base after the regulator reviews and approves their annual budgets, their projects, and the anticipated electricity losses. The power sector’s main difficulty in 2013 may have been its excessive reli- ance on explicit and implicit subsidies as indicated by its high quasi-fiscal deficit. Because much of this is related to underpricing, it is important to get the pricing signals right. This helps managing demand. The main difficulty is that this would have to be done in a politically sensitive context: the poor, including many refu- gees, will have to be protected as much as possible from brutal price shocks while efforts to cut the sector’s fiscal costs are implemented. Average tariffs can help ensure financial and fiscal viability. But tariff composition matters just as much to equity as the social, financial, and political viability of energy prices. Finally, to do the job right, a regulator critically needs information. Morocco: Benefits and Challenges of Multiservice Providers The case study of Morocco should be put in the context of 2013—the year of the data used for this analysis—recognizing that the Moroccan power sector has gone through some important changes since then. From a technical perspective, Morocco’s biggest private GU (Jorf Lasfar Energy Company [JLEC]) does much better than its MENA peers, with high capacity and availability factors. This is probably because in its power purchase agreement (PPA), electricity purchase is guaranteed and the utility is encouraged to maximize utilization of its generation capacity. The large multi-utility com- pany, the Office National de l’Electricité et de l’Eau Potable (ONEE), is below Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Synopses of the Case Studies 161 par, with a capacity factor of 31 percent, suggesting that ONEE’s facilities are operated as load followers and for peaking. The availability factor of ONEE’s generation facilities, at 53 percent, can mainly be explained by temporary repair and maintenance issues. At the operational level, JLEC’s cost performance is strong and the share of energy purchases and cost of fuel, lubricant, gas, and coal in total OPEX are high (94.5 percent). The sector’s commercial performance is excellent by any standard as indicated by cost-recovery and receivables. The cost-recovery rate in the case of JLEC is comfortably high (153 percent for recovery of OPEX from sales), as would be expected from an IPP. This level reflects the strong profitability of the business. Morocco’s private power producers enjoy attractive contractual arrangements, and PPAs are designed to pass most market and institutional risks to ONEE. The cost-recovery performance of ONEE’s generating activity is difficult to assess without detailed analysis of the overall costs of the vertically integrated utility (VIU), its sales, and the extensive use of cross-subsidies. The only hard evidence is that ONEE’s electricity sales alone were not sufficient in 2013 to fully recover total costs. At 45 days, JLEC’s receivables are almost the same as MENA’s 40 days median. However, ONEE is quite high at 159 days. The financial management is quite reasonable by MENA standards for the private GUs but unsustainable for ONEE. The debt-to-equity ratio of JLEC is 277 percent, which is relatively high but smaller than for many GUs in the region. It is much worse for ONEE (1,240 percent). At 17 percent, JLEC’s ROE was high, while its ROA was 4 percent. ONEE was not profitable in 2013, with an ROA valued at −4.40 percent. Morocco is working on address- ing the issue, but it is not an easy challenge because a large proportion of risks are passed to ONEE (for example, fuel price and exchange rate) to protect private GUs. An assessment of DUs in Morocco is particularly challenging because they tend to deliver both electricity, water and sanitation services. In addition, all the small DUs are public (municipal distributors) except for AMENDIS Tetouan, which is private, whereas three of the five medium utilities are private. Accounting for this limitation, several characteristics can be sketched out. Technically, the DUs’ load factors are all close to the MENA median value of 60 percent. Distribution losses in Morocco are lower than or equal to the MENA median of 10 percent. Operationally, the DUs are not doing as well as their peers. While labor costs of all DUs make up 8 percent to 14 percent of OPEX, the OPEX per employee values are generally higher than the MENA median. This assessment is, however, biased by the fact that key information is not available for the private operators. The share of labor costs in total OPEX remains low for the utilities for which values were reported, ranging from 8 percent to 14 percent. This suggests that the OPEX is mainly made up of other costs, such as the costs of purchasing electricity from ONEE. OPEX per connection, per kilowatt-hour (kWh), and per kilometer (km) are almost all higher than the MENA medians. The measures are, however, imperfect because these companies’ coverage of both electricity and water blurs Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 162 Synopses of the Case Studies key information on cost allocation. Moreover, electricity services are often used to cross-subsidize the water services and the heavy investments required in sanita- tion-related activities. Hence, tariffs and prices are often unrelated to sector-spe- cific costs. They can include fees used by the operators to compensate for low national tariffs. From a commercial perspective, Morocco’s utilities do reasonably well with OPEX because for most of the DUs they are fully or almost recovered. Full cost- recovery is not the norm, but the country is working on closing the gap. Morocco has indeed worked on tariff adjustments designed to improve cost-­ recovery, while not affecting households with monthly electricity consumptions below 100 kWh. This offers a model for many other countries in the region. From a financial perspective, all private utilities show strong profitability ratios with the exception of AMENDIS Tetouan, which shows negative ROA and ROE. This can be explained by the fact that AMENDIS Tangier and Tetouan are a joint concession, and the private operator compensates Tetouan with the business in Tangier. Data availability was the main obstacle to analyzing the financial perfor- mance of the municipal distributors. Morocco’s experience adds to the evidence provided by the other countries on a limited availability of access to performance indicators. Very little informa- tion on Morocco’s electricity utilities is publicly available. The issue of data qual- ity is also of concern, particularly for public DUs. Oman: A Remarkably Sophisticated Power Market The case study of Oman should be put in the context of 2013—the year of the data used for this analysis—recognizing that the Omani power sector has gone through some important changes since then. At the technical and operational levels, performance varies across Oman’s GUs. The capacity factors vary between 32 percent and 69 percent in the sample. Availability factors are notably high, ranging from 85 percent to 96 percent. This is because all the generating plants on the main interconnected system use gas turbines to generate electricity. Some of the differences in performance can be explained by the fact that a portion of the generation plants are also involved in desalination activities and that a few are used only for peaking. Most are private entities, which may explain why they have relatively small staff numbers and higher OPEX per employee than the region’s median. The cost of gas accounts for a significant share of OPEX, representing as much as 78 percent. In general, fuel costs’ share of total OPEX is similar or lower than the MENA median of 75 percent. Oman’s GUs rely on a fuel mix exclusively based on natural gas and use some of the most efficient technologies in the region for fossil-fuel-based elec- tricity generation. At the commercial level, almost all utilities recover their total OPEX from the sales of energy. This good performance is, partially at least, the result of the Oman Power and Water Procurement Company purchasing all the electricity from the GUs based upon PPAs. OPEX recovery values range from one to three times the Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Synopses of the Case Studies 163 MENA median. The strength of the commercial performance is confirmed by the relatively short average time lapse between accounts receivable and sales, guaran- teeing a constant flow of cash on hand. As in many countries of the region, financial performance is an issue. The debt-to-equity ratio of the GUs in Oman varies from 72 percent to 94 percent to 357 percent. For utilities with a small ratio, the issue may be excess cash. For utilities with a high ratio, the explanation is an excessive tendency to borrow to finance projects and operations, even when cash is available. For most of the utilities, the current assets to current liabilities ratio is divided into those utili- ties with a ratio below 100 percent and those above. Al Batinah Power Company the Al Suwadi Power Company, Al-Kamil Power Plant, Barka Power and Desalination Plant, and the United Power Company showed values in the range of 38 percent to 54 percent. For these utilities, current liabilities are very high compared to the current assets and are, therefore, not liquid. The most puzzling indicators concern profitability. In view of the sector’s strong com- mercial performance, it is surprising to see such low ROAs, even if they are quite in line with the MENA median. They range from 0.13 percent to 9 per- cent. The range for ROE is broader: 0.2 percent to 24 percent. The main drivers of this dispersion are availability and reliability. Changes in the demand and supply landscape will not affect profits, because there is a pass-through cost for the generators. The three DUs in Oman are public. Our detailed analysis indicates average technical and operational performance, poor commercial performance, and very good financial performance by MENA standards—despite poor cost-recovery rates. At the technical level, the load factors are low by MENA standards and losses are at the MENA median or worse. The diversity largely reflects differ- ences in the customer base. At the operational level, OPEX per employee is average by MENA standards, although labor costs’ share of OPEX is notably low, at half or less the MENA median. The OPEX per connection is above the MENA median, although there is no clear pattern for OPEX per km. Regarding commercial performance, the data available suggest average performance by MENA standards, and low by global ones. Accounts are received within 110– 122 days, and collection rates are significantly lower than the region’s median, with values ranging from 74 percent to 79 percent. The recovery of OPEX (and hence total costs) ranges from 61 percent to 80 percent, which highlights the importance of other forms of sector financing. In this case, however, debt plays a smaller role than in other countries of the region. With debt-to-equity ratios ranging from 109 percent to 148 percent, Oman’s utilities seem to be doing much better than the MENA median. But this is not confirmed by the ratio of current assets to current liabilities, which ranges from 18 percent to 46 percent. This is low by any standard and hints at a limited ability to repay current liabili- ties from cash or assets. For now, this does not affect the profitability of the DUs, which benefit from public transfers to help maintain their positive finan- cial performance. Muscat Electricity Distribution Company and Majan Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 164 Synopses of the Case Studies Electricity Company have similar ROAs (8 percent each). Muscat also has the highest ROE (16 percent), followed by both Majan and Mazoon Electricity Distribution Company at 14 percent each. This seems to contradict the low performance reported by indicators such as accounts receivable, collection rate, and recovery of OPEX from sales, which should be expected to indicate poor overall financial performance. Ultimately, the review of Oman’s experience provides evidence that the sig- nificant restructuring adopted by the country is technically and institutionally feasible. In 2013, the main doubt was related to the financial sustainability of the model. If the subsidy policy was to be maintained, it seems likely that the expected growth in demand would increase fiscal pressure. Unless this risk is addressed through closer monitoring of financial and commercial performance, the relative and absolute costs of the sector will continue to grow, despite a strong reliance on private operators to finance specific needs. More data are essential to all these efforts. Note 1. The cost of fuel was estimated for the Jordanian generation utilities based upon the aver- age cost of fuel per kWh from the regulator and the kWhs generated by each utility in 2013. This cost of fuel was then added to the operating costs to obtain the total OPEX as per the definition used in this study. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Conclusion The potential management and financing payoffs of the new MENA Electricity Database produced in the context of this analysis are hard to ignore. Because the database covers most power utilities in the Middle East and North Africa (MENA), it is broad enough to produce robust insights into the sector’s achieve- ments and its challenges. Those insights, synthesized in this concluding section, can be turned into con- crete management and policy decisions. But it should be remembered that the baseline year of the study is 2013, and the power sector has changed since then, in some economies more than others. An appropriate response, of course, is to expand and extend the analysis and data collection begun here. And this is what we hope each economy will decide to do working with its utilities. A fundamen- tal lesson of this study is that data analysis is essential to performance diagnostics at the utility level and at the sector level. Cutting Hidden Costs in the Power Sector Is Key to Financing Sorely Needed Investment Explicit and implicit subsidies of MENA’s power sector impose a very heavy burden on taxpayers and power users. The burden can be measured in the utili- ties’ hidden costs, or quasi-fiscal deficits (QFDs), which express the cost of not operating in the manner of a well-run utility. The QFD encompasses four types of inefficiencies: collection losses, transmission and distribution losses, underpric- ing, and overstaffing. Estimates of the power sector’s QFD range between −0.1 percent of gross domestic product (GDP) for the West Bank to 8.9 percent in Lebanon. To put this in context, consider that in Sub-Saharan Africa, where social concerns are at least as large as in the MENA region, the sector’s QFD ranges from −0.3 percent to 6.0 percent. Half of the 14 MENA economies studied have a QFD in excess of 4 percent of the entire economy’s GDP. The QFD share of GDP is relatively small in Maghreb economies and large in some Mashreq and Gulf Cooperation Council economies. The median value of about 4 percent of GDP represents one Shedding Light on Electricity Utilities in the Middle East and North Africa   165   http://dx.doi.org/10.1596/978-1-4648-1182-1 166 Conclusion and a half times the average investment needed in the region’s electricity sector, estimated at about 3 percent of GDP. In other words, the sector’s investment gap could be filled simply by halving the current level of inefficiency. At the utility level, performance varies widely. When measured as a share of utilities’ revenue, QFDs range from 25 percent for a West Bank distribu- tion utility (DU), Northern Electric Distribution Company (NEDCO), to almost 1,300 percent for the vertically integrated Iraqi power ministry. The QFD of at least 13 utilities exceeds their revenue. These figures reveal the extent to which utility-specific inefficiencies common in the region may be preventing self-financing. Underpricing Is the Major Source of Inefficiencies, Although Otherwise Inefficiencies Are Economy and Utility Specific The inefficiencies reflected in the QFD are linked both to policy and manage- ment decisions. The sources of inefficiencies, and hence the nature of the solu- tions, vary across economies. About two-thirds of the QFDs we detected can be traced to tariffs being set below cost-recovery levels in most economies, which nearly always reflects a political decision intended to protect current users. Even under such circumstances, however, managing costs can go far to enhance reve- nues. For example, Jordan’s high levels of cost inefficiency are due largely to electricity production costs that reflect the preponderant role of diesel and fuel oil in generation. The remaining one-third is explained by commercial losses, collection failures, and overstaffing, which are all mostly management decisions, although overstaff- ing may sometimes represent a political decision if it is an issue for all utilities in a given economy. These sources of inefficiencies should not be underestimated, as they represent half of the resources needed for the sector’s investment needs. Overstaffing is of particular concern in only a few utilities, almost all of them DUs in Egypt. Collecting bills seems to be a significant challenge for DUs in Djibouti, Jordan, and the West Bank. Technical losses are significant for two of the West Bank operators (Jerusalem District Electricity Company and NEDCO) and for the Republic of Yemen’s vertically integrated utility (VIU). Low tariffs and overstaffing often reflect good intentions, but they are not the most effective ways to ensure that the poor can afford electricity or to boost employment. Moreover, given their present macroeconomic prospects, many MENA economies cannot afford to continue to lavish on average 2 percent of GDP on poorly targeted electricity subsidies. Improving the sector’s performance will allow economies to increase the social returns on fiscal resources by allocating savings where they will do the most good, whether within the sector or outside it. Identifying and unbundling hidden cost drivers and inefficiencies at the utility level can pinpoint areas for improvement—whether financial, technical, com- mercial, or labor—and, from a regulatory perspective, improve the accountability of key actors. From the perspective of sector policy, quantifying the QFD provides governments with a rough order of magnitude of the improvements. ­ Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Conclusion 167 Taking advantage of readily available opportunities to reduce cost inefficiencies in the generation and distribution of electricity will also make the sector more sustainable and increase the creditworthiness of utilities, thus facilitating access to commercial financing. MENA’s Power Sector Must Match Its Technical Success with Improvements in Commercial and Financial Management For more than half of the indicators selected—most of them technical—the region’s economies tend to perform better than the sample of economies outside MENA. Unfortunately, there does not seem to be a clear correlation between good technical performance and sustainable financial performance, and unless the sector can increase its revenue or better manage its costs, the current techni- cal level is unlikely to be sustainable (table CL.1). On the technical and operational side, the international comparison and the trend analysis point to a significant increase in operating expenses (OPEX) Table CL.1  Comparing Median Utility Performance in the MENA Region and Elsewhere Vertically integrated All utilities Distribution utilities utilities Technical and operational OPEX/connection ($) — MENA higher MENA higher OPEX/kWh sold ($) — MENA lower Samples too small Residential connections/employee — MENA lower MENA lower Distribution losses Equivalent — — Commercial Energy sold (kWh)/connection MENA higher — — Total billing/connection MENA somewhat higher — — Collection rate MENA somewhat lower — — Financial Sales/OPEX (%) — MENA somewhat lower MENA somewhat higher Sales/total costs (%) — MENA higher (depending MENA lower (depending on subsidies) on subsidies) Accounts receivable/sales (days) MENA much higher — — Debt/equity MENA much higher and — — essentially unsustainable Current assets/current liabilities Equivalent but not ideal — — Return on assets (%) MENA somewhat higher but — — not high enough to stimulate financing Return on equity (%) MENA higher but not — — commensurate with risk Source: World Bank calculations. Note: Comparisons are only made for all utilities together when the indicator has the same meaning for different types of utilities. Otherwise, comparisons are made separately for distribution utilities and vertically integrated utilities. kWh = kilowatt-hours; MENA = Middle East and North Africa; OPEX = operating expenses; — = not applicable. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 168 Conclusion during the period covered, which is consistent with the increase in oil prices from 2009 to 2013. On commercial management, the indicators reveal (a) a high dependence on subsidies to recover costs and (b) a high tolerance for nonpay- ment (with a ratio of accounts receivable to sales that is almost three times that of non-MENA economies). On financial dimensions, despite return-on-assets and return-on-equity values that are somewhat better than those of non-MENA peers, the region appears to be relying on a risky strategy as indicated by (a) a low ratio of current assets to current liabilities (lower than 100 percent) and (b) an exceptionally high debt-to-equity ratio (almost four times the non-MENA median), leaving utilities highly exposed to external shocks. The importance of labor costs highlighted by the QFD analysis is likely to be a particularly sensitive topic in any policy discussion of the data reported here. In a region where underemployment is a major problem, it is impossible not to rec- ognize the political sensitivity of efforts to improve labor indicators. Where the mat- ter is so sensitive that overstaffing in the power sector simply cannot be broached, it may nevertheless be useful to quantify the costs of not addressing the issue, thus clarifying the implications for subsidy levels (if revenues cannot be increased). Because partial indicators of utility performance can lead to heterogeneous rankings of utilities, we applied an “average rank score” methodology to help utilities assess their performance against other utilities across a set of relevant indicators. The average rank score makes it possible to identify the better-per- forming utilities within a group that share a common set of data and for which reliance on a single indicator could be misleading. The main takeaways from this diagnostic across utility types are as follows: (a) for generation utilities (GUs), the best-performing utility is Qatrana Electric Power Company (Jordan), followed by Al-Kamil Power Plant (Oman) and ACWA Power Barka (Oman); (b) for DUs, Electricity Distribution Company (Jordan) is the best- performing utility in the group, followed by LYDEC (Morocco) and Jordan Electric Power Company (Jordan); and (c) for VIUs, the best performance is by Saudi Electricity Company (Saudi Arabia), followed by Société Nationale de l’Électricité et du Gaz (Algeria). Well-Targeted Institutional and Economic Reforms Would Boost MENA’s Power Sector The variety of organizational structures found in the electricity sector around the world is quite striking. This study reveals that the MENA region is no exception. Utilities are central to all the organizational models encountered in ­ the region, but otherwise these models show substantial institutional and contex- tual differences, some of which have been credited with, or blamed for, differ- ences in utilities’ performance. Our assessment of the correlations between various institutional and ­contextual characteristics (utility type, size, ownership, the presence of a separate regulatory agency, and national income) and performance indicators, despite limitations (notably the use of cross-sectional rather than time-series data), suggests how and Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Conclusion 169 where reform policies may be most effective. Of the 36 performance indicators used for this analysis, 25 showed impact for one of the characteristics; in 14 cases, more than one characteristic (or “driver”) was statistically significant. The results support the hypothesis that performance differences between utilities are likely to be correlated with institutional and economic policy variables, although a more thorough analysis is needed to be able to establish causality. Utility type and size are the policy-related drivers that were most often signifi- cant (each for 30 percent of the indicators tested), while ownership type (public or private) and the presence of an independent regulator were significant for about 20 percent of the indicators tested. National income level was significant in 35 percent of the tests, indicating that this variable should be considered in any comparison across economies. The impacts of reform would not be felt across all indicators but are likely to be concentrated in certain aspects of performance. Table CL.2 shows that the significant results for each driver are concentrated within two or three categories of indicators. For example, utility type has a substantial proportion of significant links to the indicator categories of losses efficiency, profitability, and consump- tion and billing, and no links at all to the categories of systems and operational efficiency, cost structure, cost-recovery, balance sheet, and metering. Ownership and regulation are linked to cost efficiency and labor efficiency. This suggests that improvements in cost efficiency and labor efficiency are particularly susceptible to reform efforts, because ownership and regulation are relatively easy factors to adjust. Other categories of indicators may be influenced by other drivers or by a complex combination of factors that simple testing of one characteristic at a specific point in time was unable to duplicate. The Case Studies Yield Valuable Insights on the Variety and Nature of Reform Paths Egypt, Jordan, Morocco, and Oman are analyzed in detail in the case studies presented in part II. The four countries represent the diverse challenges faced by economies in the region, as well as different paths taken toward electricity reform over the past 10–15 years. The four countries are characterized by quite different economic and political environments, which affect the degree of ease or difficulty involved in implementing reforms. Egypt has not enjoyed the political stability often needed when undertaking significant reforms. Its experience indicates that demand shocks linked to politi- cal tensions may have a much stronger impact on the sector’s commercial and financial performance than on its technical and operational performance. Jordan has had to address both a demand and a supply shock. On the supply side, it has been affected by the need to drastically change its sources of energy owing to a break in gas supply from its main supplier in 2012. On the demand side, it has had to deal with unexpected increases resulting from a large inflow of refugees. The case study illustrates the impact of efforts to significantly scale up the role of the private sector in absorbing these shocks. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 170 Table CL.2   Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities Separate Utility regulatory Classes of utilities included Indicator Category Number Mean type Size Income Ownership agency present VIU vs. DU Load factor System and 23 0.56 0.80 0.25 0.96 0.07* 0.63 VIU vs. GU Capacity factor operational 20 0.54 0.07* 0.61 0.12 0.43 S efficiency VIU vs. GU Availability factor 11 0.93 0.50 0.71 0.04** 0.50 0.50 VIU vs. TU vs. DU Network maintenance 10 0.02 0.79 0.41 0.85 0.52 0.20 VIU vs. DU Share of meters replaced (%) 9 0.02 0.41 0.40 0.70 0.29 0.82 VIU vs. TU Transmission losses Losses efficiency 3 0.03 0.86 S S S S VIU vs. DU Distribution losses 37 0.13 0.001** 0.76 0.52 0.63 0.69 VIU vs. DU Technical losses 18 0.075 0.0003** 0.37 0.32 0.22 0.14 VIU vs. DU Nontechnical losses 18 0.049 0.0003** 0.88 0.32 0.13 0.48 VIU vs. GU vs. TU vs. DU OPEX/employee Cost efficiency 48 274,000 n.a. 0.0001** 0.003** 0.006** 0.39 VIU vs. DU OPEX/connection 36 723.0 n.a. 0.16 0.0001** 0.99 0.80 VIU vs. DU OPEX/kWh sold 36 0.11 n.a. 0.002** 0.51 0.41 0.0001** VIU vs. TU vs. DU OPEX/km 37 24,381.0 n.a. 0.006** 0.95 0.02** 0.001** VIU vs. DU Residential connections/employee Labor efficiency 24 238 n.a. 0.09* 0.54 0.15 0.02** VIU vs. DU Energy sales/employee 31 170,000 n.a. 0.03** 0.48 0.0007** 0.005** VIU vs. DU Total revenues/employee 34 212,000 n.a. 0.1* 0.70 0.004** 0.001** VIU vs. GU Cost fuels/OPEX Cost structure 22 0.65 0.12 0.16 0.47 0.97 0.02** VIU vs GU Energy purchases + fuels/OPEX 8 0.77 S 0.05** 0.23 S 0.70 VIU vs. GU vs. DU Labor cost/OPEX 35 0.13 0.22 0.03** 0.02** 0.13 0.29 VIU vs. DU Energy sales/OPEX Cost recovery 32 0.95 0.42 0.49 0.07* 0.83 0.15 VIU vs. DU Energy sales/costs 19 0.82 0.11 0.13 0.03** 0.54 0.48 table continues next page Table CL.2   Tests of Equality between Subgroups of Factors Related to Indicator Mean Values (Probabilities) Using One-at-a-Time Testing, MENA Utilities (continued) Separate Utility regulatory Classes of utilities included Indicator Category Number Mean type Size Income Ownership agency present VIU vs. DU Accounts receivable Balance sheet 26 161 0.11 0.22 0.06* 0.84 0.63 VIU vs. GU vs. TU vs. DU Debt/equity 47 7.08 0.24 0.05** 0.04** 0.62 0.67 VIU vs. GU vs. TU vs. DU Assets/liabilities 53 1.17 0.32 0.0005** 0.31 0.56 0.84 VIU vs. GU vs. TU vs. DU Return on assets Profitability 49 0.3% 0.39 0.07* 0.22 0.05* 0.40 VIU vs. GU vs. TU vs. DU Return on equity 46 4.6% 0.009** 0.10 0.15 0.03** 0.12 VIU vs. DU Total energy volume/connection Consumption 35 6.4 0.002** 0.36 0.001** 98.0 0.21 VIU vs. DU Residential energy volume/ and billing connection 23 4.0 0.01** 0.72 0.0001** 0.62 0.51 VIU vs. DU Total billing/connection 27 297 0.17 0.005** 0.0001** 0.037** 0.09* VIU vs. DU Residential billing/connection 22 258 0.59 0.0001** 0.007** 0.37 0.34 VIU vs. DU Collection rate 15 88% 0.03** 0.003** 0.86 0.51 0.08* VIU vs. DU Share of installed meters (%) Metering 15 96% 0.32 0.33 0.02** 0.72 0.75 VIU vs. TU vs. DU SAIFI Customer 15 1.6 0.02** 0.70 0.06* 0.37 0.69 management VIU vs. TU vs. DU SAIDI 12 28.6 0.46 0.35 0.72 0.49 0.57 and service VIU vs. TU vs. DU CAIDI quality 9 52 0.21 0.46 S S 0.20 VIU vs. TU vs. DU Duration of interruptions 5 2.0 S 0.99 0.03** 0.32 0.03** Source: World Bank calculations. Note: Significant results are shaded in light red; performance indicators for which more than one factor gave significant results in one-at-a time testing are shaded in green; tests that are inappropriate are shaded in blue. CAIDI = Customer Average Interruption Duration Index; DU = distribution utility; GU = generation utility; km = kilometer; kWh = kilowatt-hour; MENA = Middle East and North Africa; n.a. = not applicable (tests are inappropriate); OPEX = operating expenses; S = singular dataset so estimation is not possible; SAIDI = System Average Interruption Duration Index; SAIFI = System Average Interruption Frequency Index; TU = transmission utility; VIU = vertically integrated utility. Significance level: * = 10 percent, ** = 5 percent. 171 172 Conclusion Morocco illustrates how electricity reforms can be implemented in a hybrid market in which regional utilities cover electricity as well as water and sanitation. This peculiarity makes it difficult to differentiate the allocation of resources across the two activities but does allow for the introduction of cross-subsidies. Finally, Oman is a relatively small economy where policy reforms have eased access to private financing in the sector. It now has long experience with an unbundled electricity sector. Private GUs are also involved in the desalination efforts that ensure the sultanate’s water supply. More Systematic Monitoring of Power Sector Performance Is Needed The MENA Electricity Database can be used not only to produce a snapshot of the region’s power sector but also to clarify the managerial, technical, and policy steps that might be required to meet fast-growing demand from all eco- nomic actors, including residential users. Just as important, and perhaps more subtly, the database provides a baseline against which future progress can be tracked and measured. To be effective and to ensure accountability of policy makers and managers, progress needs to be measured from baseline to target, which is how comparisons can become an input for policy. Targets are best set at the firm level for most operational matters, but sector-level targets are needed as well if governments are to address the fiscal and social concerns and constraints raised in the analysis. The new database produced for this report offers the region access to a com- parable dataset for a statistically significant sample of economies both within and outside the region. The comparable components of the dataset cover indicators in three broad performance categories: (a) technical and operational, (b) finan- cial, and (c) commercial. But the dataset also exposes the monitoring weaknesses of the region. Very little comparable information exists for GUs, for example. On many performance indicators, comparability is not possible, either for lack of data or because the indicators have different meanings for the different types of utility. The gaps in the data needed for good policy and management are real but not unsurmountable. To help fill them, authorities in the region may wish to impose on regulated industries guidelines and other information-sharing requirements derived from modern regulatory practice. For unregulated companies, standard accounting reports and annual balance sheets can go a long way toward supplying the raw data needed to improve monitoring of the region’s electricity sector, provided the will to use that information is present. Without a political commitment to improve the dataset and to use it to moni- tor progress and fine-tune policy, it will be difficult for the sector’s decision mak- ers to track efforts to cut the sector’s financing deficits and close its service gaps. The analysis provided here has shown how much room there is to cut specific costs and to enhance revenue. It has also shown, for many economies in the region, the unsustainability of a business-as-usual approach. Without the checks and balances provided by an effective monitoring system, progress in addressing Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Conclusion 173 challenges cannot be tracked adequately. The case for change in the region’s monitoring practices is thus strong—and change is possible. Many policy makers are already moving in the right direction by making important institutional changes. How fast and how intensively they move is likely to determine how quickly the financing and service needs of the sector are met. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 APPENDIX A Manual of Indicators and Data Sources MENA Electricity Database The MENA Electricity Database covers 67 utilities in 14 economies of the Middle East and North Africa (MENA) region. This information was gathered through a questionnaire sent to electricity utilities, in which they were asked if they performed one or several of the following activities: generation, transmis- sion, and distribution. According to their responses, the utilities were categorized as vertically integrated utilities (VIUs), generation utilities, distribution utilities, or transmission utilities. Detailed data collected through the questionnaire enabled the calculation of 36 indicators of the utilities’ performance across two key measures of efficiency (operational and financial) and two measures of service quality (technical and commercial). Data were obtained for the years 2009–14, making it possible to follow the evolution of the performance of one specific utility. Core Performance Indicators for 67 Utilities in the MENA Region A list of key data that would be required to calculate a set of priority indicators was identified. An attempt was made to focus on primary data, that is, data that would be commonly generated through internal processes and reported to the utility’s management or contained in standard financial, technical, and commer- cial reports. The indicators selected are those that can provide insights into key technical, commercial, and financial elements of utility performance. These indicators are commonly used by the industry, regulators, and academic and international orga- nizations to assess the different dimensions of utility performance, taking into account key specificities. Shedding Light on Electricity Utilities in the Middle East and North Africa   175   http://dx.doi.org/10.1596/978-1-4648-1182-1 176 Manual of Indicators and Data Sources The final choice of data and indicators was based on a review of international experiences of similar data collection and benchmarking exercises. Various reports related to global and regional initiatives were reviewed, as well as similar initiatives and programs of national or local energy regulators such as the Office of Gas and Electricity Markets (OFGEM, United Kingdom), the Ontario Energy Board (OEB, Canada), and the Australian Energy Market Commission (AEMC), as well as specialized reports and analysis from international consultants (for example, Hesmondhalgh and others 2012). It was also ensured that major inputs and out- puts required for statistical benchmarking analysis would be available if our information requests were satisfied. Table A.1 contains a total of 36 indicators that are grouped into (a) 16 technical and operational indicators; (b) 10 financial indicators; and (c) 10 commercial indicators, which are further subdivided into four, four, and three subgroups respectively. Table A.1  Descriptions of the 36 Core Indicators Name Unit Description Subactivity Technical and operational indicators System and operational efficiency Load factor % (Energy delivered to distribution network in MWh/8,760 DU-VIU hours)/maximum demand on the interconnected system in MW Capacity factor % (Total net generation in MWh/8,760 hours)/total installed GU-VIU generation capacity in MW Availability factor % [(Total installed generation capacity * 8,760) − (total GU-VIU capacity hours out of service)] * 100/(total installed generation capacity * 8,760) Network maintenance % Length of existing network subject to major repair or TU-DU-VIU replacement/(length of transmission network + length of distribution network) Number of meters % Number of meters of existing connections replaced/total DU-VIU replaced/total number of meters number of meters Losses efficiency Transmission losses % Energy lost during transmission of power as a percentage TU-VIU of the sum of the total net energy generated and the energy purchased Distribution losses % Energy lost during distribution of power as a percentage DU-VIU of the sum of the total net energy generated and the energy purchased Technical losses % Distribution losses due to the technical characteristics of DU-VIU the distribution network Nontechnical losses % Distribution losses due to unmetered and unbilled DU-VIU consumption due to illegal connections, inaccurate estimations of consumptions, billing errors table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Manual of Indicators and Data Sources 177 Table A.1  Descriptions of the 36 Core Indicators (continued) Name Unit Description Subactivity Cost efficiency (total OPEX) Total OPEX/FTE employee $/employee Total OPEX/(number of FTE own employees + number GU-TU-DU-VIU of FTE employees from outsourced contracts) Total OPEX/connection $/connection Total OPEX/number of connections DU-VIU Total OPEX/kWh sold $/kWh Total OPEX/energy billed (excluding exports) DU-VIU Total OPEX/km of $/km Total OPEX/(length of transmission network + length of TU-DU-VIU network distribution network) Labor efficiency # of residential Connections/ (Total number of residential connections)/(number of FTE DU-VIU connections/FTE employee own employees + number of FTE employees from employee outsourced contracts) Energy sales ($)/FTE $/employee Total sales in $ related to energy service (consumption + DU-VIU employee fixed charges)/(number of FTE own employees + number of FTE employees from outsourced contracts) Total revenues ($)/FTE $/employee Total utility’s revenues in $/(number of FTE own employees DU-VIU employee + number of FTE employees from outsourced contracts) Financial indicators Cost structure Share of cost of fuel, % Cost of fuel, lubricant, gas, and coal/total OPEX GU-VIU lubricant, gas, and coal in total OPEX Share of energy % (Cost of fuel, lubricant, gas, and coal + energy purchases)/ VIU purchases and cost of total OPEX fuel, lubricant, gas, and coal in total OPEX Share of labor cost in total % Labor cost/total OPEX GU-TU-DU-VIU OPEX Cost-recovery Energy sales/total OPEX % Revenues related to energy consumption and service DU-VIU in $/OPEX in $ Energy sales/total costs % Revenues related to energy consumption and service DU-VIU in $/(OPEX + depreciation of fixed assets + other depreciation and provisions − net interests) Balance sheet (Accounts receivable/ Days (Accounts receivable at year end/energy sales in $) * 365 DU-VIU sales) * 365 Debt/equity % Total debt at year end in $/total equity GU-TU-DU-VIU Current assets/current % Total of current assets/total of current liabilities GU-TU-DU-VIU liabilities Profitability Return on assets % Net profit of the year/net fixed assets at year end GU-TU-DU-VIU Return on equity % Net profit of the year/total equity GU-TU-DU-VIU table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 178 Manual of Indicators and Data Sources Table A.1  Descriptions of the 36 Core Indicators (continued) Name Unit Description Subactivity Commercial indicators Average consumption and billing Total energy volume sold kWh/ Total sales of energy in kWh/# of connections DU-VIU (kWh)/connection connection Residential energy kWh/ Residential sales in kWh/# of residential connections DU-VIU volume sold (kWh)/ connection connection Total billing ($)/ $/connection Total sales related to energy service (consumption + fixed DU-VIU connection charges)/# of connections Residential billing ($)/ $/connection Residential energy sales in $/# of residential connections DU-VIU connection Collection rate % Income effectively collected during year/income billed DU-VIU Metering Share of installed meters % # of meters/# of connections DU-VIU Customer management and service quality SAIFI Thousands Takes into account interruptions affecting customers due TU-DU-VIU to planned and unplanned events SAIDI Minutes Takes into account interruptions affecting customers due TU-DU-VIU to planned and unplanned events CAIDI Minutes Tracks interruptions due to planned and unplanned events TU-DU-VIU Duration of interruption Minutes Corresponds to the minimum duration of interruptions in TU-DU-VIU taken into the consumer-side customers that are considered consideration for toward customer reliability indicators (planned and system interruptions unplanned events) affecting customers (for example, SAIDI, SAIFI, and CAIDI customer measures) Source: World Bank calculations. Note: CAIDI = Customer Average Interruption Duration Index; DU = distribution utility; FTE = full-time equivalent; GU = generation utility; km = kilometer; kWh = kilowatt-hours; MW = megawatts; MWh = megawatt-hours; OPEX = operating expenses; SAIDI = System Average Interruption Duration Index; SAIFI = System Average Interruption Frequency Index; TU = transmission utility; VIU = vertically integrated utility. Data Collection: Results and Challenges Data Results and Sources The main source of data for the MENA Electricity Database was an elaborate questionnaire developed for the purpose of the study. Supplementary data were obtained from the utilities’ annual, financial, and activity reports, as needed. For the non-MENA comparative portion of the database, data were obtained for 181 economies from several sources, mainly from the following: • European countries from the AF-MERCADOS EMI report, 2013 (except for Denmark) • African utilities from the Africa Infrastructure Country Diagnostic, 2005; (see Eberhard and others 2008) Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Manual of Indicators and Data Sources 179 • African utilities from Africa Power Subsidies, 2010; (see Trimble and others 2016) • Latin American utilities from the Commission for Energy Regulation (CER) report, 2010; (see Andres and others 2012) • Latin American utilities from the Latin America and Caribbean (LAC) Database, 2007; (see Andres and others 2012) • Australia, Denmark, Sri Lanka, and Vietnam from Readiness for Investment in Sustainable Energy, 2013; (see Banerjee and others 2016) Following is a summary of the number of relevant observations obtained for the non-MENA utilities, by type of utility (table A.2). A total of 1,041 indicator points were obtained for the 36 key indicators. The dominant type in our sample is distribution utilities, most of them from the LAC region. Regarding the collection of data on MENA, two main points are to be highlighted: • Relevance of indicators to the type of utility Some indicators are relevant only to certain utility types. For example, distribu- tion losses are experienced by both vertically integrated and distribution utilities but are not relevant to generation or transmission utilities. Other indicators, such as those relating to operating expenses (OPEX) are not comparable between structure types—the nature of the distribution business is such that the OPEX of a distribution utility would be much lower than that of a VIU, which includes some generation to serve the same number of customers. • Availability of indicators (and data used to calculate indicators) from the utilities Because not all these utilities actually collect the relevant data required in our data collection exercise, missing observations were commonly encountered. This means that for some indicators the number of available observations can be small. Table A.2  Number of Indicator Points and Number of Utilities, by Type of Utility and Region Vertically integrated utility Distribution utility Generation utility Transmission utility Number of Number of Number of Number of indicator Number indicator Number indicator Number indicator Number points of utilities points of utilities points of utilities points of utilities Africa 124 25 27 5 0 0 0 0 Asia 7 1 40 9 4 2 2 2 Australia 0 0 13 1 0 0 0 0 Europe 0 0 12 3 8 2 5 2 LAC 46 9 721 117 0 0 0 0 Israel 5 1 0 0 0 0 0 0 United States 27 2 0 0 0 0 0 0 Source: World Bank calculations. Note: LAC = Latin America and the Caribbean. “Africa” here refers to Sub-Saharan Africa. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 180 Manual of Indicators and Data Sources Table A.3  Number of Indicator Points Collected, 2009–13 2009 2010 2011 2012 2013 Total collected 703 744 817 873 1,164 Not applicable 93 97 111 124 187 Total collected of 1,600 applicable 610 647 706 749 977 Share of collected and applicable (%) 38 40 44 47 61 Source: World Bank calculations. Table A.4  Number of Data Points Collected, 2009–13 2009 2010 2011 2012 2013 Total collected 812 865 930 974 1,271 Not applicable 3 3 4 5 19 Total collected of 2,027 applicable 809 862 926 969 1,252 Share of collected and applicable (%) 40 43 46 48 63 Source: World Bank calculations. Table A.3 presents the number of actual observations collected and available for the 36 core indicators used in this study. If all the 67 utilities were to ideally provide all the data required, we would have expected to obtain 1,600 indicator observations per year. Yet this is not the case. The highest number of relevant (or applicable) observations were obtained for the year 2013, where 977 observations were obtained out of 1,600 (equivalent to 61 percent). These indicators were calculated from the data points collected (40 data points required for the calculation of the 36 core indicators), shown in table A.4. Main Challenges Most of the gaps encountered involved financial and technical data. The most common challenges encountered during the data collection process are pre- sented below. Different types of data. Different types of data were requested, that is, regulatory, financial, technical, and commercial. Therefore, responding to the questionnaire required the collaboration of different departments within the utility and the involvement of different senior executives who were not always accustomed to working together. Format of financial information. The reporting format of utilities’ financial state- ments differ depending on the legal environment (for example, if accounting plans and financial reporting formats are defined by local law or not) and accounting practices and philosophies (for example, the Anglo-Saxon accounting model versus the French or Spanish model). Many utilities have already or are progressively adopting international financial reporting standards (IFRS), but others have not done so yet. ­ Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Manual of Indicators and Data Sources 181 Generally, the financial and accounting data requested here were to be easily identifiable in the main financial statements of the utility: balance sheet, income, cash flow, and use and sources of funds (the last two were not always available). When there was a doubt about a financial or accounting variable, instructions were given to refer to specific definitions presented. Network management practices. Indicators related to service interruptions, such as direct indicators of frequency, durations, and related indexes are commonly used in the industry to assess the availability of the service. See, for instance, the System Average Interruption Frequency Index (SAIFI), System Average Interruption Duration Index (SAIDI), and Customer Average Interruption Duration Index (CAIDI). Still, the accuracy and comparability of these indica- tors will depend on the implementation and configuration of network manage- ment automated recording systems (such as fault incidence recording systems or supervisory control and data acquisition [SCADA]/energy management system [EMS] with capabilities to perform these function). References AF-MERCADOS EMI. 2013. “Final Report on Proposal of Key Performance Indicators in the Energy Sector in Serbia.” Unpublished report, Madrid, Spain. Andres, Luis A., J. Schwartz, and J. L. Guasch. 2012. “Uncovering the Drivers of Utility Performance: Lessons from Latin America and the Caribbean on the Role of the Private Sector, Regulation, and Governance in the Power, Water, and Telecommunication Sectors.” World Bank, Washington, DC. Banerjee, Sudeshna Ghosh, A. Moreno, J. Sinton, T. Primiani, and J. Seong. 2016. “Regulatory Indicators for Sustainable Energy: A Global Scorecard for Policy Makers.” World Bank, Washington, DC. Eberhard, A., V. F oster, C. Briceño-Garmendia, F. Ouedraogo, D. Camos, and M. Shkaratan. 2008. “Underpowered: The State of the Power Sector in Sub-Saharan Africa.” Africa Infrastructure Country Diagnostic (AICD), summary of background paper 6, World Bank, Washington, DC. Hesmondhalgh, Serena, W. Zarakas, T. Brown. 2012. “Approaches to Setting Electric Distribution Reliability Standards and Outcomes.” The Brattle Group. Trimble, C., M. Kojima, I. P. Arroyo, and F. Mohammadzadeh. 2016. “Financial Viability of Electricity Sectors in Sub-Saharan Africa: Quasi-Fiscal Deficits and Hidden Costs.” Policy Research Working Paper 7788, World Bank, Washington, DC. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 APPENDIX B Utilities Considered and Their Basic Characteristics The MENA Electricity Database (MED) covers 67 utilities in 14 MENA econo- mies. Table B.1 provides basic power sector characteristics of the 14 MENA economies of interest, including its size, market structure, and key stakeholders. Table B.2 lists the 67 MENA utilities included in the MED. Table B.3 presents the five categories of institutional and contextual characteristics of utilities used in chapter 5 to assess the drivers of performance: type, size, ownership, economy income level, and presence of separate regulatory agency. Finally, table B.4 pro- vides a list of the utilities and their respective economies for which observations were available and used in the non-MENA analysis of this study. Non-MENA data used in the study were obtained from 181 economies. Owing to the lack of a time series for the non-MENA data, the latest available year was used from the range 2005 to 2016. Because data used in the MENA analysis were limited to the period 2009 to 2014, non-MENA observations from 2005 to 2008 were associated with the year 2009. Shedding Light on Electricity Utilities in the Middle East and North Africa   183   http://dx.doi.org/10.1596/978-1-4648-1182-1 Table B.1  Summary of the Electricity Sector for 14 MENA Economies, 2013 184 Country or Installed Vertical economy capacity (MW) Generation Distribution Sector regulator? integration? Comments Algeria 12,949 Société Algérienne de Production Société de Distribution de Yes Single buyer, Socièté Nationale de de l’Electricité (SPE) l’Electricité et du Gaz d’Alger (Commission de unbundled l’Electricité et du Gaz (SDA), Société de Distribution Régulation de (SONELGAZ) is a holding de l’Electricité et du Gaz du l’Electricité et du company with SPE, Centre (SDC), Société de Gaz, CREG) transmission assets, ISO, Distribution de l’Electricité et and four distribution du Gaz de l’Est (SDE), and companies. For the Société de Distribution de purpose of this study, l’Electricité et du Gaz de SONELGAZ is considered l’Ouest (SDO) as a VIU in Algeria because of data constraints. Bahrain 3,934 BOOT model, with several Electricity and Water Authority No Single buyer, Single buyer. EWA is privately owned facilities (EWA) (part of Ministry of partially considered as a VIU. Electricity and Water) unbundled Djibouti 100 Electricité de Djibouti (EDD) EDD No Yes One of only two countries (with the Republic of Yemen) in which access (53%) is an issue. Egypt, Arab 29,312 Six public (regional) and nine Nine regional: Alexandria, South Yes Single buyer, Egyptian Electricity Holding Rep. private (three BOOT and six Cairo, North Cairo, El-Behera, (Egyptian Electric unbundled Company (EEHC) and its IPPs) South Delta, North Delta, Upper Utility and affiliates are responsible Egypt, Middle Egypt, Canal Consumer for generation, Protection transmission, and Regulatory distribution. EEHC is a Agency, EgyptERA) single buyer with elements of a monopoly. For the purpose of this study, we consider the six public GUs and the nine regional DUs, and the Egyptian Electricity Transmission Company (EETC) (a TU). table continues next page Table B.1  Summary of the Electricity Sector for 14 MENA Economies, 2013 (continued) Country or Installed Vertical economy capacity (MW) Generation Distribution Sector regulator? integration? Comments Iraq 19,354 Electric Energy Production (EEP) Electric Energy Distribution (EED) No Yes EEP, EED, and the Electric Energy Transmission are three divisions within the Ministry of Electricity. Jordan 3,452 Samra Electric Power Generating Three private: Jordan Electric Yes Single buyer, Transmission activities are Company (SEPCO), state Power Company (JEPCO), (Electricity Regulatory unbundled under the National owned (25%); Central Electricity Distribution Company Commission, ERC) Electricity Power Electricity Generating (EDCO), Irbid District Electricity Company (NEPCO). Company (CEGCO), private Company (IDECO) (56%); Amman East Power Plant (AES PSC), private (16%) Lebanon 2,313 Electricité du Liban (EdL) and IPPs EdL and three DSPs with two-year No Yes “concession” contracts Morocco 6,677 Office National de l’Electricité ONEE (50% customers) and four No Single buyer, There are some private et l’Eau Potable (ONEE) and concessions—Lyonnaise des partially players in generation and several IPPs Eaux de Casablanca (LYDEC) for unbundled distribution. Casablanca, REDAL for Rabat, AMENDIS-TA for Tanger, AMENDIS-TE for Tetouan—and seven municipal multiservice distribution utilities Oman 4,938 Some public and several private Three state-owned distribution Yes Single buyer, Oman Electricity actors: United Power Company and supply companies (Authority for unbundled Transmission Company (UPC), Sembcorp Salalah Power Electricity (OTEC) is responsible for Company (SSPWC), Al Rusail Regulation, AER) transmission of electricity Power Plant (ARPP), Barka Power and Oman Power and and Desalination Plant (BPDP), Water Procurement Al Kamil Power Plant (AKPP), Company (OPWPC) is the Phoenix Power Company (PPC), single buyer of electricity ACWA Power Barka (APBS), and water. Shoar Power Plant (SPP). table continues next page 185 186 Table B.1  Summary of the Electricity Sector for 14 MENA Economies, 2013 (continued) Country or Installed Vertical economy capacity (MW) Generation Distribution Sector regulator? integration? Comments Qatar 8,756 Private sector and IWPPs Qatar General Electricity and Water No Single buyer, Corporation (KAHRAMAA) unbundled Saudi Arabia 53,588 Saudi Electricity Company (SEC) SEC Yes Yes SEC is a vertically integrated and several private: Saline (Electricity and monopoly. Power for Water Conversion Corporation Cogeneration desalination is an (SWCC), Saudi Aramco, Regulatory important player. Tihamah, Power and Utility Authority, ECRA) Saudi Arabia plans to Company for Jubail and Yanbu unbundle SEC soon. (MARAFIQ), Water and Electricity LLC. (WEC) including several large industrial firms Tunisia 4,095 Société Tunisienne de l’Electricité STEG No Yes et du Gaz (STEG) and two IPPs West Bank 125 Electricity supply almost entirely Five: Northern Electricity Yes Single buyer, dependent on the Israel Distribution Company (NEDCO) (Palestinian Electricity unbundled Electric Corporation (IEC) and for North West Bank, Southern Regulatory Gaza IPP (2x70 MW) Electricity Company (SELCO) for Council, PERC) South West Bank, Hebron Electricity Corporation (HEPCO) for Hebron, Jerusalem District Electricity Company (JEDCO) for Jerusalem, Gaza Electricity Distribution Company (GEDCO) for Gaza Yemen, Rep. 1,520 Public Electricity Corporation PEC No Yes PEC is vertically integrated. (PEC) Yemen, Rep., is one of only two countries of the sample in which access (48%) is an issue, along with Djibouti. Source: World Bank calculations. Note: BOOT = build-own-operate-transfer; DSP = distribution service provider; DU = distribution utility; GU = generation utility; IPP = independent power producer; ISO = independent system operator; IWPP = independent water and power producer; MENA = Middle East and North Africa; MW = megawatts; TU = transmission utility; VIU = vertically integrated utility. Utilities Considered and Their Basic Characteristics 187 Table B.2  Names and Abbreviations of MENA Utilities Country or economy Utility name Abbreviation Algeria Société Nationale de l’Électricité et du Gaz SONELGAZ Bahrain Electricity and Water Authority EWA Djibouti Électricité de Djibouti EdD Egypt, Arab Rep. Alexandria Electricity Distribution Company AEDC Cairo Electricity Production Company CEPC Canal Electricity Distribution Company CEDC East Delta Electricity Production Company EDEPC Egyptian Electricity Transmission Company EETC El-Behera Electricity Distribution Company EEDC Middle Delta Electricity Production Company MDEPC Middle Egypt Electricity Distribution Company MEEDC North Cairo Electricity Distribution Company NCEDC North Delta Electricity Distribution Company NDEDC South Cairo Electricity Distribution Company SCEDC South Delta Electricity Distribution Company SDEDC Upper Egypt Electricity Distribution Company UEEDC Upper Egypt Electricity Production Company UEEPC West Delta Electricity Production Company WDEPC Iraq Ministry of Electricity MOE Jordan AES Levant Holding BV Jordan PSC AES Levant Amman East Power Plant AES PSC Amman-Asia Electric Generating Company AAEPC Central Electricity Generating Company CEGCO Electricity Distribution Company EDCO Irbid District Electricity Company IDECO Jordan Electric Power Company JEPCO National Electric Power Company NEPCO Qatrana Electric Power Company QEPCO Samra Electric Power Generating Company SEPCO Lebanon Électricité du Liban EdL Morocco AMENDIS Tanger AMENDIS TANGER AMENDIS Tetouan AMENDIS TETOUAN Lyonnaise des Eaux de Casablanca LYDEC Office National de l’Électricité et de l’Eau Potable ONEE RADEEL RADEEL REDAL Rabat REDAL Régie Autonome de Distribution d’Eau d’Électricité et d’Assainissement RAK liquide de la Province de Kenitra Régie Autonome de Distribution d’Eau et d’Électricité de Marrakech RADEEMA Régie Autonome de Distribution d’Eau et d’Électricité de Meknès RADEM Régie Autonome de Distribution d’Eau, d’Électricité et d’Assainissement RADEEJ liquide des Provinces d’El Jadida et de Sidi Bennour Régie Autonome Intercommunale de Distribution d’Eau et d’Électricité de Safi RADEES Régie Autonomie Intercommunale de Distribution d’Eau et d’Électricité de Fès RADEEF table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 188 Utilities Considered and Their Basic Characteristics Table B.2  Names and Abbreviations of MENA Utilities (continued) Country or economy Utility name Abbreviation Oman ACWA Power Barka APBS Al Batinah Power Company ABPC Al Suwadi Power Company ASPC Al-Ghubra Power and Desalination Company GPDCO Al-Kamil Power Plant AKPP Al-Rusail Power Plant ARPP Barka Power and Desalination Plant BPDP Dhofar Power Company DPC Majan Electricity Company MJEC Mazoon Electricity Distribution Company MZEC Muscat Electricity Distribution Company MEDC Oman Electricity Transmission Company OETC Phoenix Power Company PPC Rural Areas Electricity Company RAECO Sembcorp Salalah Power and Water Company SSPWC Sohar Power Plant SPP United Power Company UPC Wadi Al-Jizzi Power Company WAJPCO Qatar Qatar General Electricity and Water Corporation KAHRAMAA Saudi Arabia Saudi Electricity Company SEC Tunisia Société Tunisienne de l’Électricité et du Gaz STEG West Bank Jerusalem District Electricity Company JDECO Northern Electricity Distribution Company NEDCO Tubas District Electricity Company TUBAS Yemen, Rep. Public Electricity Corporation PEC Source: MENA Electricity Database. Note: MENA = Middle East and North Africa. Table B.3  Characteristics of MENA Utilities Presence of Country or Country income separate economy Utility name Type Size Ownership level regulatory agency Algeria Société Nationale de VIU Big Public Upper middle Yes l’Électricité et du Gaz Bahrain Electricity and Water VIU Medium Public High No Authority Djibouti Électricité de Djibouti VIU Small Public Lower middle No Egypt, Arab Rep. Alexandria Electricity DU Big Public Lower middle Yes Distribution Company Cairo Electricity GU Big Public Lower middle Yes Production Company Canal Electricity DU Big Public Lower middle Yes Distribution Company East Delta Electricity GU Big Public Lower middle Yes Production Company table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Utilities Considered and Their Basic Characteristics 189 Table B.3  Characteristics of MENA Utilities (continued) Presence of Country or Economy separate economy Utility name Type Size Ownership income level regulatory agency Egyptian Electricity TU Big Public Lower middle Yes Transmission Company El-Behera Electricity DU Medium Public Lower middle Yes Distribution Company Middle Delta Electricity GU Big Public Lower middle Yes Production Company Middle Egypt Electricity DU Big Public Lower middle Yes Distribution Company North Cairo Electricity DU Big Public Lower middle Yes Distribution Company North Delta Electricity DU Big Public Lower middle Yes Distribution Company South Cairo Electricity DU Big Public Lower middle Yes Distribution Company South Delta Electricity DU Big Public Lower middle Yes Distribution Company Upper Egypt Electricity DU Big Public Lower middle Yes Distribution Company Upper Egypt Electricity GU Big Public Lower middle Yes Production Company West Delta Electricity GU Big Public Lower middle Yes Production Company Iraq Ministry of Electricity VIU Big Public Upper middle No Jordan AES Levant Holding BV GU Small Private Upper middle Yes Jordan psc Amman-Asia Electric GU Medium Private Upper middle Yes Generating Company Amman East Power Plant GU Small Private Upper middle Yes Central Electricity GU Big Private Upper middle Yes Generating Company Electricity Distribution DU Small Private Upper middle Yes Company Irbid District Electricity DU Medium Private Upper middle Yes Company Jordan Electric Power DU Medium Private Upper middle Yes Company National Electric Power TU Big Public Upper middle Yes Company Qatrana Electric Power GU Small Private Upper middle Yes Company Samra Electric Power GU Big Public Upper middle Yes Generating Company Lebanon Électricité du Liban VIU Medium Public Upper middle No Morocco AMENDIS Tanger DU Medium Private Lower middle No AMENDIS Tetouan DU Small Private Lower middle No table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 190 Utilities Considered and Their Basic Characteristics Table B.3  Characteristics of MENA Utilities (continued) Presence of Country or Economy separate economy Utility name Type Size Ownership income level regulatory agency Lyonnaise des Eaux de DU Medium Private Lower middle No Casablanca Office National de VIU Big Public Lower middle No l’Électricité et de l’Eau Potable RADEEL DU Small Public Lower middle No REDAL Rabat DU Medium Private Lower middle No Régie Autonome de DU Small Public Lower middle No Distribution d’Eau d’Électricité et d’Assainissement liquide de la province de Kenitra Régie Autonome de DU Medium Public Lower middle No Distribution d’Eau et d’Électricité de Marrakech Régie Autonome de DU Small Public Lower middle No Distribution d’Eau et d’Électricité de Meknès Régie Autonome de DU Small Public Lower middle No Distribution d’Eau, d’Électricité et d’Assainissement liquide des Provinces d’El Jadida et de Sidi Bennour Régie Autonomie DU Medium Public Lower middle No Intercommunale de Distribution d’Eau et d’Électricité de Fès Régie Autonome DU Small Public Lower middle No Intercommunale de Distribution d’Eau et d’Électricité de Safi Oman ACWA Power Barka GU Small Private High Yes Al Batinah Power GU Medium Public High Yes Company Al Suwadi Power GU Medium Public High Yes Company Al-Ghubra Power and GU Small Public High Yes Desalination Company Al-Kamil Power Plant GU Small Private High Yes Al-Rusail Power Plant GU Medium Private High Yes Barka Power and GU Medium Private High Yes Desalination Plant table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Utilities Considered and Their Basic Characteristics 191 Table B.3  Characteristics of MENA Utilities (continued) Presence of Country or Economy separate economy Utility name Type Size Ownership income level regulatory agency Dhofar Power Company VIU Small Private High Yes Majan Electricity Company DU Small Public High Yes Mazoon Electricity DU Medium Public High Yes Distribution Company Muscat Electricity DU Medium Public High Yes Distribution Company Oman Electricity TU Big Public High Yes Transmission Company Phoenix Power Company GU Big Private High Yes Rural Areas Electricity VIU Small Public High Yes Company Sembcorp Salalah Power GU Small Private High Yes and Water Company Sohar Power Plant GU Medium Private High Yes United Power Company GU Small Private High Yes Wadi Al-Jizzi Power GU Small Public High Yes Company Qatar Qatar General Electricity VIU Medium Public High No and Water Corporation Saudi Arabia Saudi Electricity Company VIU Big Public High Yes Tunisia Société Tunisienne de VIU Big Public Upper middle Yes l’Électricité et du Gaz West Bank Jerusalem District DU Small Public Lower middle Yes Electricity Company Northern Electricity DU Small Public Lower middle Yes Distribution Company Tubas District Electricity DU Small Private Lower middle Yes Company Yemen, Rep. Public Electricity VIU Medium Public Lower middle Yes Corporation Source: MENA Electricity Database. Note: DU = distribution utility; GU = generation utility; MENA = Middle East and North Africa; TU = transmission utility; VIU = vertically integrated utility. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 192 Utilities Considered and Their Basic Characteristics Table B.4  Names and Abbreviations of Non-MENA Utilities Economy Utility name Abbreviation Angola Empresa de Electricidade de Luanda ANG-EDEL Antigua and Barbuda Antigua Public Utilities Authority ANT-APUA Argentina Empresa Distribuidora de Energía Atlántica S.A. ARG-EDE Empresa Distribuidora de Electricidad de Mendoza S.A. ARG-EDMSA Empresa Distribuidora de Electricidad de Santiago del Estero S.A. ARG-EDSTE Empresa Distribuidora de Electricidad de San Luis S.A. ARG-EDSL Empresa Distribuidora Norte S.A. ARG-EDNR Empresa Distribuidora Sur S.A. ARG-EDSR Australia AGL Electricity Ltd. AUS-AGL Belize Belize Electricity Limited BEL-BECOL Benin Electricité de Benin BEN-CEB Bolivia Cooperativa Rural de Electrificación Ltda. BOL-CRE Empresa de Luz y Fuerza Eléctrica Cochabamba  BOL-ELFEC Botswana Botswana Power Corporation BOT-BPC Burkina Faso Société Nationale d’Electricité du Burkina BUR-SONABEL Brazil AES SUL Distribuidora Gaúcha de Energia S.A. BRA-AESUL Ampla Energia e Serviços S.A. BRA-AMPLA Caiuá Serviços de Eletricidade S.A. BRA-CAI Centrais Elétricas de Santa Catarina S. A. BRA-CELSC Centrais Elétricas do Pará S.A. BRA-CELPA Centrais Elétricas Matogrossenses S.A. BRA-CMT CMG BRA-CMG Companhia Campolarguense de Energia BRA-COC Companhia de Eletricidade do Amapá BRA-CEA Companhia de Eletricidade do Amapá BRA-CEAM Companhia de Eletricidade do Estado da Bahia BRA-COELB Companhia de Energia Elétrica do Estado do Tocantins BRA-CELT Companhia Energética de Alagoas BRA-CEAL Companhia Energética de Brasília BRA-CEB Companhia Energética de Goiás BRA-CELG Companhia Energética de Minas Gerais S.A. BRA-CEMIG Companhia Energética de Pernambuco BRA-CELPE Companhia Energética de Roraima BRA-CER Companhia Energética de São Paulo BRA-CPSA Companhia Energética do Ceará BRA-COELC Companhia Energética do Maranhão BRA-CMR Companhia Energética do Rio Grande do Norte BRA-COS Companhia Estadual de Energia Elétrica BRA-CEEE Companhia Força e Luz do Oeste BRA-CFLO Companhia Hidroelétrica São Patrício BRA-CHSP Companhia Jaguari de Energi BRA-CJE Companhia Luz e Força Mococa BRA-MOCCA Companhia Luz e Força Santa Cruz BRA-SANT Companhia Nacional de Energia Elétrica BRA-CNEE table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Utilities Considered and Their Basic Characteristics 193 Table B.4  Names and Abbreviations of Non-MENA Utilities (continued) Economy Utility name Abbreviation Companhia Paranaense de Energia BRA-COP Companhia Paulista de Força e Luz BRA-PAUL Companhia Piratininga de Força e Luz BRA-PIRA COO BRA-COO CPE BRA-CPE CRN BRA-CRN CSPE BRA-CSPE Departamento Municipal de Energia de Ijuí BRA-DEM DME Distribuição S.A. BRA-EBO Electricidade de São Paulo S.A. BRA-ELETROPAULO Empresa Bandeirante de Energia BRA-BAND Empresa Elétrica Bragantina S.A. BRA-EEB Energisa Borborema BRA-BOA Energisa Borborema S.A. BRA-DME Metropolitana Eletricidade de São Paulo S.A. BRA-AES Cabo Verde Electra CAB-ELECTRA Cameroon Cameroon Electricity Corporation CAM-CEC Chad Société Tchadienne d’Eau et d’Electricité CHA-STEE Chile Compañía General de Electricidad Distribución S.A. CHL-CGED Chilectra S. A. CHL-CHIL Compañía Nacional de Fuerza Eléctrica S.A. CHL-CON Colombia Centrales Eléctricas de Nariño S.A. E.S.P. COL-CEDENAR Centrales Eléctricas del Cauca S.A. E.S.P. COL-CEDELCA Centrales Eléctricas del Norte de Santander S.A. E.S.P. COL-CENS CHC Energía COL-CHC COD COL-COD Codensa S.A. E.S.P. COL-CODENSA Electrificadora de Santander COL-ESSA Electrohuila S.A. E.S.P COL-ELECTROHUILA Empresa Eléctrica de Colina Ltda. COL-EEC Comoros Electricite et Eaux des Comores COM-EEDC Congo, Dem. Rep. Société Nationale d’Electricité DRC-SNEL Costa Rica Compañia Nacional de Fuerza y Luz COS-CNFL Instituto Costarricense de Electricidad COS-ICE Croatia Distribucijskog sustava CRO-HEP-ODS Hrvatska elektroprivreda CRO-HEP Denmark Dong Energy DEN-DONG Dominica Dominica Electricity Services Limited DOM-DOMLEC Dominican Republic Empresa Distribuidora de Electricidad del Este DOM-EDESTE Empresa Distribuidora de Electricidad del Norte DOM-EDENOTRE Empresa Distribuidora de Electricidad del Sur DOM-EDESUR Ecuador Corporación para la Administración Temporal Eléctrica de ECU-CATEG-D/EMELEC Guayaquil—Distribución y Comercialización Empresa Eléctrica Ambato S.A. ECU-AMBATO table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 194 Utilities Considered and Their Basic Characteristics Table B.4  Names and Abbreviations of Non-MENA Utilities (continued) Economy Utility name Abbreviation Empresa Eléctrica Azogues S.A. ECU-AZOGUES Empresa Eléctrica Bolivar S.A. ECU-BOLIVAR Empresa Eléctrica Cotopaxi S.A. ECU-COTOPAXI Empresa Eléctrica El Oro S.A. ECU-ELORO Empresa Eléctrica Esmeraldas S.A. ECU-ESMERALDAS Empresa Eléctrica Galapagos S.A. ECU-GALAPAGOS Empresa Eléctrica Guayas Los Ríos S.A. ECU-GUAYAS-LOSRÍOS Empresa Eléctrica Los Ríos S.A. ECU-LOS RIOS Empresa Eléctrica Manabí S.A. ECU-MANABÍ Empresa Eléctrica Milagro S.A. ECU-MILAGRO Empresa Eléctrica Notre S.A. ECU-NORTE Empresa Eléctrica Quito S.A. ECU-QUITO Empresa Eléctrica Regional Centro Sur S.A. ECU-CENTROSUR Empresa Eléctrica Regional Sur S.A. ECU-SUR Empresa Eléctrica Riobamba S.A. ECU-RIOBAMBA Empresa Eléctrica Santa Elena S.A. ECU-STA.ELENA Empresa Eléctrica Santo Domingo S.A. ECU-STO.DOMINGO Empresa Eléctrica Sucumbíos S.A. ECU-SUCUMBÍOS El Salvador AES-El Salvador ELS-AES CAESS ELS-CAE CLESA ELS-CLSA Distribuidora Eléctrica de Usulután (DEUSEM) ELS-DEU Duke Energy El Salvador Co. ELS-DEL La Empresa Eléctrica de Oriente ELS-EEO Ethiopia Ethiopian Electric Power ETH-EEP France Réseau de Transport d’Électricité FRA-RTE Ghana Electricity Company of Ghana Ltd. GHA-ECG Grenada Grenada Electricity Services Ltd. GRE-GRENLEC Guinea Electricité de Guinée GUI-EDG Guinea-Bissau National Electricity and Water Corporation GUB-EAGB Honduras Empresa Nacional de Energía Eléctrica HON-ENEE India Ajmer Vidyut Vitran Nigam Ltd. IND-AVVNL BSES Yamuna Power Limited IND-BYPL Dakshin Gujarat Vij Company Ltd. IND-DGVCL Gujarat Energy Transmission Corporation Limited IND-GETCO Gujarat State Electricity Corporation Limited IND-GSECL Jaipur Vidyut Vitran Nigam Ltd. IND-JVVNL Jodhpur Vidyut Vitran Nigam Ltd. IND-JdVVNL Madhya Gujarat Vij Company Limited IND-MGVCL North Delhi Power Limited IND-NDPL Paschim Gujarat Vij Company Ltd. IND-PGVCL Rajasthan Rajya Vidyut Prasaran Nigam Ltd. IND-RVPNL Rajasthan Rajya Vidyut Utpadan Nigam Ltd. IND-RVUNL Israel Israel Electric Corporation ISR-IEC table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Utilities Considered and Their Basic Characteristics 195 Table B.4  Names and Abbreviations of Non-MENA Utilities (continued) Economy Utility name Abbreviation Jamaica Jamaica Public Service Company Ltd. JAM-JPSco Kenya Kenya Power KEN-KPLC Lesotho Lesotho Electricity Company LES-LEC Malawi Electricity Supply Corporation of Malawi Ltd. MLW-ESCOM Mauritius Central Electricity Board MAU-CEB Mexico Comisión Federal de Electricidad MEX-CFE Luz y Fuerza del Centro MEX-LyFC Mozambique Electricidade de Moçambique MOZ-EDM Namibia Nampower NAM-NAMPOWER Paraguay Administración Nacional de Electricidad PAR-ANDE Peru Consorcio Eléctrico de Villacurí S.A.C. PER-COELVISAC EDELNOR PER-EDELNOR Electro Centro S.A. PER-ELC Electro Nor Oeste S.A. PER-ENOSA Electro Norte S.A. PER-ENSA Electro Oriente S. A PER-ELOR Electro Pangoa S.A. PER-Pangoa Electro Puno S.A.A. PER-ELPUNO Electro Sur Este S.A. PER-ELSE Electro Sur Medio S.A.A. PER-ELSM Electro Sur S.A. PER-ELS Electro Tocache PER-Tocache Electro Ucayali S.A. PER-ELU Electronorte Medio S.A.-Hidradina S.A. PER- ELECTRONORTEMEDIO Empresa de Distribución Eléctrica Cañete S.A. PER-EDECAÑETE Empresa de Servicios Eléctricos Municipales de Paramonga S. A. PER-EMSEMSA Empresa Municipal de Servicios Eléctricos Utcubamba PER-EMSEU Luz del Sur PER-LUZ del Sur Servicios Eléctricos Rioja PER-SERSA Sociedad Electrica del Sur Oeste S.A. PER-SEAL Portugal EDP Distribuição POR-EDP-DIS Energias de Portugal POR-EDP Rwanda Rwanda Energy Group RWA-REG Senegal Société National d’Éléctricité du Sénégal SEN-SENELEC Sierra Leone National Power Authority SIE-NPA South Africa The Electricity Supply Commission ZAF-ESKOM Spain Red Eléctrica de España SPA-REE Sri Lanka Ceylon Electricity Board SRI-CEB St. Kitts and Nevis St. Kitts Electricity Department STK-SED St. Lucia St. Lucia Electricity Services Limited STL-LUCELEC Swaziland Swaziland Electricity Company SWA-SEC Tanzania Tanzania Electric Supply Company Limited TAN-TANESCO Togo Compagnie Energie Electrique du Togo TOG-CEET table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 196 Utilities Considered and Their Basic Characteristics Table B.4  Names and Abbreviations of Non-MENA Utilities (continued) Economy Utility name Abbreviation Uganda Uganda Electricity Board UGA-UEB Uruguay Administración Nacional de Usinas y Trasmisiones Eléctricas URU-UTE United States Consolidated Edison Inc. USA-ConEdison Duke Energy USA-Duke Venezuela, RB La Electricidad de Caracas S.A. EDC-AES Vietnam VietNam Electricity VIET-EVN Zambia Zambia Electricity Supply Corporation Limited ZAM-ZESCO Source: MENA Electricity Database. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 APPENDIX C Quasi-Fiscal Deficit: Hypothesis and Methodology Data Sources and Definitions of Key Variables Table C.1 presents the sources of data that are used for each variable of the economy-level quasi-fiscal deficit (QFD) calculations for the 14 economies considered. As far as the utility-level QFD is concerned, the MENA Electricity Database (MED) was used to fill in all variables except for the gross domestic product, for which the World Development Indicators were used. However, in some cases alternate sources to the MED were used. These exceptions are listed in table C.2. In order to compute the components of the economy-level QFD, a series of assumptions and approximations were used in some cases, as described in table C.3. Shedding Light on Electricity Utilities in the Middle East and North Africa   197   http://dx.doi.org/10.1596/978-1-4648-1182-1 198 Table C.1  Sources of Data Used for the Economy-Level QFD Calculations Qe: end-user Lm: Number of Country or consumption Tc: cost-recovery technical customers Number of Cost of Rct: collection economy (kWh) tariff Te: avg. End-user tariff loss ratesa (connections) employees (FTE) labor c rates GDP Algeria MED Calculations (WDI; Arab Union of WDI MED MED MED Online WDI ESMAP META Electricity (2014), Model; Lazard’s Electricity Tariff in LCOE Analysis, the Arab Countries 2014) Bahrain WDI MED Online MED MED Djibouti MED Onlineb MED MED MED MED Egypt, Arab Rep. WDI WDI EEHC Annual EEHC Annual Estimation MED (average) Report 2014 Report 2014 Iraq MED WDI MED Onlined MED Online (World Bank)e Jordan WDI WDI NEPCO Annual NEPCO Annual Estimation MED (average) Report 2013 Report 2013 MED Lebanon WDI WDI MED MED MED Onlinef Morocco WDI WDI Estimation ONEE contact MED Oman WDI WDI AER Annual Report AER Annual Estimation Estimated 2013 Report 2013 MED table continues next page Table C.1  Sources of Data Used for the Economy-Level QFD Calculations (continued) Qe: End-user Lm: Number of Country or consumption Tc: Cost-recovery Technical customers Number of Cost of Rct: Collection economy (kWh) tariff Te: Avg. End-user tariff loss ratesa (connections) employees (FTE) laborb Rates GDP Qatar WDI WDI KAHRAMAA KAHRAMAA KAHRAMAA MED Sustainability Sustainability Annual Report 2013 Report 2013 Report 2014 Saudi Arabia WDI WDI MED MED MED SEC statistics 2000 to 2014 Tunisia WDI WDI MED Data from utility West Bank MEDg MED MED MED (average) (average) Yemen, Rep. WDI WDI MEDh Estimated MED Source: World Bank calculations. Note: AER = Authority for Electricity Regulation; EEHC = Egyptian Electricity Holding Company; ESMAP = Energy Sector Management Assistance Program; FTE = full-time equivalent employee; GDP = gross domestic product; KAHRAMAA = Qatar General Electricity and Water Corporation; kWh = kilowatt-hours; LCOE = levelized cost of electricity; MED = MENA Electricity Database; META = Model for Electricity Technology Assessment; NEPCO = National Electric Power Company; ONEE = Office National de l’Électricité et de l’Eau Potable; SEC = Saudi Electricity Company; WDI = World Development Indicators. a. WDI technical losses (distribution and transmission losses). b. EUEI (European Union Energy Initiative) 2013. c. Refer to appendix tables C.7 onward for calculation details. d. Iraq Energy Institute 2015. e. World Bank 2016a. f. Lebanon Ministry of Environment and UNDP. g. Calculated as the sum of energy volume billed (from MED) for the three distribution utilities in the West Bank (TUBAS, JDECO, and NEDCO). h. Used 2012 value in the case of the Republic of Yemen owing to lack of data for 2013. 199 200 Quasi-Fiscal Deficit: Hypothesis and Methodology Table C.2  List of Alternate Sources for the Utility-Level QFD Utility Indicator Source Algeria: Share of loss (%) WDI. Socièté Nationale de l’Electricité Bill collection rate L’Algérie profonde/Ouest. n.d. “Plus de 184 millions de DA de et du Gaz (SONELGAZ) pertes pour la Sonelgaz.” http://www.liberte-algerie.com​ /­ouest/plus-de-184-millions-de-da-de-pertes-pour-la​ -sonelgaz-228183/print/1. Bahrain: Electricity and Water Share of loss (%) WDI. Authority (EWA) Djibouti: Total electricity EUEI (European Union Energy Initiative). 2013. Country Power Electricité de Djibouti (EDD) billed Market Brief: Djibouti. Africa-EU Energy Partnership. http:// www.euei-pdf.org/sites/default/files/field_publication_file​ / AEEP_Djibouti_Country_market_brief_EN.pdf. Bill collection rate MED: 2013 value used for 2011 calculations for lack of data. Iraq: Bill collection rate World Bank. 2016a. “eC2: Electricity Services Restoration and Ministry of Electricity (MoE) Operations Efficiency.”  ToR for World Bank Assignment Title: 1223732—IRAQ, Netherlands for the World Bank, July 24. https://nl4worldbank.org/2016/07/14/ec2electricity-services​ -restoration-and-operations-efficiency. Number of Approximation for number of employees in the Ministry of employees Electricity. Lebanon: Electricité du Liban Number of new Estimated from WDI population growth figures for 2011. (EdL) customers Length of AUE (Arab Union of Electricity). 2013. Statistical Bulletin 2013. transmission Amman, Jordan: AUE. network Share of loss (%) WDI. Saudi Arabia: Saudi Electricity Bill collection rate Calculated from SEC document “Statistics 2000 to 2014.” See Company (SEC) appendix table C.4 for methodology. Tunisia: Bill collection rate Value obtained from STEG. Société Tunisienne de l’Electricité et du Gaz (STEG) West Bank: Energy purchased Calculated from electricity IEC sold to West Bank in 2013. Northern Electricity Distribution World Bank. 2014b . West Bank and Gaza: Assessment and Action Company (NEDCO) Plan to Improve Payment for Electricity Services in the Palestinian Territories: Study on Electricity Sector Contribution to Net Lending. Report No: ACS9393. Washington, DC: World Bank. http:// documents.worldbank.org/curated/en/120271468317065014​ /pdf/ACS93930WP0P1469990Box385388B00OUO090.pdf. Total electricity World Bank. 2014b. West Bank and Gaza—Assessment and Action billed Plan to Improve Payment for Electricity Services in the Palestinian Territories: Study on Electricity Sector Contribution to Net Lending. Report No: ACS9393. Washington, DC: World Bank. West Bank: Tubas District Energy purchased World Bank. 2014b. West Bank and Gaza—Assessment and Action Electricity Company (TUBAS) Plan to Improve Payment for Electricity Services in the Palestinian Territories: Study on Electricity Sector Contribution to Net Lending. Report No: ACS9393. Washington, DC: World Bank. Yemen, Rep.: Length of AUE (Arab Union of Electricity). 2013. Statistical Bulletin 2013. Public Electricity Corporation transmission Amman, Jordan: AUE. (PEC) network Bill collection rate World Bank. 2014a. “YEM Power Ministerial Note.” Unpublished paper, Washington, DC. January 3. Note: IEC = Israel Electric Corporation; MED = MENA Electricity Database; MENA = Middle East and North Africa; WDI = World Development Indicators. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 201 Table C.3  Descriptions and Assumptions of Economy-Level QFD Components Element Description and assumptions (Qe) Calculated by multiplying the electric power consumption per capita by the total population of End-user consumption the economy for the year 2013. (Te) Taken to be the average residential tariff for a consumption of 250 kWh/month for the year 2013. Average end-user tariff Values for all economies were calculated based upon the Arab Union of Electricity’s (2014), “Electricity Tariff in the Arab Countries.” In the case of Djibouti, calculations were based upon the official tariff document published by the economy. (Tc) Not readily available and had to be estimated using the LCOE. The LCOE unit cost of energy per Cost-recovery tariff rate technology type was obtained in $/kWh and then weighted according to the energy mix of each economy. Sources used were WDI for the energy mix information, and an LCOE modeling tool developed by ESMAPa for most of the LCOE values. Because the unit cost of fuel and renewables used in the modeling tool did not reflect the current state of energy sources in the MENA region, values from Lazard’s LCOE Analysis 2014 were used instead (see also appendix table C.5). These values do not consider the T&D contribution to the unit cost, and for this reason, a factor of ¢ 3.2/kWh was added to ensure that the T&D costs were considered in the calculations. (Lm) The technical loss rate is defined as the electric power transmission and distribution losses (% of Technical loss rate output) and was obtained from WDI database. WDI did not include data for West Bank (calculated alternatively as the average of the technical losses of West Bank distribution utilities in the MED) and Djibouti (value obtained as the grid losses from an online source).b (Ln) The choice of 5% was done so as to have values of Ln below the region’s best-performing Normative loss rate economies, namely Bahrain and Qatar with technical loss rates of 5.2% and 6.0%, respectively. (Rct) The bill collection rate indicates the income effectively collected during the year by the utility in Collection rate relation to the income billed. In the cases where a single utility existed (a VIU in the case of Algeria, for example), the collection rate of the economy was that of the utility. When more than one utility existed, the average value of the distribution utilities was used (in the case of Egypt, Arab Rep. for example). The collection rate was one of the most challenging indicators to collect from utilities in the MENA region, and when this was not possible, the methodology detailed in appendix C was used with the data presented in table C.19. (NC) This figure was easily obtained for economies with a single VIU. For economies with several Number of customers utilities, the presence of a regulator would allow for an aggregate official figure to be obtained (connections) from the regulator’s annual report. However, in the case of no regulator present, the sum of individual utility customers was calculated. (NE) The methodology used to obtain this figure was similar to that of the number of customers. The Number of employees number of FTE employees was used for all utilities, except in the case of Oman, where the number of total (direct and indirect) employees was used. This is because several utilities in Oman have a very low number of FTE whereas the number of outsourced (or indirect) employees is high. For example, in 2013, the indirect employees in Oman represented 67% of the total of 8,277 employees. table continues next page Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 202 Quasi-Fiscal Deficit: Hypothesis and Methodology Table C.3  Descriptions and Assumptions of Economy-Level Quasi-Fiscal Deficit Components (continued) Element Description and assumptions (CL) The cost of labor is defined as the annual cost of personnel directly employed by the utility and Cost of labor was sourced mainly from the financial statements of utilities. However, when this was not available, estimates were made to calculate a unit labor cost per employee, which was then multiplied by the number of employees present in the utilities for which labor cost data were not available. A calculated sum then allowed the economy-level aggregated estimate of the cost of labor to be obtained (see also methodology in appendix tables C.8 to C.16). In the case of the Republic of Yemen, where no labor cost data were available for the VIU (Public Electricity Corporation, PEC), an average unit cost of labor per employee was obtained from average earnings figures from the ILO (see also table C.17 and table C.18). (413) Customer per employee is an indicator of performance with values commonly above 500 in the Benchmark number of OECD economies.b The value of 413 used in this study was obtained using the same customers per benchmark value for the number of customers per employees in low-income countries as in employee in LICs the AICD methodology. Source: World Bank calculations, except where noted below. Note: A compilation of economic costs of more than 50 electricity generation and delivery technologies, META was rolled out to the World Bank Group and selected partners and clients in June 2012. Since then, META has been used in Dominica, the Arab Republic of Egypt, Kosovo, the former Yugoslav Republic of Macedonia, Morocco, and Vietnam as part of the World Bank’s engagement in these countries, and by consultants in Haiti and Jamaica. It can be downloaded here: http://esmap.org/META. AICD = Africa Infrastructure Country Diagnostic; ESMAP = Energy Sector Management Assistance Program; FTE = full-time equivalent; ILO = International Labour Organization; kWh = kilowatt-hours; LCOE = levelized cost of electricity; LICs = low-income countries; MED = MENA Electricity Database; MENA = Middle East and North Africa; META = Model for Electricity Technology Assessment; OECD = Organisation of Economic Co-operation and Development; T&D = transmission and distribution; VIU = vertically integrated utility WDI = World Development Indicators. a. ESMAP META Model. b. Eberhard and others 2011. We now provide the methodology used to estimate collection rates in Oman, Saudi Arabia, and Qatar, for which we did not have direct data from the MED. The bill collection rate is defined as the income effectively col- during the year in relation to the income billed, and is calculated using lected ­ ­equation C.1. Income effectively collected from customers for energy consumption and related service Bill collection rate = revenues related to energy consumption and service (C.1) When the collection rate was not available, it was calculated from the annual reports and financial statements of the utilities. In other words, the rate is the revenues collected divided by the billed amount. Since the annual reports do not provide a value for billed amounts, it was approximated as follows: 1. The income effectively collected is considered to be the figure of annual sales of, or annual revenues from, electricity in the financial statement. 2. The income not collected is considered as the receivables from customers, as stated in the financial report. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 203 3. The billed amount is therefore the sum of what was not collected (the receivables) and what was actually collected (the sales revenue reflected in the financial report). 4. The collection rate is therefore calculated using equation C.2 sales revenue The collection rate = (C.2) sales revenue + receivables from customers 5. If the economy has several utilities, steps 1–4 above were applied to each utility and the average of all utilities was taken to be the economy collection rate. This methodology was used to calculate the economy QFDs for Oman, Saudi Arabia, and Qatar, as shown in panels a, b, and c of table C.4. Table C.4  Data and Sources Used for Calculating Collection Rates Economy Oman Oman Oman a. Oman Utility name Muscat Electricity Majan Electricity Mazoon Electricity Distribution Company Company Distribution Company Source of data Annual report 2013 Annual report 2013 Annual report 2013 Amounts due from private 33,562,000 17,357,000 20,344,000 customers (RO) Amounts due from government 13,610,000 6,029,000 5,776,000 customers (RO) Electricity sales to private customers 98,814,000 79,265,000 67,567,000 (RO) Electricity sales to government 37,479,000 10,221,000 18,815,000 customers (RO) Collection rate (%) 74 79 77 b. Saudi Arabia Economy Saudi Arabia Utility name Saudi Electricity Company (SEC) Source of data SEC publication: electric data 2000–14 Receivables from customers and Saudi riyal (SRl) 18,452,000,000 revenues accrued net Total electricity sales SRl 32,878,000,000 Collection rate (%) 64 c. Qatar Economy Qatar Utility name KAHRAMAA Source of data KAHRAMAA Annual Report 2013 Accounts receivable Qatari riyal (QR) 585,434,000 Revenues from sale of electricity QR 1,553,741,000 Collection rate (%) 73 Source: World Bank calculations; RO: Omani Riyal. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 204 Quasi-Fiscal Deficit: Hypothesis and Methodology Methodology for Estimating the Economy-Level Cost Recovery Tariff and Collection Rates in Selected Economies Economy-Level QFD Cost-recovery tariffs were calculated using the basis of the economy fuel mix, and the levelized cost of electricity (LCOE) from different energy sources, as follows: Tc = Weighted LCOE = (LCOECoal × %Coal) + (LCOEHydro × %Hydro) + (LCOEN.gas × %N.gas) + (LCOEFuel × %Fuel) + (LCOERenewables × %Renewables) The shares of energy mix in each country used to compute the cost-recovery tariff are in table C.5. The LCOE values corresponding to each generation source are presented in table C.6. The Energy Sector Management Assistance Program’s (ESMAP’s) Model for Electricity Technology Assessment (META) considered 2010 as the base year; transmission and distribution (T&D) costs were not included and neither were environmental costs. To account for T&D losses, a value of US¢ 3.2 per kilowatt-hour (kWh) was added. Table C.5  Share of Energy Mixes Used in the Calculation of Tc (%) Economy Coal Hydro Natural gas Fuel Renewables Algeria 0 1 93 7 0 Bahrain 0 0 100 0 0 Djibouti 0 0 0 100 0 Egypt, Arab Rep. 0 8 77 15 1 Iraq 0 8 55 19 0 Jordan 0 0.3 25 74 0.1 Lebanon 0 7 0 93 0 Morocco 43 10 21 21 5 Oman 0 0 97 3 0 Qatar 0 0 100 0 0 Saudi Arabia 0 0 53 24 0 Tunisia 0 0.3 96 0.4 2 Yemen, Rep. 0 0 32 68 0 Israela 54 0 42 36 1 Source: WDI. a. in the case of West Bank, all electricity is imported from Israel, therefore the LCOE of Israel is used for Tc. Table C.6  LCOE Values Used to Calculate the Cost-Recovery Tariffs and Their Sources Generation type LCOE (US$ cents) /kWh Source Coal 7.44 ESMAP META Model Hydro 2.86 ESMAP META Model Natural gas 8.12 ESMAP META Model Fuel 31.45 Average Lazard Renewables 6.9 Average Lazarda Source: World Bank calculations based on ESMAP META Model and Lazard. 2014. Note: ESMAP = Energy Sector Management Assistance Program; LCOE = levelized cost of electricity; META = Model for Electricity Technology Assessment. a. Considering utility-sized photovoltaics (PV) and wind only. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 205 Utility-Level QFD: Calculating the Unit Historical Cost The unit historical cost, Tc, is composed of three main components that account for both capital expenditure (CAPEX) and operating expenses (OPEX). Equation C.3 below is used for calculating the unit historical cost made up of three annualized components: Infrastructure CAPEX + Connection Unit historical cost =  CAPEX + OPEX (C.3) Where: cost for power generation + cost for T & D Infrastructure CAPEX = 0.95 × ( kWh generated + kWh purchased ) CAPEX of T&D connection per customer × number of new customers Connection CAPEX = 0.95 × ( kWh generated + kWh purchased ) Total OPEX from statements OPEX = kWh billed Differences between the economy-level and the utility-level QFD values can be explained primarily in the calculation of the cost-recovery tariff, Tc. While the economy-level QFD does not take into consideration the energy purchased and imported, and only considers the energy generated within an economy, this is accounted for in the utility-level QFD and can be observed in the cases of Djibouti1 and the Republic of Yemen, for example. The effective tariff was approximated to that used previously in the economy-level QFD calculations, that is, for an average monthly consumption of 250 kWh in the residential sector. The Infrastructure CAPEX component of the Unit Historical Cost Using the CAPEX figures mentioned in the financial statements of utilities can be misleading in MENA. Since most utilities are public vertically integrated utilities (VIUs), the CAPEX is often obtained in the form of subsidies from the state and this is not always properly reflected in the financial statements. The annualized CAPEX related to infrastructure is made up of the costs related to power generation and those related to investments in T&D infrastructure. Calculating the Generation CAPEX The annualized CAPEX for generation was calculated depending upon the installed capacity of the plant, the technology type, and its economic life. This is shown in equation C.4. r Amortized capital cost = Installed capacity × CAPEX per kWh × 1 1− (1 + r ) T (C.4) Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 206 Quasi-Fiscal Deficit: Hypothesis and Methodology Table C.7  Components Used to Calculate the CAPEX According to Technology Type Amortization factor Technology CAPEX a per kW ($/kW) Economic life (T) r ÷ [ 1 − (1 + r)−T] Biomass 2,500 30 0.1062 Coal 2,403 30 0.1062 Co-Gen 917 30 0.1062 Diesel 1,070 30 0.1062 Gas CC 917 30 0.1062 Gas OC 603 30 0.1062 Geothermal 4,362 30 0.1062 HFO 1,250 30 0.1062 Hydropower 1,500 35 0.1037 Nuclear 4,102 60 0.1003 Solar 2,500 25 0.1102 Wind 2,000 25 0.1102 Source: Trimble and others 2016. Note: CAPEX = capital expenditure; CC = combined cycle; HFO = heavy fuel oil; kW = kilowatts; OC = open cycle. a. World Bank 2016b: 70. Data were used from table C.7, as well as from data on the installed capacities of the utilities involved. A discount rate (r) of 10 percent was used and the assumption was made that the hydro plants in MENA are big since the MENA region does not have a significant amount of hydro installed capacity. This avoided a degree of complexity in this calculation. Calculating the T&D CAPEX The annualized CAPEX for the T&D infrastructure depends upon the type of line voltage, its economic life, and the length of the network as shown in table C.8. This was calculated using a discount rate (r) of 12 percent and equation C.5: Amortized capital cost = Length of T & D network × r CAPEX per km × (C.5) 1 1− (1 + r )T Table C.8  Components Used to Calculate the CAPEX of the T&D Network Assumed CAPEX Economic life in Amortization factor ($/km) years (T) r ÷ [ 1 - (1 + r)-T] Lines 110 kV or above 165,000 50 0.1204 Lines below 110 kV down to 66 kV 65,000 40 0.1213 Lines below 66 kV down to 1 kV 10,000 30 0.1241 Source: World Bank 2016b: 71. Note: CAPEX = capital expenditure; km = kilometers; kV = kilovolts; T&D = transmission and distribution. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 207 The Customer Connection CAPEX component of the Unit Historical Cost The cost of connecting a customer is calculated using equation C.6: CAPEX of T&D connection per customer × Total number of new customers (C.6) 0.95 × ( kWh generated + kWh purchased ) The CAPEX related to T&D for connecting each customer is considered to be $100. Multiplying the electricity generated plus the energy the utility purchases, 0.95 corresponds to the ratio of electricity actually dispatched if the normative losses are considered to be 5 percent. The OPEX component of the Unit Historical Cost Annualized OPEX is expressed as a share of the electricity generated, as shown in equation C.7 below: Total OPEX from statememts (C.7) Electricity billed Estimating Labor Costs for the Arab Republic of Egypt, Djibouti, Jordan, Morocco, Oman, and the Republic of Yemen Labor costs were unavailable for several utilities and were obtained based on calculations making use of an estimated unit labor cost, as described for the countries listed below. Egypt Egypt has an unbundled electricity sector with a total of 12 utilities (including generation, distribution, and transmission) under the Egyptian Electricity Holding Company (EEHC). The cost of labor for all utilities except the Hydro Power Plants Electricity Production Company was available in the MENA database. However, the number of employees for all the utilities was not available. To calculate the total cost of labor, including that of the Hydro Power Plants Electricity Production Company and EEHC, a unit average cost per employee was calculated from the data for the utilities with labor costs and number of employees available. This unit cost was then multiplied by the total number of employees to obtain the value for the total labor cost for Egypt. The values used are found in table C.9. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 208 Quasi-Fiscal Deficit: Hypothesis and Methodology Table C.9  Values and Methodology Used in Calculating Labor Costs for the Arab Republic of Egypt Equation Description Value A Number of employees without the Hydro Power Plants Electricity 172,733 Production Company and without EEHC B Cost of labor in all utilities except the EEHC and Hydro Power Plants $1,359,678,577 Electricity Production Company C = B/A Unit cost of labor $7,872 D Number of employees in the EEHC 3,586 E Number of employees in the Hydro Power Plants Electricity 3,038 Production Company F = (D+E) × C Cost of employees in the EEHC and Hydro Power Plants Electricity $52,141,228 Production Company G = F+B Total estimated cost of labor including EEHC and Hydro Power Plants $1,411,819,806 Electricity Production Company Source: MENA Electricity Database and World Bank calculations. Note: EEHC = Egyptian Electricity Holding Company; MENA = Middle East and North Africa. Jordan For Jordan, several utilities had no data. Initially out of 10 utilities, the number of employees was available for 9, and labor costs for 8. Data from the report of the Jordanian regulator, the Energy and Minerals Regulatory Commission (EMRC), were used for the utility with the missing number of employees (Qatrana Electric Power Company, QEPC). Since the Amman Asia utility was not operational in the year of study (2013), it was neglected. As a result, there were nine utilities with nine employee numbers available, and seven with labor costs available. The same methodology as used in the case of Egypt was applied here to calculate the total cost of labor for the nine utilities in Jordan. The number of employees and labor costs values available are shown in table C.10. Table C.11 shows the total number of employees and the total labor costs for the 7 utilities for which data was available in Jordan. Table C.10  Utilities and Data Available for Jordanian Utilities No. employees Labor costs in $ 1 AES Levant Holding B.V. 47 Not available 2 Amman East Power Plant (AES) 51 3,248,314 3 Central Electricity Generating Company 1,037 18,788,759 4 Electricity Distribution Company 1,320 19,813,536 5 Irbid District Electricity Company 1,088 16,270,190 6 Jordan Electric Power Company 2,602 86,150,700 7 National Electric Power Company 1,373 22,166,850 8 Qatrana Electric Power Company 78 Not available 9 Samra Electric Power Generation Company 345 6,096,730 Source: MENA Electricity Database. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 209 Table C.11  Calculating the Unit Labor Cost for Jordan Description Value Total number of employees in 7 utilities 7,941 Total labor costs for 7 utilities $175,535,079 Unit labor cost calculated using equation C.6 $22,075 Source: World Bank calculations based on MENA Electricity Database. Note that total labor costs and total number of employees in equation C.8 do not include data for AES Levant Holding B.V and Qatrana Electric Power Generation Company. Total labor costs Unit labor cost = (C.8) Total number of employees Equation C.9 is used to calculate the labor costs of the two utilities with missing values and table C.12 shows the results obtained. Number of employees in Utility × Estimated labor cost =  Unit labor cost (C.9) The total cost of labor for Jordan is shown in table C.13 and was obtained using the equation C.10. Total labor costs for 7 utilities with data + Total labor cost =  labor costs of remaining two utilities (C.10) Table C.12  Calculating the Cost of Labor for the Two Utilities with Missing Values for Jordan Description Value ($) Estimated labor costs for Qatrana 1,721,819 Estimated labor costs for AES Levant 1,037,506 Source: World Bank calculations based on MENA Electricity Database. Table C.13  Calculating the Total Labor Costs for Jordan Description Value ($) Total cost of labor for 9 utilities 175,294,404 Source: World Bank calculations based on MENA Electricity Database. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 210 Quasi-Fiscal Deficit: Hypothesis and Methodology Morocco In the case of Morocco, the number of employees of all the utilities were available, but not the labor costs per utility. This is shown in table C.14 while table C.15 details how the unit labor cost was obtained. Table C.16 shows how the missing values were obtained using values from table C.14. Finally, the total cost of labor for all the utilities for Morocco is shown in table C.17. Table C.14  Utilities and Data Available for Moroccan Utilities No. employees (A) Labor costs in $ (B) 1 AMENDIS Tanger 401 25,306,122 2 AMENDIS Tetouan 468 25,772,595 3 LYDEC 1,432 92,912,657 4 ONEE 8,796 252,453,751 5 RADEEL 134 6 REDAL 511 44,702,600 7 Regie de Kenitra 196 8 Regie de Marrakech 370 8,355,024 9 Regie de Meknes 208 10 RADEEJ 188 4,131,731 11 Regie de Fes 439 12 Regie de Safi 118 Source: MENA Electricity Database. Table C.15  Calculating the Unit Labor Cost for Morocco Equation Description Value A12 Total number of employees available 13,261 C= ∑A1 B12 Total labor costs available $453,634,480 D= ∑B1 E = D/C Unit labor cost $34,208 Source: World Bank calculations based on MENA Electricity Database. Table C.16  Calculating the Cost of Labor for the Utilities with Missing Values for Morocco Equation Description Value F = E × (A5 + A9 + A11 + A12) Labor cost in remaining utilities $37,457,940 Source: World Bank calculations based on MENA Electricity Database. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 211 Table C.17  Calculating the Total Labor Costs for Morocco Equation Description Value G=D+F Total cost of labor for all utilities $491,092,421 Source: World Bank calculations based on MENA Electricity Database. Oman Omani utilities often have a larger number of outsourced employees than full- time employees. For consistency, it was decided to include the total number of employees in the labor cost estimates. Twelve utilities were found to have data for both the total number of employees and the labor costs. A unit cost of labor was calculated from these twelve utilities, which was equal to $19,742. The total number of employees for the 12 utilities was 5,085 and the total cost of labor obtained for these 12 utilities was $100,385,560. After obtaining an aggregate value of the total direct and indirect employees in 2013 from the Authority for Electricity Regulation (AER) annual report for 2014, it can be estimated that the remaining number of employees (8,277 – 5,085), is 3,192. (3,192 x 19,74) + $100,385,560 The total estimated labor costs in Oman =  = $163 million Republic of Yemen The cost of labor for the Public Electricity Corporation (PEC), the Yemeni public VIU, was unavailable. Data on the number of employees were obtained. An estimate of the cost of labor was done using average values from the International Labor Organization (ILO) for the Republic of Yemen. This is shown in table C.18. Using an exchange rate of $1 = 203.4 Yemeni riyals (corresponding to January 1, 2013), the values listed in table C.19 were obtained for the average unit annual cost of labor. Table C.18  Calculating Average Monthly Earning Based upon ILO Data for the Republic of Yemen Position Monthly salary in YRls (Yemeni Riyals) Managers 30,290 Clerical support workers 42,591 Technicians and associate professionals 69,439 Average monthly earning calculated 47,440 Source: World Bank calculations based on International Labor Organization. Note: ILO = International Labor Organization. Table C.19  Calculating the Cost of Labor for the Republic of Yemen Assuming salary paid for 12 months Average annual cost in U.S. dollars per employee $2,797 Number of employees in PEC 18,126 Total estimated salary bill in U.S. dollars (cost of labor) $50,706,483 Source: World Bank calculations based on MENA Electricity Database and ILO. Note: PEC = Public Electricity Corporation. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 212 Quasi-Fiscal Deficit: Hypothesis and Methodology Note 1. In the case of Djibouti, the electricity volume billed was assimilated to the energy generated + energy imported, since the figures obtained otherwise did not seem to reflect the fact that Djibouti imported an amount equivalent to 73 percent of the electricity generated in the economy. A factor of 0.86 was added to account for the 16 percent system losses in Djibouti. References AER (Authority for Electricity Regulation). 2013. Annual Report 2013. http://www.aer​ -oman.org/pdfs/Annual%20Report%202013%20-%20Eng.pdf Arab Union of Electricity. 2014. “Electricity Tariff in Arab Countries.” Eberhard, A., O. Rosnes, M. Shkaratan, H. Vennemo. 2011. Africa’s Power Infrastructure: Investment, Integration, Efficiency. Directions in Development; infrastructure. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle​ /0986/2290. EUEI (European Union Energy Initiative). 2013. Country Power Market Brief: Djibouti. Africa-EU Energy Partnership http://www.euei-pdf.org/sites/default/files/field​ _­publication_file/AEEP_Djibouti_Country_market_brief_EN.pdf. EEHC (Egyptian Electricity Holding Company). 2014. Annual Report 2013/2014. http:// www.moee.gov.eg/english_new/EEHC_Rep/REP-EN2013-2014.pdf. ESMAP (Energy Sector Management Assistance Program). ESMAP Meta Model. http:// esmap.org/META. Iraq Energy Institute. 2015. “Iraq’s Biggest Power Threat.” Iraq Energy Forum, August 6, 2015, Iraq Energy Institute, Baghdad. http://iraqenergy.org/home/articles_details​ .php?id=5. KAHRAMMA. Annual Report 2013. ———. Sustainability Report 2013. ———. Annual Report 2014. Lazard. 2014. “Lazard’s Levelized Costs of Energy Analysis, version 8.0.” https://www​ .lazard.com/media/1777/levelized_cost_of_energy_-_version_80.pdf. Lebanon Ministry of Environment and United Nations Development Programme (UNDP). http://climatechange.moe.gov.lb/viewfile.aspx?id=64. NEPCO (National Electric Power Company). 2013. Annual Report 2013. Jordan: NEPCO. SEC. “Statistics 2000 to 2014.” Trimble, C., M. Kojima, I. P. Arroyo, and F. Mohammadzadeh. 2016. “Financial Viability of Electricity Sectors in Sub-Saharan Africa: Quasi-Fiscal Deficits and Hidden Costs.” Policy Research Working Paper 7788, World Bank, Washington, DC. World Bank. 2013. World Development Indicators (database). World Bank, Washington, DC. https://data.worldbank.org/data-catalog/world-development-indicators. ———. 2014a. “YEM Power Ministerial Note.” Unpublished paper, World Bank, Washington, DC. January 3. ———. 2014b. West Bank and Gaza: Assessment and Action Plan to Improve Payment for Electricity Services in the Palestinian Territories: Study on Electricity Sector Contribution to Net Lending. Report No: ACS9393. Washington, DC: World Bank. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Quasi-Fiscal Deficit: Hypothesis and Methodology 213 http://­d ocuments.worldbank.org/curated/en/120271468317065014/pdf/ ACS93930WP0P​1469​990​Box385388B00OUO090.pdf. ———. 2016a. “eC2: Electricity Services Restoration and Operations Efficiency.” ToR for World Bank Assignment Title: 1223732—IRAQ, Netherlands for the World Bank, -and​ July 24. https://nl4worldbank.org/2016/07/14/ec2electricity-services-restoration​ -operations-efficiency. ———. 2016b. “Financial Viability of Electricity Sectors in Sub-Saharan Africa: Quasi- Fiscal Deficits and Hidden Costs.” Policy Research Working Paper WPS 7788, World Bank Group, Washington, DC. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 APPENDIX D Methodology for the Analysis of Drivers of Performance There are several equivalent approaches to testing the equality of the means of two subgroups of observations. This study uses regression on dummy variables because it is immediately generalizable to testing for the equality of three (or more) subgroup means and to testing equality for means of each subgroup when two or more subgroups are analyzed at the same time. The dummy variable approach defines a variable of interest (the benchmark indicator) whose observations can be ascribed to one of two subgroups using a 1/0 classification. For example, data were available for the load factor—denoted YL(i) for the i’th observation—for 23 utilities, of which six were vertically inte- grated utilities (VIUs) and 17 were distribution utilities (DUs). The null hypothesis is that there is no structure effect so that the means of the two groups are equal. The mean load factor for the first group was 0.567, and for the second it was 0.554. Rather than use the standard test statistic for the equality of two means based on these values and the estimated variance for the pooled sample, a regression approach can be used. Let D1(i) take the value 1 if the observation is from a VIU and zero if it is from a DU, and let D2(i) take the value 1 if it is from a distribution utility and value 0 if it is from a VIU. The regression model expresses the load factor in terms of the two dummy variables and an error term as shown in equation (D.1): YL(i) = β1 D1(i) + β2 D2(i) + u(i) (D.1) where the values of the β are to be estimated. This equation can be interpreted as saying that all values of the load factor for VIUs are equal to β1 + u(i), while for DUs they are equal to β2 + u(i). Assuming that the means of the error term are zero for both groups, estimates of the β values can be obtained by ordinary least squares and are denoted b1 and b2. It is important to note that this equation does not include a constant. Indeed, attempting to estimate an equation with a constant and two dummy variables that correspond to the two states of the Shedding Light on Electricity Utilities in the Middle East and North Africa   215   http://dx.doi.org/10.1596/978-1-4648-1182-1 216 Methodology for the Analysis of Drivers of Performance dichotomy under consideration would lead to exact singularity of the data matrix. The values obtained are b1 = 0.557 and b2 = 0.554, the same as obtained by simply finding means for the two sets of observations. The test for the equality of the two means is then equivalent to a test for the equality of the β in the regression model. A test of a linear restriction (β1 = β2) for a regression model is provided by Wald’s test based on an F statistic. For the load factor, the Wald test indicates that the probability of observing a difference between them at least as large as that estimated is 0.80. That is, for the load factor, there is an 80 percent chance of observing a difference between the subgroup mean load factors of 0.003 or greater. A probability of 5 percent is regarded as indicating a significant result that supports the alternative hypothesis that the group means are not equal. A 10 percent probability is regarded as being worthy of note, if not highly significant. The same results can be obtained through a different formulation of the dummy variable model. A dummy variable DC(i) is defined as taking the value 1 for all observations, while D2(i) is defined as before. The model is now written as in equation (D.2): YL(i) = γ1 DC(i) + γ2 D2(i) + u(i)(D.2) Noting that DC(i) is constant for all observations, equation (D.2) is equivalent to equation (D.3): YL(i) = γ1 + γ2 D2(i) + u(i)(D.3) which corresponds to a single variable regression model with a constant. The interpretation of this model is that observations for VIUs have a mean of γ1, while those for distribution utilities (DU) have a mean of γ1 + γ2. The coeffi- cient on the DU (γ2) is now the difference from the VIU. A standard t test for the hypothesis that this difference is zero is equivalent to a test of the equality of the two means. In the example above, g1 = 0.567 and g2 = −0.013, and the probability level for the t statistic on the difference coefficient g2 is 0.80. Certain factors in the study were categorized as falling into one of three classes (for example, big, medium, or small). The null hypothesis that the mean load factor is the same for all three groups is tested by constructing three dummy variables (D1 = 1 for big, = 0 for medium or small; D2 = 1 for medium, = 0 for big or small; D3 = 1 for small, = 0 for big or medium). From the regression of the load factor on these three variables the coefficients are equal to the subgroup means. Equality of the three means can be tested with a Wald test based on two linear restrictions (β1 = β2; β2 = β3). For testing the effect of more than one categorization on the indicator of inter- est, it is simplest to use the approach of equation (D.2). Consider the case of the load factor in which both structure and ownership are to be considered at the same time. There are four different combinations of states: publicly owned VIU, privately owned VIU, publicly owned DU, and privately owned DU. The model Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Methodology for the Analysis of Drivers of Performance 217 includes a constant factor, a dummy variable for the DU, and a dummy variable for privately owned DU. The three variables fully define all four states. A publicly owned VIU takes the value of the constant, while DUs (whether public or pri- vate) have an incremental effect given by the coefficient on the distribution dummy, and private utilities (whether VIU or DU) have an incremental effect given by the coefficient on the ownership dummy. The hypothesis test that both the structure effect and ownership effect are zero can be carried out by a Wald test (β1 = β2; γ1 = γ2). This approach can be easily generalized to the case where all five factors are included and where some factors (size, income) are catego- rized into three states. For indicators in which there are three states (size, income), the Wald test is carried out in two stages. First, the two restrictions β1 = β2 and β2 = β3 are simul- taneously tested, and then the pairwise restrictions β1 = β2, β1 = β3, and β2 = β3 are tested one at a time, so as to identify which variables (if any) have different means from the others. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 APPENDIX E Core Values for MENA Indicators This appendix provides the values used for the static analysis of this report, that is, year 2013 or where data were missing for that year, the most recent year for which data were available between 2009 and 2012. Table E.1 lists the abbrevia- tions and full names of indicators used in the following tables E.2, E.3, and E.4. These tables report values of the latest year available for MENA electricity utili- ties, for technical and operational indicators, financial indicators, and commercial indicators, respectively. Shedding Light on Electricity Utilities in the Middle East and North Africa   219   http://dx.doi.org/10.1596/978-1-4648-1182-1 220 Core Values for MENA Indicators Table E.1  Indicator Names and Their Abbreviations, as Used in Tables E.2–E.4 of This Appendix Indicator name Abbreviated name Technical and Load factor Load Factor operational Capacity factor Cap. Factor indicators Availability factor Av. Factor Transmission losses Tran. Losses Distribution losses Dis. Losses Technical losses Tech. Losses Nontechnical losses N.tech Losses Network maintenance Network maint. # of meters replaced/total # of meters Share of meters replaced Total OPEX/full-time equivalent (FTE) employee OPEX per emp. Total OPEX per connection OPEX per con. Total OPEX/kWh sold OPEX per kWh sold Total OPEX/km of network OPEX per km # of residential connections/FTE employee # of res. con. per emp. Energy sales ($)/FTE employee Energy sales per emp. Total revenues ($)/FTE employee Total rev. per emp. Financial indicators Share of cost of (fuel, lubricant, gas and coal) in total OPEX Share of cost of fuel in OPEX Share of (energy purchases and cost of fuel, lubricant, gas Share of energy purchased in OPEX and coal) in total OPEX Share of labor cost in total OPEX Labor cost in OPEX Energy sales/total OPEX Engy sales/OPEX Energy sales/total costs Engy sales/tot. costs (Accounts receivable/sales) × 365 Acc. rec./sales Debt/equity Debt/equity Current assets/current liabilities Assets/liab. ROA (return on assets) ROA ROE (return on equity) ROE Commercial Total energy volume sold (kWh)/connection Engy vol. sold per con. indicators Residential energy volume sold (kWh)/connection Res. engy vol. sold per con. Total billing ($)/connection Billing per con. Residential billing ($)/connection Res. billing per con. Collection rate Collection rate Share of installed meters (%) Share of installed meters SAIFI SAIFI SAIDI SAIDI CAIDI CAIDI Duration of interruption taken into consideration for system Duration of interruption interruptions affecting customers (including SAIDI, SAIFI, and CAIDI customer measures). Source: World Bank calculations. Note: CAIDI = Customer Average Interruption Duration Index; km = kilometer; kWh = kilowatt-hours; OPEX = operating expenses; SAIDI = System Average Interruption Duration Index; SAIFI = System Average Interruption Frequency Index. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 Table E.2a  Technical and Operational Indicators Technical and operational indicators System and operational efficiency Share of N.tech Network meters Load factor Cap. factor Av. factor Tran. losses Dis. losses Tech. losses losses maint. replaced Country or economy Utility type Utility (%) (%) (%) (%) (%) (%) (%) (%) (%) Algeria VIU SONELGAZ 50 29 19 10 11 Bahrain VIU EWA 52 96 6 Djibouti VIU EDD 0.4 Egypt, Arab Rep. DU AEDC 61 n.a. n.a. n.a. 11 7 4 GU CEPC n.a. 58 n.a. n.a. n.a. n.a. n.a. n.a. DU CEDC 38 n.a. n.a. n.a. 6 4 3 GU EDEPC n.a. 60 n.a. n.a. n.a. n.a. n.a. n.a. TU EETC n.a. n.a. n.a. n.a. n.a. n.a. DU EEDC 62 n.a. n.a. n.a. 10 5 5 GU MDEPC n.a. 65 n.a. n.a. n.a. n.a. n.a. n.a. DU MEEDC 69 n.a. n.a. n.a. 11 4 4 DU NCEDC 62 n.a. n.a. n.a. 10 5 4 DU NDEDC n.a. n.a. n.a. 9 7 4 DU SCEDC 64 n.a. n.a. n.a. 8 6 2 DU SDEDC 60 n.a. n.a. n.a. 10 DU UEEDC 68 n.a. n.a. n.a. 8 GU UEEPC n.a. 70 n.a. n.a. n.a. n.a. n.a. n.a. GU WDEPC n.a. 57 n.a. n.a. n.a. n.a. n.a. n.a. Iraq VIU MoE 37 Jordan GU AES Levant n.a. 28 99 n.a. n.a. n.a. n.a. n.a. n.a. GU AAEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU AES PSC n.a. 80 97 n.a. n.a. n.a. n.a. n.a. n.a. GU CEGCO n.a. 50 n.a. n.a. n.a. n.a. n.a. n.a. DU EDCO n.a. n.a. n.a. 12 221 table continues next page Table E.2a  Technical and Operational Indicators (continued) 222 Technical and operational indicators System and operational efficiency Share of N.tech Network meters Load factor Cap. factor Av. factor Tran. losses Dis. losses Tech. losses losses maint. replaced Country or economy Utility type Utility (%) (%) (%) (%) (%) (%) (%) (%) (%) DU IDECO n.a. n.a. n.a. 11 1 DU JEPCO 51 n.a. n.a. n.a. 14 11 3 0 TU NEPCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU QEPCO n.a. 75 97 n.a. n.a. n.a. n.a. n.a. n.a. GU SEPCO n.a. 50 n.a. n.a. n.a. n.a. n.a. n.a. Lebanon VIU EdL 33 15 17 Morocco DU AMENDIS n.a. n.a. n.a. 10 8 2 Tanger DU AMENDIS n.a. n.a. n.a. 11 9 2 Tetouan DU LYDEC n.a. n.a. n.a. 7 1 1 VIU ONEE 66 4 15 8 6 DU RADEEL n.a. n.a. n.a. 8 DU REDAL 56 n.a. n.a. n.a. 8 5 3 1 DU RAK n.a. n.a. n.a. 8 DU RADEEMA 51 n.a. n.a. n.a. 5 3 DU RADEM n.a. n.a. n.a. 7 DU RADEEJ 64 n.a. n.a. n.a. 4 2.7 DU RADEEF n.a. n.a. n.a. DU RADEES n.a. n.a. n.a. 3 Oman GU APBS n.a. 59 93 n.a. n.a. n.a. n.a. n.a. n.a. GU ABPC n.a. 41 96 n.a. n.a. n.a. n.a. n.a. n.a. GU ASPC n.a. 32 90 n.a. n.a. n.a. n.a. n.a. n.a. GU GPDCO n.a. 58 85 n.a. n.a. n.a. n.a. n.a. n.a. table continues next page Table E.2a  Technical and Operational Indicators (continued) Technical and operational indicators System and operational efficiency Share of N.tech Network meters Load factor Cap. factor Av. factor Tran. losses Dis. losses Tech. losses losses maint. replaced Country or economy Utility type Utility (%) (%) (%) (%) (%) (%) (%) (%) (%) GU AKPP n.a. 67 89 n.a. n.a. n.a. n.a. n.a. n.a. GU ARPP n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU BPDP n.a. n.a. n.a. n.a. n.a. n.a. n.a. VIU DPC 15 0 1.5 DU MJEC 71 n.a. n.a. n.a. 13 7 6 DU MZEC 44 n.a. n.a. n.a. 11 DU MEDC 55 n.a. n.a. n.a. 9 5 5 TU OETC n.a. n.a. n.a. 3 n.a. n.a. n.a. n.a. GU PPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. VIU RAECO 11 1 2.1 GU SSPWC n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU SPP n.a. 69 93 n.a. n.a. n.a. n.a. n.a. n.a. GU UPC n.a. 50 91 n.a. n.a. n.a. n.a. n.a. n.a. GU WAJPCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. Qatar VIU KAHRAMAA Saudi Arabia VIU SEC 57 5 0.1 Tunisia VIU STEG 61 37 2 14 8 7 0.4 West Bank DU JDECO n.a. n.a. n.a. 26 DU NEDCO 38 n.a. n.a. n.a. 13 9.1 DU TUBAS 28 n.a. n.a. n.a. 16 11 5 3 1 Yemen, Rep. VIU PEC 55 46 36 0 Note: DU = distribution utility; GU = generation utility; n.a. = not applicable; TU = transmission utility; VIU = vertically integrated utility. 223 Table E.2b  Technical and Operational Indicators 224 Technical and operational indicators Cost-efficiency (Total OPEX) Labor efficiency OPEX per OPEX per OPEX per # of res. con. Energy sales Total rev. per emp. con. kWh sold OPEX per km per emp. per emp. emp. Country or economy Utility type Utility $/emp. $/con. $/kWh $/km #con./emp. $/emp. $/emp. Algeria VIU SONELGAZ 31,050 304 0.05 7,730 Bahrain VIU EWA 0.08 Djibouti VIU EDD 79,469 1,612 0.43 75,869 44 121,124 121,124 Egypt, Arab Rep. DU AEDC 23,911 134 0.04 15,207 155 19,920 22,783 GU CEPC 138,270 n.a. n.a. n.a. n.a. n.a. n.a. DU CEDC 47,446 230 0.04 10,406 178 42,083 43,943 GU EDEPC 91,535 n.a. n.a. n.a. n.a. n.a. n.a. TU EETC n.a. n.a. n.a. n.a. n.a. DU EEDC 34,612 157 0.04 8,513 188 31,038 34,683 GU MDEPC 75,610 n.a. n.a. n.a. n.a. n.a. n.a. DU MEEDC 37,360 115 0.03 6,199 296 32,182 37,258 DU NCEDC 46,455 157 0.04 12,780 252 42,052 44,462 DU NDEDC 38,554 101 0.03 8,502 315 38,420 40,756 DU SCEDC 46,892 169 0.04 14,788 233 43,406 45,113 DU SDEDC 27,665 75 0.03 9,923 319 28,216 29,709 DU UEEDC 37,974 119 0.03 6,373 287 33,131 38,279 GU UEEPC 178,678 n.a. n.a. n.a. n.a. n.a. n.a. GU WDEPC 77,375 n.a. n.a. n.a. n.a. n.a. n.a. Iraq VIU MOE 820 0.07 30,885 Jordan GU AES Levant n.a. n.a. n.a. n.a. n.a. n.a. GU AAEPC 311,051 n.a. n.a. n.a. n.a. n.a. n.a. GU AES PSC n.a. n.a. n.a. n.a. n.a. n.a. GU CEGCO n.a. n.a. n.a. n.a. n.a. n.a. DU EDCO 230,005 26,497 126 223,713 228,204 DU IDECO 196,963 547 0.10 11,900 310 211,193 table continues next page Table E.2b  Technical and Operational Indicators (continued) Technical and operational indicators Cost-efficiency (Total OPEX) Labor efficiency OPEX per OPEX per OPEX per # of res. con. Energy sales Total rev. per emp. con. kWh sold OPEX per km per emp. per emp. emp. Country or economy Utility type Utility $/emp. $/con. $/kWh $/km #con./emp. $/emp. $/emp. DU JEPCO 447,628 0.14 43,084 364 416,395 423,073 TU NEPCO n.a. n.a. n.a. n.a. n.a. GU QEPCO n.a. n.a. n.a. n.a. n.a. n.a. GU SEPCO n.a. n.a. n.a. n.a. n.a. n.a. Lebanon VIU EdL 573,990 1,575 0.29 155,298 162,865 Morocco DU AMENDIS 321,620 508 0.12 35,835 355,728 Tanger DU AMENDIS 151,086 346 0.15 30,610 142,100 Tetouan DU LYDEC 527,093 836 0.20 96,152 527,759 565,777 VIU ONEE 283,710 510 0.09 9,362 333,564 347,685 DU RADEEL 202,559 361 0.12 18,849 174,709 221,208 DU REDAL 642,046 644 0.17 49,382 969 663,010 736,208 DU RAK 247,581 412 0.12 19,758 231,793 305,496 DU RADEEMA 287,299 410 0.10 31,932 372,131 DU RADEM 253,858 309 0.11 21,096 246,725 305,224 DU RADEEJ 190,092 396 0.10 21,285 403 258,237 258,791 DU RADEEF 185,909 318 0.11 37,114 182,504 230,975 DU RADEES 199,577 339 0.13 32,438 177,773 454,427 Oman GU APBS n.a. n.a. n.a. n.a. n.a. n.a. GU ABPC 675,865 n.a. n.a. n.a. n.a. n.a. n.a. GU ASPC 816,078 n.a. n.a. n.a. n.a. n.a. n.a. GU GPDCO 304,725 n.a. n.a. n.a. n.a. n.a. n.a. table continues next page 225 226 Table E.2b  Technical and Operational Indicators (continued) Technical and operational indicators Cost-efficiency (Total OPEX) Labor efficiency OPEX per OPEX per OPEX per # of res. con. Energy sales Total rev. per emp. con. kWh sold OPEX per km per emp. per emp. emp. Country or economy Utility type Utility $/emp. $/con. $/kWh $/km #con./emp. $/emp. $/emp. GU AKPP n.a. n.a. n.a. n.a. n.a. n.a. GU ARPP n.a. n.a. n.a. n.a. n.a. n.a. GU BPDP n.a. n.a. n.a. n.a. n.a. n.a. VIU DPC 328,856 1,438 0.05 19,460 173 265,289 370,361 DU MJEC 226,672 0.05 92 157,010 273,953 DU MZEC 174,580 1,150 14,107 115 107,185 227,820 DU MEDC 1,698 42,246 399 TU OETC 58,413 n.a. n.a. 8,990 n.a. n.a. n.a. GU PPC n.a. n.a. n.a. n.a. n.a. n.a. VIU RAECO 89,332 4,917 0.21 30,847 13 23,179 107,349 GU SSPWC 614,812 n.a. n.a. n.a. n.a. n.a. n.a. GU SPP 942,235 n.a. n.a. n.a. n.a. n.a. n.a. GU UPC 288,524 n.a. n.a. n.a. n.a. n.a. n.a. GU WAJPCO n.a. n.a. n.a. n.a. n.a. n.a. Qatar VIU KAHRAMAA 1,519 Saudi Arabia VIU SEC 278,984 1,237 0.03 16,992 179 276,918 300,451 Tunisia VIU STEG 258,487 948 20,409 142,249 142,249 West Bank DU JDECO 0.19 DU NEDCO 125,720 684 0.16 15,350 147 135,888 144,991 DU TUBAS 759 0.10 22,697 73 79,724 56,434 Yemen, Rep. VIU PEC 16,590 158 0.06 16,712 90 18,630 Source: MENA Electricity Database. Note: DU = distribution utility; GU = generation utility; km = kilometer; kWh = kilowatt-hours; n.a. = not applicable; OPEX = operation expenses; TU = transmission utility; VIU = vertically integrated utility. Table E.3  Financial Indicators Financial indicators Cost structure Cost-recovery Balance sheet Profitability Share of Share of cost of energy Engy Engy fuel in purchased Labor cost sales/ sales/Tot. Acc. rec./ Debt/ Assets/ Utility OPEX in OPEX in OPEX OPEX costs sales equity liab. ROA ROE Country or economy type Utility % % % % % days % % % % Algeria VIU SONELGAZ 92 56 428 146 −1.74 −7 Bahrain VIU EWA 67 5 37 205 67 84 0.88 1 Djibouti VIU EDD 22 56 17 152 110 192 222 274 Egypt, Arab Rep. DU AEDC n.a. n.a. 83 81 77 0.18 0.26 GU CEPC 88 n.a. 8 n.a. n.a. n.a. 5 0.0 1 DU CEDC n.a. n.a. 20 89 83 685 66 2 8 GU EDEPC 88 n.a. 10 n.a. n.a. n.a. 3,484 37 0.05 0.3 TU EETC n.a. n.a. n.a. n.a. n.a. 53 DU EEDC n.a. n.a. 26 90 80 175 527 103 0.04 0.12 GU MDEPC 79 n.a. 12 n.a. n.a. n.a. 2,509 68 0.03 0.41 DU MEEDC n.a. n.a. 27 86 77 115 501 85 0.06 0.14 DU NCEDC n.a. n.a. 21 91 87 183 850 71 0.19 0.61 DU NDEDC n.a. n.a. 24 90 677 97 0.30 0.83 DU SCEDC n.a. n.a. 21 93 87 276 1,282 81 2.6 8.77 DU SDEDC n.a. n.a. 35 523 103 0.23 0.46 DU UEEDC n.a. n.a. 26 87 75 178 571 113 0.06 0.17 GU UEEPC 93 n.a. 5 n.a. n.a. n.a. 1,270 56 0.35 3.02 GU WDEPC 81 n.a. 15 n.a. n.a. n.a. 3,074 67 0.01 0.11 Iraq VIU MOE 30 84 Jordan GU AES Levant n.a. n.a. n.a. n.a. GU AAEPC 63 n.a. n.a. n.a. n.a. 290 123 GU AES PSC n.a. n.a. n.a. n.a. 333 287 36 GU CEGCO n.a. n.a. n.a. n.a. 354 95 12 21 227 table continues next page Table E.3  Financial Indicators (continued) 228 Financial indicators Cost structure Cost-recovery Balance sheet Profitability Share of Share of cost of energy Engy Engy fuel in purchased Labor cost sales/ sales/Tot. Acc. rec./ Debt/ Assets/ Utility OPEX in OPEX in OPEX OPEX costs sales equity liab. ROA ROE Country or economy type Utility % % % % % days % % % % DU EDCO n.a. n.a. 6 97 117 1,476 99 5 16 DU IDECO n.a. n.a. 7 107 99 120 981 84 6 20 DU JEPCO n.a. n.a. 7 93 122 576 80 25 12 TU NEPCO n.a. n.a. n.a. n.a. n.a. 126 GU QEPCO n.a. n.a. n.a. n.a. 621 488 5 25 GU SEPCO n.a. n.a. n.a. n.a. 876 113 4 17 Lebanon VIU EdL 82 88 5 27 27 15 −150 Morocco DU AMENDIS Tanger n.a. n.a. 3 3 DU AMENDIS Tetouan n.a. n.a. −1 −2 DU LYDEC n.a. n.a. 12 100 89 76 279 72 18 VIU ONEE 81 10 118 87 159 3,327 63 −4 −127 DU RADEEL n.a. n.a. 86 6 7 DU REDAL n.a. n.a. 14 103 92 121 92 2 10 DU RAK n.a. n.a. 94 DU RADEEMA n.a. n.a. 8 130 205 41 DU RADEM n.a. n.a. 97 21 22 DU RADEEJ n.a. n.a. 12 136 119 106 66 64 DU RADEEF n.a. n.a. 98 DU RADEES n.a. n.a. 89 14 16 Oman GU APBS n.a. n.a. n.a. n.a. 249 121 8 24 GU ABPC 59 n.a. n.a. n.a. n.a. 303 54 GU ASPC 61 n.a. n.a. n.a. n.a. 294 53 GU GPDCO 75 n.a. 13 n.a. n.a. n.a. 443 1 0.2 table continues next page Table E.3  Financial Indicators (continued) Financial indicators Cost structure Cost-recovery Balance sheet Profitability Share of Share of cost of energy Engy Engy fuel in purchased Labor cost sales/ sales/Tot. Acc. rec./ Debt/ Assets/ Utility OPEX in OPEX in OPEX OPEX costs sales equity liab. ROA ROE Country or economy type Utility % % % % % days % % % % GU AKPP 78 n.a. n.a. n.a. n.a. 94 79 9 15 GU ARPP 77 n.a. n.a. n.a. n.a. 156 GU BPDP 52 n.a. n.a. n.a. n.a. 1,857 42 3 VIU DPC 6 81 72 263 315 46 DU MJEC n.a. n.a. 6 69 119 109 43 8 14 DU MZEC n.a. n.a. 61 110 148 18 6 14 DU MEDC n.a. n.a. 5 80 122 147 46 8 16 TU OETC n.a. n.a. n.a. n.a. n.a. 192 7 20 GU PPC n.a. n.a. n.a. n.a. 0.1 VIU RAECO 51 53 15 365 316 128 3 11 GU SSPWC 51 n.a. n.a. n.a. n.a. 357 179 3 13 GU SPP 68 n.a. n.a. n.a. n.a. 1,399 118 3 GU UPC n.a. n.a. n.a. n.a. 72 38 5 7 GU WAJPCO 47 n.a. 25 n.a. n.a. n.a. 504 8 2 Qatar VIU KAHRAMAA 48 12 97 138 74 214 Saudi Arabia VIU SEC 14 99 39 205 392 86 2 5 Tunisia VIU STEG 82 88 6 55 47 99 596 89 −4 −22 West Bank DU JDECO n.a. n.a. 10 260 126 −20 −19 DU NEDCO n.a. n.a. 7 108 103 166 86 275 3 4 DU TUBAS n.a. n.a. 276 7 Yemen, Rep. VIU PEC 74 100 75 Source: MENA Electricity Database. Note: DU = distribution utility; GU = generation utility; n.a. = not applicable; OPEX = operation expenses; ROA = return on assets; ROE = return on equity; TU = transmission utility; VIU = vertically integrated utility. 229 Table E.4  Commercial Indicators 230 Commercial indicators Average consumption and billing Metering Customer management and service quality Engy vol. Res. Engy Share of sold per vol. sold Billing per Res. billing Collection installed con. per con. con. per con. rate meters SAIFI SAIDI CAIDI Duration of Country or economy Utility type Utility kWh/con. kWh/con. $/con. $/con. % % 000 minutes minutes interruption Algeria VIU SONELGAZ 5,814 Bahrain VIU EWA 97 Djibouti VIU EDD 3,713 2,997 37 Egypt, Arab Rep. DU AEDC 3,658 2,200 111 41 99 100 0.12 2.86 24.47 GU CEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. DU CEDC 5,862 2,194 197 41 94 100 0.17 4.59 27.12 GU EDEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. TU EETC n.a. n.a. n.a. n.a. n.a. n.a. DU EEDC 4,392 1,851 132 33 95 99 2.24 186.02 83.17 GU MDEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. DU MEEDC 3,746 2,686 96 92 0.50 58.30 117.76 DU NCEDC 4,340 2,690 138 59 93 0.53 21.54 40.42 DU NDEDC 3,133 2,323 97 50 84 0.93 17.26 18.48 DU SCEDC 4,584 2,633 148 60 86 100 2.15 48.57 DU SDEDC 2,438 1,839 68 33 93 100 DU UEEDC 3,570 2,275 101 36 88 100 0.17 18.11 106.93 GU UEEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU WDEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Iraq VIU MOE 6,341 182 Jordan GU AES Levant n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU AAEPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU AES PSC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU CEGCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. DU EDCO 6,429 3,487 335 12.71 table continues next page Table E.4  Commercial Indicators (continued) Commercial indicators Average consumption and billing Metering Customer management and service quality Engy vol. Res. Engy Share of sold per vol. sold Billing per Res. billing Collection installed con. per con. con. per con. rate meters SAIFI SAIDI CAIDI Duration of Country or economy Utility type Utility kWh/con. kWh/con. $/con. $/con. % % 000 minutes minutes interruption DU IDECO 5,591 3,054 586 247 100 3.84 DU JEPCO 7,437 3,638 356 97 2.11 2.81 TU NEPCO n.a. n.a. n.a. n.a. n.a. n.a. GU QEPCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU SEPCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Lebanon VIU EdL 5,386 529 Morocco DU AMENDIS 4,312 473 Tanger DU AMENDIS 2,292 299 Tetouan DU LYDEC 4,223 520 102 0.92 13.36 1 VIU ONEE 5,634 1,103 190 102 100 3.69 3.70 1.00 DU RADEEL 2,953 DU REDAL 3,759 1,773 442 186 DU RAK 3,532 306 DU RADEEMA 4,047 466 1.15 2 DU RADEM 2,750 301 DU RADEEJ 4,048 1,615 436 168 1 table continues next page 231 232 Table E.4  Commercial Indicators (continued) Commercial indicators Average consumption and billing Metering Customer management and service quality Engy vol. Res. Engy Share of sold per vol. sold Billing per Res. billing Collection installed con. per con. con. per con. rate meters SAIFI SAIDI CAIDI Duration of Country or economy Utility type Utility kWh/con. kWh/con. $/con. $/con. % % 000 minutes minutes interruption DU RADEEF 2,814 312 DU RADEES 2,621 302 Oman GU APBS n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU ABPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU ASPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU GPDCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU AKPP n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU ARPP n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU BPDP n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. VIU DPC 27,586 13,630 400 DU MJEC 521 3 DU MZEC 432 DU MEDC 582 15 3 TU OETC n.a. n.a. n.a. n.a. n.a. n.a. GU PPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. VIU RAECO 23,011 925 459 71 1.75 1.88 GU SSPWC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU SPP n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU UPC n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. GU WAJPCO n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Qatar VIU KAHRAMAA table continues next page Table E.4  Commercial Indicators (continued) Commercial indicators Average consumption and billing Metering Customer management and service quality Engy vol. Res. Engy Share of sold per vol. sold Billing per Res. billing Collection installed con. per con. con. per con. rate meters SAIFI SAIDI CAIDI Duration of Country or economy Utility type Utility kWh/con. kWh/con. $/con. $/con. % % 000 minutes minutes interruption Saudi Arabia VIU SEC 35,937 22,154 100 4.09 Tunisia VIU STEG 3,749 377 125 West Bank DU JDECO 5,988 3,826 97 616 DU NEDCO 4,307 2,476 504 90 100 DU TUBAS 7,330 3,427 426 62 100 Yemen, Rep. VIU PEC 2,631 1,922 178 100 Source: MENA Electricity Database. Note: CAIDI = Customer Average Interruption Duration Index; DU = distribution utility; GU = generation utility; kWh = kilowatt-hours; SAIDI = System Average Interruption Duration index; SAIFI = System Average Interruption Frequency Index; TU = transmission utility; VIU = vertically integrated utility. 233 Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled ­ ­ content. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. Shedding Light on Electricity Utilities in the Middle East and North Africa http://dx.doi.org/10.1596/978-1-4648-1182-1 The electricity sector in the Middle East and North Africa (MENA) is in the grip of an apparent paradox. The region holds the world’s largest oil and gas reserves and has been able to universalize access to electricity in most economies, but it may not be able to make the investments required to meet the future power needs of its fast-growing populations. The annual investments to keep pace with the demand for electricity have been estimated at about 3 percent of the region’s projected GDP. In most of the region’s economies, however, financial constraints limit the ability to make those investments. The power sector needs to find its own financing sources—and quickly. Shedding Light on Electricity Utilities in the Middle East and North Africa provides quantitative evidence on how better utility management; sustainable pricing; and selective, context-specific reforms would free enough resources to make the needed investments and lower the operating costs of the sector. The solution involves cutting costs and raising revenues through well-targeted and well-identified improvements. These improvements would generate more financing than the sector’s investment needs. The report provides detailed evidence of the size of the potential gain in each of the 14 MENA economies covered. The analysis is based on the MENA Electricity Database, a new dataset covering 67 electricity utilities, as well as a sample of utilities in comparable economies from other regions. The authors hope that their benchmarking efforts will provide a regional- and utility-level frame of reference for sector performance in the region. The book will be of interest to managers of electricity utilities, regulators, policy makers, and other stakeholders concerned with the performance of utilities in the region. ISBN 978-1-4648-1182-1 SKU 211182