COOKING WITH ELECTRICITY A COST PERSPECTIVE © 2020 September | 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 This work is a product of the staff of the World Bank and the Modern Energy Cooking Services (MECS) program. The findings, interpretations, 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. The MECS program is funded by UK aid from the UK government; however, the views expressed do not necessarily reflect the UK government’s official policies. 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: Energy Sector Management Assistance Program (ESMAP). 2020. Cooking with Electricity: A Cost Perspective. Washington, DC: World Bank. 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 contained 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. Production Credits Editor | Barbara Karni Production Editor | Heather Austin, The World Bank Designer | Will Kemp, The World Bank Images | Jon Leary @ Loughborough University and Gamos (p. x, p. 4, p. 6, p. 12, p. 14, p. 18, p. 20, p. 24, p. 27, p. 35 [Figure 3.7], p. 36, p. 37, p. 39, p. 41 [Figure 3.10], p. 42, p. 43, p. 45, p. 47, p. 51 [Figure 3.16], p. 52 [Figure 3.17, Figure 3.18], p. 53, p. 55, p. 66, p. 69 [Figure 3.28, Figure 3.29], p. 70, p. 73, p. 74, p. 81, p. 82, p. 87, p. 88, p. 89, p. 94, p. 123), Shima Sago @ TaTEDO (p. 57 [Figure 3.21], p. 61, p. 67, p. 83, p. 92, Hannah Blair @ CLASP (cover, p. 7, p. 63), Mercy Kamau @ SCODE (p. 68 [Figure 3.26], p. 69 [Figure 3.27, Figure 3.28]), Jacob Fodio Todd @ University of Sussex (p. 95, p. 110). All images remain the sole property of their source and may not be used for any purpose without written permission from the source. ii ESMAP  |  Cooking with Electricity: A Cost Perspective CONTENTS Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Report Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv 1 | BACKGROUND. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1. State of Access to Clean Cooking and Electricity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2. The Burden of Cooking with Biomass Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3. Objective and Scope of the Report. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4. The Case for Cooking with Electricity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.5. A New Generation of Highly Efficient eCooking Appliances. . . . . . . . . . . . . . . . . . . . . . . . . 9 1.6. Electrical Infrastructure in Sub-Saharan Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2 | METHODOLOGY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.1. Techno-Economic Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2. System Architectures for eCooking in Strong-Grid, Weak-Grid, and Off-Grid Contexts. . 15 2.3. Cost Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.4. Summary of Contexts, Systems, and Fuel Prices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.5. Demand for Electricity for Cooking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.6. Business Models and Financing Horizons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 3 | ECOOKING IN GRID-CONNECTED AND OFF-GRID SYSTEMS: MODELLING RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.1. Overview of Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.2. eCooking on National Grids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Case Study 1: Building on the Success of LPG to Displace Charcoal in Urban East African Kitchens with a Clean Fuel Stack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Case Study 2: Tackling Load Shedding in Lusaka, Zambia, by Time Shifting and Reducing Electricity Demand for Cooking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 3.3. eCooking on Mini Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Case Study 3: Enabling 24-Hour eCooking on Micro-Hydro Mini Grids in Myanmar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Case Study 4: Exploring the Range of Opportunities for eCooking on Solar Hybrid Mini Grids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.4. eCooking with Stand-alone Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Case Study 5: The Next Generation of Cooking-Enabled Solar Home Systems. . . 68 3.5. Implications for eCooking in Off-Grid and Grid‑Connected Contexts . . . . . . . . . . . . . . . . 74  4 | DELIVERY APPROACHES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.1. Appliance Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 4.2. Peer-to-Peer Women-Led Product Distribution Models. . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.3. Pay-as-You-Go Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 4.4. Productive Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 4.5. Utility Model: Cooking as a Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.6. Distribution through Consumer Lending Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 5 | FINANCING THE TRANSITION TO ECOOKING. . . . . . . . . . . . . . . . . . . . . . . . . . . 89 5.1. Consolidating Investment Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 5.2. Financing the Cost of eCooking for Households. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.3. Financing Developers’ Capital Expenses and Working Capital. . . . . . . . . . . . . . . . . . . . . . 91 5.4. Results-Based Financing and Impact-Linked Financing. . . . . . . . . . . . . . . . . . . . . . . . . . . 93 6 | DISCUSSION, RECOMMENDATIONS, AND AREAS FOR FURTHER RESEARCH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 6.1. Support Policy Makers’ Efforts to Create an Enabling Environment that Bridges the Division between the Electrification and Clean Cooking Sectors. . . . . . . . . . . . . . . . . . . 95 6.2. Conduct Strategic Evidence-Based Research to Inform Decision Makers, Private Sector Players, and Consumers of Emerging Opportunities . . . . . . . . . . . . . . . . . . . . . . . 99 6.3. Support Private Sector Efforts to Develop Products and Services Tailored to the Needs and Aspirations of the Poor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 6.4. Help Consumers Understand the Benefits of Adopting Modern eCooking Solutions and Reduce Barriers to Behavioral Change. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 7 | CONCLUSION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Appendix A: The Modern Energy Cooking Solutions Program. . . . . . . . . . . . . . . . . . . 119 Appendix B: Typology of eCooking System Architectures. . . . . . . . . . . . . . . . . . . . . . . 121 Appendix C: Assessing Electricity Demand for Cooking . . . . . . . . . . . . . . . . . . . . . . . . 122 Appendix D: Comparison of eCooking Appliances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Appendix E: Outline of the eCooking Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Appendix F: Model Input Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Appendix G: Solar eCooking Cross-Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Appendix H: Multi-Tier Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 iv ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURES Figure ES.2 Comparison of system architectures using aggregated data from all case studies . . . . . . . . xxi Impact of energy-efficient appliances and fuel stacking on cost of AC and Figure ES.3  battery-supported DC eCooking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii Figure 1.1 Actual and projected global access to electricity and clean cooking, 2000–30 . . . . . . . . . . . . 2 Figure 1.2 Share of population with access to clean cooking fuels and technologies, by region, 2017 . . . . 2 Figure 1.3 Access to electricity and clean cooking in Zambia, Myanmar, and Kenya. . . . . . . . . . . . . . . . . 8 Figure 1.4 Assessment of eCooking appliances featured in this report . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 2.1 Actual and projected prices for PV modules and lithium-ion battery storage, 2010–22. . . . . . 16 Figure 3.1 Comparison of the five case studies and rationale for selection . . . . . . . . . . . . . . . . . . . . . 26 Figure 3.2 Percentage of households cooking primarily with electricity in Sub-Saharan African  and South/Southeast Asian countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 3.3 Sensitivity analysis comparing the cost of eCooking with the cost of cooking with charcoal  across Sub-Saharan Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure 3.4  ercentage of households with grid connections that still cook primarily with fuels other P than electricity in Sub-Saharan Africa and South/Southeast Asia. . . . . . . . . . . . . . . . . . . . . 33 Figure 3.5 Percentage of households cooking primarily with commercialized polluting fuels and  technologies (charcoal, coal, or kerosene) in Sub-Saharan Africa and South/Southeast Asia . 33 Figure 3.6 Primary cooking fuel used in selected countries in East and Southern Africa . . . . . . . . . . . . 34 Figure 3.7  uel stacking using LPG for manual control and an electric pressure cooker for F automatic control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Figure 3.8 Monthly cost of cooking with main fuels in Nairobi, 2020 and 2025. . . . . . . . . . . . . . . . . . . 38 Figure 3.9 Sensitivity of modelling results to charcoal price in Nairobi, Dar es Salaam, and Kampala, 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Charcoal market in Lusaka, Zambia, alongside electricity distribution infrastructure. . . . . . . . 41 Figure 3.10  Figure 3.11 Comparison of mbaula, hot plate, and electric pressure cooker. . . . . . . . . . . . . . . . . . . . . . 42 Figure 3.12 Monthly cost of cooking using main fuels in Lusaka, Zambia, 2020 and 2025. . . . . . . . . . . . 44 Figure 3.13 Sensitivity of modelling results to potential tariff increases by ZESCO, 2020 and 2025. . . . . 46  reak-even tariffs for typical solar hybrid mini grid in India at different levels of energy Figure 3.14 B consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Figure 3.15 Effect of increasing load factor on levelized cost of electricity of power-limited mini grids. . . 50 Powerhouse at one of the many small community-owned micro-hydro systems in Figure 3.16  Shan State, Myanmar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Figure 3.17 Voltage stabiliser in Myanmar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Voltmeter installed in kitchen in Myanmar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Figure 3.18  Figure 3.19 Monthly cost of cooking using main fuels in Shan State, Myanmar, 2020 and 2025. . . . . . . . 54 Figure 3.20 Sensitivity of modelling results to mini grid tariffs in Myanmar, 2020 and 2025 . . . . . . . . . . 56 Kibindu village residents experimenting with range of efficient eCooking appliances Figure 3.21  during a focus group session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure 3.22 Monthly cost of cooking using main fuels in Kibindu, Tanzania, 2020 and 2025 . . . . . . . . . . 59 Figure 3.23 Break-even analysis for mini grid tariffs for household cooking, 2020 and 2025. . . . . . . . . . 60 Figure 3.24 Break-even analysis for mini grids tariffs for microenterprise cooking, 2020 and 2025. . . . . 62 Figure 3.25 Monthly cost of cooking with different fuel options projected by various models . . . . . . . . . 65 Early prototype of a battery-supported DC electric pressure cooker designed by SCODE . . . 68 Figure 3.26   v Participatory cooking session with prototype of DC electric pressure cooker in Echariria,  Figure 3.27  Kenya. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Residents of Echariria, Kenya at a community meeting with a DC electric pressure cooker . . 69 Figure 3.28  Charcoal stove and battery that is regularly charged at Echariria’s solar hub . . . . . . . . . . . . 69 Figure 3.29  Figure 3.30 Monthly cost of cooking using main fuels in Echariria, Kenya, 2020 and 2025 . . . . . . . . . . . . 71  ensitivity of solar battery–eCooking and fuel-stacking scenarios to charcoal price with a Figure 3.31 S five-year repayment horizon, 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72  reakdown of solar eCooking and fuel costs for systems sized to meet needs of average Figure 3.32 B Kenyan household in 2025. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure 3.33 Emerging opportunities for cost-effective eCooking identified in each of the five case studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75  ptimal-system diagrams for household cooking, based on electricity/charcoal price Figure 3.34 O combination and quality of the grid, 2025. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76  ptimal-system diagram for productive-use case (precooking beans/cereals with an Figure 3.35 O electric pressure cooker) on reliable grid, 2025. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Figure 3.36 Comparison of system architectures using aggregated data from all case studies . . . . . . . . 79 Impact of energy-efficient appliances and fuel stacking on cost of AC and battery- Figure 3.37  supported DC eCooking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Figure 5.1 Market financing of electric cooking appliances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Figure 5.2 Range of appliance financing options for utilities and mini grid developers . . . . . . . . . . . . . . 91 Figure C.3 Enumerator training study participant to record cooking diary data in Nairobi . . . . . . . . . . . 125 TABLES Table 1.1 Types of electric cooking physically possible with each tier of electricity access. . . . . . . . . . . 7 Table 2.1 Summary of data collection methodologies used in this report. . . . . . . . . . . . . . . . . . . . . . . 14 Table 2.2 Simplified typology of eCooking devices for strong, weak, and off-grid settings. . . . . . . . . . . 15 Table 2.3 Parameter values used in high- and low-cost scenarios for eCooking systems. . . . . . . . . . . . 17 Table 2.4 System architectures and modelling parameters in each case study context . . . . . . . . . . . . . 19 Table 2.5 Measured energy consumption for eCooking and modelling assumptions. . . . . . . . . . . . . . . 21 Table 2.6 Normalized energy consumption cooking with traditional fuel, by fuel type. . . . . . . . . . . . . 22 Table 3.1 Electricity supply factors in Kenya, Tanzania, Uganda, and Zambia . . . . . . . . . . . . . . . . . . . 29 Table 3.2 Fuel prices in Nairobi, Kampala, and Dar es Salaam in 2020 and 2025 used in modelling. . . . 37 Table 3.3 Electricity tariffs in Kenya, Tanzania, and Uganda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Table 3.4 Key parameters of selected studies modelling the costs of solar eCooking systems. . . . . . . 64 Table 3.5 Range of opportunities for cost-effective eCooking that open up at different tariff levels . . . . 78 Table 4.1 Applicability of various delivery approaches to each system architecture. . . . . . . . . . . . . . . 84 Table 6.1 Targeted recommendations for creating interministerial spaces . . . . . . . . . . . . . . . . . . . . . 96 Table 6.2 Targeted recommendations for encouraging intersectoral dialogue . . . . . . . . . . . . . . . . . . 96 Table 6.3 Targeted recommendations for using lifeline tariffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Table 6.4 Targeted recommendations for diverting fossil fuel subsidies. . . . . . . . . . . . . . . . . . . . . . . 98 Table 6.5 Targeted recommendations for enabling quality-assured energy-efficient appliances. . . . . . 98 Table 6.6 Targeted recommendations for identifying culturally appropriate appliances. . . . . . . . . . . . 99 Table 6.7 Targeted recommendations for understanding target market segments . . . . . . . . . . . . . . 100 vi ESMAP  |  Cooking with Electricity: A Cost Perspective Table 6.8 Targeted recommendations for enhancing the modelling of solar battery–powered eCooking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Table 6.9 Targeted recommendations for modelling load management on grid systems . . . . . . . . . . . 101 Table 6.10 Targeted recommendations for developing utility and mini grid business models. . . . . . . . . 102 Table 6.11 Targeted recommendations for producing and selling appliances that appeal to customers at the bottom of the pyramid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Table 6.12 Targeted recommendations for developing business models for solar home systems. . . . . . 104 Table 6.13 Targeted recommendations for enhancing the role of players in the clean cooking value chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Table 6.14 Targeted recommendations for empowering women to promote eCooking . . . . . . . . . . . . . 105 Table 6.15 Targeted recommendations for balancing consumer and private sector financing needs. . . . 105 Table 6.16 Targeted recommendations for bridging initial cost–viability gaps . . . . . . . . . . . . . . . . . . . 106 Table 6.17 Targeted recommendations for developing “pay-as-you-cook” financing. . . . . . . . . . . . . . . 107 Table 6.18 Targeted recommendations for helping consumers understand the cost of eCooking. . . . . . 108 Table 6.19 Targeted recommendations for conducting eCooking demonstrations and offering trial periods for consumers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Table 6.20 Targeted recommendations for translating evidence into easy-to-understand content . . . . . 109 Table 6.21 Targeted recommendations for encouraging wider use of energy-efficient appliances . . . . . 109  vii ABBREVIATIONS AC alternating current BMS battery management system CapEx capital expense CO2 carbon dioxide CO2-eq carbon dioxide equivalent DC distributed current EPC electric pressure cooker ESMAP Energy Sector Management Assistance Program GJ gigajoule GOOGLA Global Off-Grid Lighting Association GW gigawatt KPLC Kenya Power and Lighting Company kW kilowatt kWh kilowatt hour kWp kilowatt peak LCoE levelized cost of electricity LED light emitting diode LPG liquified petroleum gas MECS Modern Energy Cooking Solutions MJ megajoule MTF Multi-Tier Framework PAYG pay-as-you-go PM2.5 atmospheric particulate matter with diameter of less than 2.5 micrometers PV photovoltaic SACCO savings and credit cooperative SCODE Sustainable COmmunity DEvelopment SDG Sustainable Development Goal SHS solar home system TANESCO Tanzania Electric Supply Company TaTEDO Tanzania Traditional Energy Development Organisation V volt Wh watt hour ZESCO Zambia Electricity Supply Corporation μg/m 3 micrograms per cubic meter All currency in United States dollars (US$, USD), unless otherwise indicated.  ix ACKNOWLEDGMENTS This report was prepared under the overall guidance of ESMAP’s Program Manager, Rohit Khanna, and MECS’ Research Director, Prof. Ed Brown. Besnik Hyseni, a World Bank energy specialist, managed the project, from inception to publication. The report’s lead authors were Dr. Jon Leary (Loughborough University and Gamos), Besnik Hyseni, Prof. Matt Leach (Gamos), and Dr. Simon Batchelor (Loughborough University and Gamos). Many of the data presented in this report were collected during the projects that laid the foundation for the MECS program. They included (a) “The Next Generation of Low Cost Energy Efficient Products for the Bottom of The Pyramid” project, supported by the Understanding Sustainable Energy Solutions (USES) program, co-funded by UK Aid from the Foreign, Commonwealth and Development Office (FCDO), the Engineering and Physical Sciences Research Council (EPSRC), the Research Councils UK (RCUK), and the Department for Energy and Climate Change (DECC) and (b) the “eCook: A Transformational Household Solar Battery- Electric Cooker for Poverty Alleviation” project, co-funded by UK Aid (FCDO) via Innovate UK Energy Catalyst and Gamos. The findings presented in this report would not have been possible without the dedication and enthusiasm of the 80 households that diligently recorded data on everything they cooked for six weeks. Their willingness to experiment with new appliances and share their experi- ences created a rich learning opportunity. We are grateful to our partners the African Centre for Technology Studies (ACTS), the Tanzania Traditional Energy Development Organisation (TaTEDO), the Renewable Energy Association of Myanmar (REAM), and the Centre for Energy, Environment and Engineering Zambia (CEEEZ) for facilitating the cooking diary studies. These were complemented by a range of activities to explore emerging opportunities in Kenya, Myanmar, Tanzania, and Zambia, including 13 focus group sessions, 6 stakeholder workshops, 14 concept prototypes, and a survey of 800 households. For their time, expertise, and thoughtful comments, we are exceptionally grateful to our peer reviewers: Eliot Avila (A2EI), Iwona Bisaga (BBOXX), Richard Blanchard (Loughborough University), Ewan Bloomfield (Power Africa), Jonathan Bowes (University of Strathclyde), William Brent (Power for All), Malcolm Bricknell (Loughborough University), Toby Couture (E3 Analytics), Aran Eales (University of Strathclyde), Chris Emmott (Fenix International), Jacob Fodio Todd (University of Sussex), Stuart Galloway (University of Strathclyde), Rupert Gammon (De Montfort University), Peter George (Clean Cooking Alliance), Ray Gorman (Power Africa), Sam Grant (CLASP), Rebecca Hanlin (ACTS), Tarek Kettleson (Amperes), Alessandra Leach (independent consultant), Eva Lee (Power for All), Jacquetta Lee (independent consul- tant/University of Surrey), Peter Lilienthal (HOMER), John Maina (Sustainable Community Development Services [SCODE]), Daisy Mkandawire (Zambia Energy Services Company [ZESCO]), Katherine Manchester (Clean Cooking Alliance), Job Ngeni (Power Africa), Samson Ondiek (Kenya Power and Lighting Company), Jessie Press-Williams (Burn Manufacturing), Surabhi Rajagopal (Hivos), Oli Rasion (Biolite), Charlotte Ray (independent consultant), Nick Rousseau (Loughborough University), Dana Rysankova (ESMAP), Estomih Sawe (TaTEDO), Nigel Scott (Gamos), Meron Tesfamichael (University College London), Dipti Vaghela (Hydropower Empowerment Network), Robert Van Buskirk (Enervee), Neal Wade (Newcastle University), and Yabei Zhang (ESMAP). x ESMAP  |  Cooking with Electricity: A Cost Perspective ABOUT ESMAP The Energy Sector Management Assistance Program (ESMAP) is a partnership between the World Bank and development partners and private nonprofit organizations that helps low- and middle-income countries reduce poverty and boost growth through sustainable energy solutions. ESMAP’s analytical and advisory services are fully integrated within the World Bank’s country financing and policy dialogue in the energy sector. Through the World Bank Group (WBG), ESMAP works to accelerate the energy transition required to achieve Sustainable Development Goal 7 (SDG 7) to ensure access to affordable, reliable, sustainable, and modern energy for all. It helps to shape WBG strategies and programs to achieve International Development Association (IDA) policy commitments and the WBG Climate Change Action Plan targets. ESMAP is funded by Australia, Austria, Canada, ClimateWorks Foundation, Denmark, the European Commission, Finland, France, Germany, Iceland, Italy, Japan, Lithuania, Luxembourg,  the Netherlands, Norway, the Rockefeller Foundation, Sweden, Switzerland, and the United Kingdom, as well as by the World Bank.  Lear more at www.esmap.org. ABOUT MECS The Modern Energy Cooking Services (MECS) Program a five-year initiative funded by UK Aid of the Foreign, Commonwealth & Development Office (FCDO) and led by Loughborough University and the World Bank’s Energy Sector Management Assistance Program (ESMAP). The MECS Program aims to accelerate the global transition from traditional biomass-based cooking to modern-energy cooking solutions. By integrating modern energy cooking services into the planning for electricity access, quality, reliability and sustainability, MECS hopes to leverage investment in renewable energies (both grid and off-grid) to address the clean cooking challenge.  MECS is implementing a strategy focused on including the cooking needs of households into the investment and action on “access to affordable, reliable, sustainable modern energy for all.”  Acknowledgments xi A hybrid clean cooking system pairing LPG with a DC EPC powered from a solar home system in Kenya (case study 5). xii ESMAP  |  |  Cooking ESMAP  Cookingwith Electricity:A withElectricity: ACost Cost Perspective Perspective EXECUTIVE SUMMARY Through five case studies, this report compares the current and projected costs to the consumer of a range of electric cooking (eCooking) solutions with the costs of cooking with currently widely-used fuels in each context. The use of energy-efficient electric cooking appliances challenges the widespread perception that electricity is too expensive for cooking. The analysis shows that eCooking can already be a cost-effective option in a variety of settings and is likely to become increasingly effective in the near future. 2.8 billion people globally are still cooking with solid biomass, however, just 789 million are now without access to electricity (ESMAP 2020). This implies that approximately 2 billion people now have access to some form of electricity, but continue to cook with biomass. The case studies show that in some settings, using modern energy-efficient appliances to cook with reliable grid electricity already offers a cost-effective opportunity to enable clean cooking. For people with unreliable electricity access, as well as people who are still not connected to the grid, a suite of new clean cooking technologies and business models is emerging. The results indicate that there is a growing potential to enable modern energy-efficient electric cooking with grid and off-grid electricity, enhancing both reliability and access. Taking the case studies as a baseline, the report extrapolates the results to illustrate the wider application of eCooking for a range of costs and fuel prices and carries out sensitivity analyses to explore emerging trends. The results highlight the cost thresholds that can be used to identify the markets where the levelized costs1 of eCooking systems are already lower than current expenditures on cooking fuels. When the models are projected to include 2025 costs and expenditures, the comparison looks even more favorable, meaning that eCooking is likely to become cost-effective in a broader range of markets. The uptake of eCooking will depend substantially on the willingness of the private sector— in particular solar companies, mini-grid operators and utilities—to adopt the technology as part of the suite of services it offers its customers. Utilities with excess generating capacity could stimulate demand by developing an on-bill financing mechanism for energy-efficient cooking appliances. Financial institutions also have an important role to play, as financing will be needed across the value chain to offset the high upfront costs of eCooking solutions, especially battery-supported models. End-users will require credit to allow them to pay for the high upfront cost of eCooking devices in affordable installments or reframe them as eCook- ing services, where the provider retains ownership of the assets, leasing or renting them to the user. 1 The net present value of investment and operating costs per month of cooking service delivered. Executive Summary x iii The report seeks to build the evidence base to assess whether cooking with electricity could make a significant contribution to the Sustainable Development Goals (SDGs) by simulta- neously enabling cost-effective access to modern energy and clean cooking. The results suggest that integrating planning and action on electrification with the need to transition away from biomass cooking could add momentum to the quest to achieve SDG7 in particular (ensuring access to affordable, reliable, sustainable, and modern energy). Commercial and political interest in eCooking is growing. With appropriate support from governments, adoption of eCooking can be accelerated, yielding substantial environmental, gender equity, and health benefits to some of the world’s most disadvantaged people. Experimenting by cooking ugali in a rice cooker at a workshop in East Africa (case study 1). xiv ESMAP  |  Cooking with Electricity: A Cost Perspective REPORT OVERVIEW Modern energy-efficient electric cooking (eCooking) has the potential to achieve a broad range of developmental goals—for energy access, the environment, gender equity, and health—by enabling access to clean cooking and reliable electricity. Battery-supported cook- ing devices can make cooking with electricity more reliable and offer the co-benefit of also making low power energy services (such as LED lighting or phone charging) more reliable. This emerging opportunity leverages rapid progress in the electricity sector to drive the clean cooking sector toward achieving the seventh Sustainable Development Goal (SDG7) of univer- sal access to affordable, reliable, sustainable, and modern energy by 2030. A new generation of highly efficient eCooking appliances is now available that can drastically lower costs by reducing the amount of electricity required to cook (Zubi and others 2017; Leary, Serenje, Mwila and others 2019; Couture and Jacobs 2019). The electric pressure cooker (EPC) is the most energy-efficient appliance for cooking the most energy-intensive foods. Recent field trials2 have shown that it is also attractive to cooks, as it cooks more quickly and includes automatic controls that allow for multitasking (Leary, Fodio Todd, Batchelor, Chepkurui and others 2019). IMARC (2019) reports that worldwide sales of EPCs totaled 8 million units or $578 million in 2018. It reports that convenience and speed are primary driv- ers of sales. Awareness of the energy efficiency potential of EPCs is still low among consum- ers, but it is growing within the development community, who are searching for cost-effective solutions to the clean cooking challenge. The prices of lithium-ion batteries and solar photovoltaic (PV) power have dropped significantly in recent years, and the cost of biomass fuels is rising rapidly in many heavily degraded or deforested areas (Batchelor 2015; Couture and Jacobs 2019). This trend is opening the door to a range of potentially transformative solutions for cooking with both alternating current (AC) electricity and battery-supported direct current (DC) devices that can enable cooking on weak grids, mini-grids, and stand-alone systems. As a result, mini-grid developers, solar home system companies, and utilities are starting to take a closer look at eCooking. In many developing countries, electricity grids are expanding their coverage and becoming more reliable (Power Africa 2015, 2018), while battery-supported appliances can support weaker grids and enable off-grid access. This development is important, as energy-efficient eCooking appliances can also be powered by batteries, as they draw much less power than conventional electric hotplates. Advancements in energy storage can shift electricity demand away from peak times and allow users to cook during blackouts or brownouts. Advancements in battery storage and solar PV also have the potential to provide electricity access in even the most remote parts of the world (Batchelor and others 2018). 2 Cooking diary studies with 80 households and 13 focus groups across Kenya, Tanzania, Zambia and Myanmar (Batchelor et al. 2019; Scott et al. 2019; Leary, Scott, Serenje, Mwila, et al. 2019b; Leary et al. 2019). Report Overview xv Case Study Methodology and Modeling This report compares the costs to the consumer of cooking with electricity versus other fuels based on detailed empirical data on cooking energy demand. Five case study sites were selected to represent a cross section of contexts in the countries where cooking energy demand data is available, including both urban and rural areas and for households with access to reliable grids, unreliable grids, and no grid access. The report identifies settings where eCooking is likely to be as affordable as (if not cheaper than) current practice by comparing typical expenditures on cooking fuels in the study sites with the levelized costs of a range of eCooking solutions. As further cost reductions of key components are expected, the report compares actual costs in 2020 with projections for 2025. The affordability of cooking is usually assessed based on the proportion of household income spent on cooking fuel, suggesting that even existing expenditures may not be considered “affordable” for households that are already spending a large proportion of their income on cooking fuels. However, this report does not seek to compare cooking fuel expenditures to household incomes. It highlights opportunities where eCooking is already, or will soon be, cost-competitive with current practice. In addition to offering benefits to individual households, eCooking could provide an opportunity to redirect expenditures away from polluting fuels and technologies,3 especially where they are used inefficiently, to support the roll-out of modern energy infrastructure. A model was constructed to simulate the monthly costs of cooking on a range of eCooking systems and compare them with typical expenditures on other fuels (Leach and others 2019). The modeling considers cooking using AC appliances and battery-supported DC appliances, connected to national grid, mini-grid, and stand-alone systems. It also compares two business models: (a) the private sector pay-as-you-go (PAYG) model, with a 5-year financing horizon and (b) the utility (or energy service) model, with a 20-year horizon. The study team collected data on energy consumption, cooking practices, and user experi- ences from households in four countries: Kenya, Myanmar, Tanzania, and Zambia (Leary, Scott, Sago and others 2019; Leary, Scott, Serenje and others 2019a; Leary, Scott, Numi and others 2019; Leary, Scott, Hlaing and others 2019). Data were collected using cooking diary studies, which included assessment of the acceptability and desirability of appliances and electricity usage based on preparation of typical dishes. The data reveal that using a mixture of conven- tional and energy-efficient appliances, the average household (assumed to include 4.2 people) in these countries can perform its daily cooking with 0.88–2.06 kilowatt hours (kWh) of elec- tricity. Under a “fuel-stacking scenario” (in which half the menu is cooked using an EPC and the other half is cooked with another fuel), daily electricity consumption is projected to be just 0.30–0.67 kWh per household. 3 According to the World Health Organization (WHO 2016, 31), polluting fuels and technologies include “biomass (wood, dung, crop residues and charcoal), coal (including coal dust and lignite) and kerosene.” xvi ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE ES.1 Comparison of the five case studies and rationale for selection DEMAND SIDE: KEY OPPORTUNITY ENERGY CASE CONTEXT SUPPLY SIDE BASELINE FUELS/ TO ENABLE 100% STORAGE LOCATION APPLIANCES CLEAN COOKING CONSIDERED LPG, charcoal Clean fuel stack: LPG and kerosene and most e cient Urban, electric appliances national grid 1 Stimulate demand (EPCs) for surplus national None Nairobi, grid electricity Kenya Ine cient electric Most e cient (EPCs) Urban, Household Mitigate load appliances (hotplates, and minimal use of national grid battery 2 shedding on oven) and charcoal less e cient appliances national grids (hotplates, oven) Lusaka, with energy Zambia storage Rural, micro- Firewood and e cient Only e cient electric hydro mini-grid electric appliances appliances (induction Mitigate peak (induction stove, rice stove, rice cooker and Household cooker and insulated insulated electric battery 3 loading constraints on electric frying pan) frying pan) Shan State, micro hydro Myanmar mini-grids with energy storage Rural, solar hybrid mini-grid Clean fuel stack: LPG Charcoal and firewood and most e cient Centralized electric appliances battery bank 4 Stimulate demand for electricity in rapidly (EPCs) Kibindu village, growing solar-hybrid Tanzania mini-grid sector Charcoal, kerosene Clean fuel stack: LPG Rural, o -grid LPG and firewood and most e cient Household electric appliances battery 5 Enable electricity access and clean (EPCs) Echariria cooking with solar village, Kenya systems Report Overview x vii CASE STUDY MODELING RESULTS The case studies illustrate real-world contexts where the levelized cost of eCooking solutions can be lower than existing expenditures on biomass. A range of system architectures and fuel-stacking scenarios was modelled, using actual costs. Figure ES.1 shows the most viable clean cooking solution in each setting. Except for the Tanzania minigrid case, modern energy cooking services are already cost-competitive with the dominant biomass fuel, including elec- tric solutions as well as clean fuel stacking with liquefied petroleum gas (LPG). In some cases, eCooking can be more cost-effective than biomass even if the appliance must be supported by a battery. The first case study explores an opportunity for urban East Africans to transition completely away from biomass by fuel stacking LPG with an EPC. Kenya Power has surplus generation capacity and is looking to increase demand for electricity, which is currently barely used for cooking. LPG is currently the aspirational fuel across most of East Africa, yet many households with an LPG stove still purchase charcoal to cook “heavy foods”. Case study 1 illustrates an urban context with high charcoal prices ($0.49/kg), low LPG prices ($1.08/kg), and average electricity prices (lifeline tariff of 100kWh/month at 0.17/kWh). It shows that a clean fuel stack of LPG and an AC EPC ($7–$10/month) is already one of the lowest-cost cooking solutions and substantially cheaper than charcoal ($23–$34/month). Cost of cooking with biomass (charcoal/firewood) versus cost of cooking with the most cost- FIGURE ES.1  effective technically viable eCooking solution in each of the five case study contexts Firewood AC eCooking Fuel Stacking: Battery-supported DC eCooking / LPG Charcoal Battery-supported DC eCooking Fuel Stacking: AC eCooking / Battery-supported DC eCooking LPG Fuel Stacking: LPG / AC eCooking 2025 40 35 30 2020 Cost of Cooking ($/month) 2025 25 2020 20 2020 15 2025 2025 2025 2025 2020 2025 10 2020 2025 2025 2020 2020 5 2020 2020 2020 2025 0 Charcoal Grid-eCook Charcoal Grid-battery- Firewood Mini-grid- Charcoal Mini-grid- Charcoal Solar-battery- and LPG eCook battery-eCook eCook and LPG eCook and LPG CASE STUDY 1 CASE STUDY 2 CASE STUDY 3 CASE STUDY 4 CASE STUDY 5 Kenya Grid Zambia Grid Myanmar Tanzania Kenya Mini-grid Mini-grid SHS Note: Case study 1, Kenya grid: Fuel stack of 50 percent liquefied petroleum gas (LPG) and 50 percent AC electric pressure cooker (EPC); private sector model (five- year financing horizon). Case study 2, Zambia grid: Hybrid AC/DC appliances with battery sized for 50 percent of cooking; utility model (20-year financing horizon). Case study 3, Myanmar mini-grid: Hybrid AC/DC appliances with battery sized to power 50 percent of cooking; utility model (20-year financing horizon). Case study 4, Tanzania mini-grid: Fuel stack of 50 percent LPG and 50 percent AC EPC; private sector model (five-year financing horizon). Case study 5, Kenya solar home system: Fuel stack of 50 percent LPG and 50 percent solar home system with DC EPC and battery sized to power 50 percent of household cooking; private sector model (five- year financing horizon). xvi i i ESMAP  |  Cooking with Electricity: A Cost Perspective The second case study illustrates an opportunity for countries with significant populations already cooking with electricity but using inefficient appliances, to optimize loading on their grids. Although electricity is already the aspirational cooking fuel in Zambia, the national utility (ZESCO) has repeatedly been forced to carry out load shedding over the past few years, as late rainfall has severely limited generation capacity on its hydropower-dominated grid. Case study 2 illustrates an urban context with lower charcoal prices ($0.21/kg) and low electricity prices (lifeline tariff of 200kWh/month at $0.01/kWh). The findings show that by 2025, a hybrid AC/DC eCooking system with a battery sized for half the day’s cooking using energy-efficient appliances and practices will be the cheapest option ($7–$8/month), substantially cheaper than charcoal ($6–$12/month) The third case study highlights the opportunity for micro-hydro minigrid developers that have already enabled cooking on their systems to allow their customers to do all of their cooking with electricity. At peak times, grids often reach capacity and the voltage dips. This case study explores the potential role of battery storage in overcoming the supply constraints on micro-hydro minigrids in Myanmar. Case study 3 shows a rural area, with moderate firewood prices ($0.12/kg) and electricity access from a micro-hydro minigrid with a low tariff ($0.16/kWh). By 2025, a battery sized to support half the day’s cooking load could enable 24-hour eCook- ing ($9–$10/month), the cost of which would be on a par with firewood ($6–$11/month). The fourth case study explores how the rapidly falling prices of batteries and solar PV are opening up new opportunities for integrating energy-efficient eCooking into solar-hybrid minigrids. Urbanization is causing many people who used to collect fuel to start paying for it, creating an opportunity to translate expenditures on biomass fuels into electricity units, which could drive down the tariff for the minigrid as a whole. Case study 4 depicts a rural area with low-cost biomass fuels available (firewood: $0.04/kg, charcoal: $0.13/kg) and access to elec- tricity via a minigrid with a very high tariff ($1.35/kWh). By 2025, tariffs in the solar hybrid mini- grid sector are expected to have fallen considerably (to $0.25–$0.38/kWh), enabling eCooking at marginal extra cost by fuel stacking an EPC. The most cost-effective clean cooking solution is a clean fuel stack of LPG and an EPC ($12–$21/month). A participant in an EPC trial on a solar-hybrid mini-grid in Tanzania (case study 4). Report Overview x ix The fifth case study describes a Kenyan village, where cooking was previously dominated by collected firewood, but dwindling forest resources and increasing livelihood opportunities have led many residents to start paying for firewood (or adopt charcoal, kerosene, or LPG). It explores whether pairing a DC EPC with lithium-ion battery storage and a suitably sized solar panel may be able to offer a cost-effective off-grid eCooking solution. Case study 5 illustrates an off-grid rural area with moderate fuel prices (charcoal: $0.30/kg; LPG: $1.33/kg). In 2025, the cheapest option is expected to be LPG ($8–$12/month). However, a clean fuel stack of LPG with a solar home system powering a DC EPC ($11–$14/month) can offer valuable co-ben- efits by enabling access to electricity for other purposes at marginal extra cost. The global perspective Figure ES.2 shows the outlook for eCooking at a global level by comparing the range of costs of the eCooking technologies explored in this paper with those of the most widely used cook- ing fuels. Input data were drawn from across the four case study countries (Kenya, Zambia, Tanzania, and Myanmar) and the three system architectures (grid, mini-grid, solar home system). The results show that AC eCooking on national grids or mini-/micro-hydropower is already cost-effective for many people today and that battery-supported DC eCooking and solar-hy- brid minigrids become cost-effective in 2025, although clean fuel stacks with LPG can make all of these technologies cost-effective today. Cooking with AC grid electricity can be the cheap- est option for many people ($3–$17/month), but it is not always possible due to access and grid stability challenges. Supporting 50 percent of cooking loads with a battery increases the cost of cooking ($5–$22/month in 2025) but is still competitive with LPG, charcoal, and fire- wood ($6–$24/month, $5–$41/month, and $0–$23/month, respectively in 2025). Supporting 100 percent of the cooking loads increases the cost substantially (to $8–$39/month in 2025) but may still be competitive in contexts with low tariffs and low energy demand. By 2025, the costs of cooking with AC appliances connected to solar hybrid mini-grids ($8–$25/month) Training a cooking diary and with DC appliances powered by solar home systems ($11–$24/month) become compet- participant on cooking itive. LPG can play an important role as a transition fuel since a clean fuel stack of electricity beans in an EPC in Kenya and LPG can make battery-supported eCooking cost-competitive for some households today (case study 1). ($6–$29/month). xx ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE ES.2 Comparison of system architectures using aggregated data from all case studies Year Electrical System Architecture Fuel stacking Cooking fuel 2020 AC without household battery AC eCooking / Battery-supported DC eCooking Charcoal 2025 Battery-supported DC LPG / AC eCooking LPG (household battery) Battery-supported DC eCooking / LPG Firewood 100% Electric Clean Fuel Stack (Electricity and LPG) Cooking Fuels 40 35 Cost of Cooking Service ($/month) 30 25 20 15 10 5 0 Note: The cost of cooking service is calculated over a five-year financing period for all system architectures. The range on each bar represents sensitivities to energy demand, to the grid tariff or solar resource and to key system performance and cost parameters. The ranges for energy demand are derived from the range of median values from the four country cooking diary studies for 100 percent eCooking (0.87–2.06kWh/household/day). The ratios of energy demand for cooking fuels: electricity calculated from the cooking diaries were used to model demand for LPG (2: 1), charcoal (10: 1) and firewood (10: 1). Grid-connected system architectures use a tariff range encompassing 90 percent of Sub-Saharan African utilities from AFREA and ESMAP (2016): $0.04–$0.25/kWh. National grids and mini-/micro-hydropower are grouped together, as tariff ranges are almost identical ($0.05–$0.25/kWh for mini-/micro-hydropower) (Skat 2019). Solar hybrid mini grid system architectures use a current tariff range of $0.55–$0.85/kWh and a range of $0.25–$0.38/kWh in 2025. The solar resource range is the range of average monthly solar irradiation in the least sunny months in each of the four case study countries (3.68–4.30kWh/kWpeak). eCook system performance and cost ranges are as reported in Table 2.3. Batteries are LiFePO4, sized to meet 100 percent and 50 percent of daily cooking loads, at 1–3kWh and 0.34–0.98kWh, respectively. PV is 300–700W for 100 percent and 100–200W for 50 percent. For full details of modelling input and output parameters, see appendix F. The critical role of energy-efficient appliances Both energy-efficient appliances and fuel stacking can substantially reduce the costs of electric cooking, with or without a battery (figure ES.3). An uninsulated four-plate cooker and oven may be cost-effective for households with reliable grid electricity and low tariffs ($7/ month at $0.04/kWh). It is unlikely that anyone would consider supporting it with a battery, which would need 4.56kWh capacity ($28/month even at $0.04/kWh). In contrast, the appli- ance stack of uninsulated (hotplate, induction, infra-red cooker, or kettle) and insulated (EPC, rice cooker, electric frying pan, or thermo-pot) appliances can offer a much more affordable solution that is capable of covering 100 percent of a household’s everyday cooking needs. It would cost $4–$13/month for AC (where the grid is reliable enough) and $13–$29/month for battery-supported DC. Simply cooking with a single uninsulated appliance will be cheaper for some AC users as the upfront cost of appliances is lower but cooking may be less convenient. For the DC systems, the cost of the battery dominates, so spending more on an additional Report Overview xxi A community solar hub acts as a demonstration, energy-efficient appliance actually reduces overall costs (from $16–$37/month to $13–$29/ distribution and after-sales month), as the battery capacity is reduced (from 2.85kWh to 2.14kWh). service centre for solar electric cooking systems Although it cannot cook all food types, the EPC is likely to be an attractive first step into in a Kenyan village (case eCooking for many, as it can deliver the cheapest cooking service by some considerable study 5). margin. Systems could be designed to cook 50 percent of the menu (at a cost of $2–$5/ month for AC or $5–$11/month for battery-supported DC) or simply what the EPC does most efficiently, which is boil heavy foods (at a cost of $2–$3/month for AC and $3–$4/month for battery-supported DC). MAIN FINDINGS Several key findings emerge from this report: ● Field trials with 80 households show that modern energy-efficient eCooking appliances (notably EPCs) are highly attractive to consumers and can substantially lower the cost of eCooking by reducing energy demand. Compared with electric hotplates, EPCs can reduce energy demand by 80 percent for “heavy foods” (foods that require boiling for more than an hour) and by 50 percent across the entire range of foods that they are able to cook.4 ● The cost of cooking with energy-efficient appliances is significantly lower than the cost of cooking with electric hotplates, but the upfront cost is higher (typically $50–$80 for an EPC, compared with $10–$30 for a hotplate). ● eCooking with AC grid electricity is already cheaper than cooking with charcoal in some of the urban centers studied, where charcoal costs more than $0.40/kg and electricity tariffs are below $0.35/kWh. 4 Analysis of the menu recorded during these trials showed that participants cooked 50 percent of their meals on an energy-efficient appliance and that with additional training this share could increase to up to 90 percent. xx i i ESMAP  |  Cooking with Electricity: A Cost Perspective Impact of energy-efficient appliances and fuel stacking on cost of AC and battery-supported FIGURE ES.3  DC eCooking Uninsulated four-plate with Appliance stack of uninsulated (hotplate, induction, 100% grid eCook integrated oven and infra-red cooker or kettle) and insulated (EPC, (AC w/o battery) rice cooker, and electric frying pan or thermo-pot) Uninsulated single-plate 100% grid-battery-eCook (hotplate and induction or Single insulated and pressurized appliance (EPC) (DC w/battery) infra-red cooker) 40 35 Cost of Cooking Service ($/month) 30 25 20 15 10 5 0 100% of household 50% of household Boiling heavy 100% of household 50% of household Boiling heavy cooking cooking foods only cooking cooking foods only Note: The cost of the cooking service is calculated over a five-year financing period for all system architectures. Component costs are from 2025. The range on each bar encompasses 90 percent of Sub-Saharan African utility tariffs from AFREA and ESMAP (2016) ($0.04–$0.25/kWh). Daily household energy demand values are from Figure 2.2 (100 percent eCooking: uninsulated plate with oven, 3kWh; uninsulated single plate, 2kWh; appliance stack, 1.5kWh; 0.5kWh. 50 percent eCooking: EPC, 0.5kWh. Boiling heavy foods only: EPC, 0.15kWh). Fuel-stacking scenarios model only the eCooking service, not the cost of the cooking fuel. ● Using a clean fuel stack of LPG and a highly efficient eCooking appliance is often the most cost-effective way to cook.5 ● Battery-supported eCooking is already cost-effective for charcoal users in urban centers with electricity tariffs below $0.15/kWh. ● By 2025, expected increases in charcoal prices and the falling costs of battery-supported solutions suggest that the cost of eCooking will likely be comparable to the cost of cooking with charcoal in weak-grid and off-grid contexts ($8–39/month vs. $5–41/month respectively). ● Battery-supported cooking devices can also provide access to other low power energy services such as lighting and mobile phone charging. ● Stand-alone solar systems start to become competitive with their grid-connected counterparts at tariffs of $0.15–$0.35/kWh (see figure ES.3). 5 LPG is a good complementary fuel to eCooking since it is popular for frying and preparing quick meals Report Overview x x iii Comparing energy- efficiency and service ● Lifeline tariffs of 100kWh/month at $0.10/kWh would be sufficient to allow most consumers delivery amongst to cook with electricity, even if the cooking appliance had to be supported by a battery. popular electric cooking appliances in Myanmar HOW CAN ECOOKING BE DELIVERED AND FINANCED? (case study 3). Innovative delivery and financing models will be needed to support the roll-out of eCooking since even where it is cost-competitive, challenges remain, especially if energy storage is required. In markets that do not require energy storage, supply chains for energy-efficient appliances are emerging but are not yet strong and the high upfront cost prevents many poorer households from accessing them. For example, private sector retail supply of EPCs is increasing in Asia, but is not yet common in Sub-Saharan Africa (IMARC 2019), where awareness among consumers remains low. In markets where energy storage will be needed, batteries further increase the upfront cost, which will require financing with longer repayment horizons, additional supply chain development, consumer awareness, and after-sales support. End-users will require credit options to break down the high upfront cost of eCooking devices into affordable installments or reframe them as eCooking services, where the provider retains ownership of the assets and rents them to the user. For example, pay-as-you-go for lease- to-own solutions and on-bill financing for energy service models.6 The uptake of eCooking will depend substantially on the willingness of energy service companies to integrate it into the suite of services they offer. For example, utilities with excess generating capacity could stimulate demand by developing an on-bill financing mechanism for EPCs and support women entrepreneurs to leverage their social networks to demonstrate new cooking technologies and practices. Grant funding could support an initial feasibility study and piloting, with results-based financing and other instruments accelerating scale up. Distributors and retailers will require working capital to finance the appliances and roll out supporting services over longer repayment 6 Pay-as-you-go systems rely on a “lock-out” mechanism to prevent the device from functioning if the user does not keep up with regular repayments. On-bill financing allows installments to be repaid automatically when topping up electricity units on prepaid meters or adding to the monthly bill on post-paid meters. xx i v ESMAP  |  Cooking with Electricity: A Cost Perspective periods. Financing instruments—including debt and equity finance, social impact investment, and results-based financing tied to environmental, gender equity, and/or health goals—will need to be combined to close the initial cost–viability gaps. A “single investment strategy” that incorporates clean cooking into electrification and renew- able energy investments could enable the existing mechanisms for mobilizing finance from the electricity sector to address the problem of cooking with polluting fuels and technologies. These include long-term loans, guarantees, and project bonds, which can offer the clean cooking sector an opportunity to leverage much larger investments. Such a strategy could synergistically position eCooking as an opportunity to improve delivery infrastructure and stimulate demand. Conclusions and Recommendations The case studies examined in this report show that in specific contexts, cooking with ener- gy-efficient electric appliances is already a cost-effective option. As prices of key components continue to fall, the range of contexts in which eCooking can offer a cost-effective alterna- tive to polluting fuels and technologies is expected to broaden, challenging the widespread perception that electricity is too expensive for cooking in developing regions. Commercial and development partners’ interest in eCooking is growing. With appropri- ate support, adoption of eCooking can be accelerated and attention focused on achieving pro-poor outcomes. Integrating planning and action on electrification with the need to transi- tion away from biomass cooking can accelerate progress toward SDG7 and yield environmen- tal, gender equity, and health benefits to some of the world’s most disadvantaged people. However, even in places where energy-efficient electric appliances are cost-effective, challenges exist. They include the lack of supply chains, high upfront costs for consumers, lack of awareness, the need for changes in the way people cook, and uncertainty about the impacts of scaled uptake on grid systems. Working together, governments, donors, and private sector can address most of these challenges—recommended actions to support the roll-out of eCooking solutions include: 1. Support policy makers to create an enabling environment that crosses the division between the electrification and clean cooking sectors • Reduce the lifetime cost of eCooking by bringing down the upfront cost of quality- assured energy-efficient appliances by streamlining supply chains (through, for example, the Global LEAP awards program for EPCs).7 • Create interministerial spaces (committees, working groups, and so forth) to develop single investment strategies that align with existing political objectives. • Create a space for dialogue between stakeholders in the clean cooking and electrification sectors. • Reduce the relative cost of cooking with electricity by diverting fossil fuel subsidies to energy access programs. • Strengthen the case for the poor through strategic use of lifeline tariffs financed by cross-subsidies or targeted subsidy programs. 7 The Global LEAP Awards is an international competition to drive innovation and performance in early-stage product markets. Awards provide market intelligence for investors, donors, policymakers, solar distributors, and other off-grid market stakeholders. Report Overview xxv 2. Conduct strategic, evidence-based research to inform decision makers, private sector players, and consumers of emerging opportunities • Identify and popularize culturally appropriate energy-efficient eCooking appliances. • Gain a deeper understanding of target market segments, particularly of their existing expenditures on cooking fuels. • Enhance techno-economic models by including the expected costs of marketing, selling and supporting solar battery–powered eCooking devices in rural areas. • Model the implications of encouraging eCooking for load management on national grids and mini-grids, in order to establish the likely impact on overall costs and the integrity of the systems. 3. Support private sector efforts to develop appropriate products and services tailored to the needs and aspirations of the poor • Enable utilities and minigrid developers to pilot, and scale up eCooking services that are compatible with their existing business models. • Enable solar home system companies to develop, pilot, and scale up innovative new eCooking products and services. • Incentivize appliance manufacturers to develop products targeted at the bottom of the pyramid, in particular DC– and battery-supported eCooking products. • Enable players in the existing clean cooking value chain to expand their product range to include eCooking appliances. • Empower women entrepreneurs to lead the development and dissemination of innovative eCooking solutions. • Identify viable business models that will both unlock consumer responses and meet private sector financing needs. • Bridge initial cost–viability gaps in new markets by combining financing instruments, including including grants, social impact investment and results-based financing tied to environmental, gender equity, and health outcomes. 4. Help consumers understand the benefits of adopting modern eCooking solutions, and reduce barriers to behavioral change • Help consumers determine how much it would really cost them to cook with electricity. • Make it possible for consumers to explore eCooking through participatory eCooking demonstrations and trial periods with limited financial risk to the consumer. • Encourage consumers to cook as much of their typical menu on energy-efficient appliances as possible. • Translate evidence-based research into easy-to-understand content that can be shared on popular media (by, for example, creating targeted content on EPCs for social media groups on cooking). • Develop “pay-as-you-cook” financing (flexible repayment schemes that are based on how consumers currently pay for biomass). The Modern Energy Cooking Services (MECS) program is supporting strategic interventions in each of the five case study contexts featured in this report (plus many more). Over the next decade, the relative price points of key technologies will continue to change, which will likely open the door to an even broader range of cost-effective eCooking solutions. The program intends to keep close track of these developments, create a range of market-ready innova- tions, and shape enabling environments to make a valuable contribution toward SDG7. xx v i ESMAP  |  Cooking with Electricity: A Cost Perspective References Batchelor, S., E. Brown, J. Leary, N. Scott, A. Alsop, and M. Leach. 2018. “Solar Electric Cooking in Africa: Where Will the Transition Happen First?” Energy Research and Social Science 40. https://doi.org/10.1016/j.erss.2018.01.019. Batchelor, S. 2015. “Solar Electric Cooking in Africa in 2020: A Synthesis of the Possibilities.” Evidence on Demand, prepared at the request of the UK FCDO (Department for International Development. https://doi.org/10.12774/eod_cr.december2015.batchelors. Batchelor, S, J Leary, S Sago, A Minja, K Chepkurui, E Sawe, J Shuma, and N Scott. 2019. “Opportunities & Challenges for ECook Tanzania—October 2019 Working Paper.” TaTEDO (Tanzania Traditional Development Organisation), Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid (FCDO) & Gamos Ltd. Available. www.MECS.org.uk. Couture, T., and D. Jacobs. 2019. “Beyond Fire: How to Achieve Electric Cooking.” HIVOS & World Future Council. ESMAP (Energy Sector Management Assistance Programme). 2020a. Tracking SDG7 Progress Towards Sustainable Energy. Washington, DC. https://trackingsdg7.esmap.org//. IMARC. 2019. “Multi Cooker Market: Global Industry Trends, Share, Size, Growth, Opportunity and Forecast 2019–2024.” Leach, M., J. Leary, N. Scott, S. Batchelor, X. Chen, K.-S. Ng, R. Oduro, and E. Brown. 2019. “ECook Modelling.” www.MECS.org.uk. Leary, J., J. Fodio Todd, S. Batchelor, K. Chepkurui, M. Chepkemoi, A. Numi, R. Hanlin, N. Scott, and E. Brown. 2019. The Kenya ECookBook: Beans & Cereals Edition. MECS, ACTS, Loughborough University, Gamos and University of Sussex supported by EPSRC and UK Aid (FCDO): Available from: www.MECS.org.uk. Leary, J., N. Scott, W. W. Hlaing, A. Myint, S. Sane, P. P. Win, T. M. Phyu, et al. 2019. “ECook Myanmar Cooking Diaries—October 2019 Working Paper.” REAM, Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid (FCDO) & Gamos Ltd. Available. www.MECS.org.uk. Leary, J., N. Scott, A. Numi, K. Chepkurui, R. Hanlin, M. Chepkemoi, S. Batchelor, M. Leach, and E. Brown. 2019. “ECook Kenya Cooking Diaries—September 2019 Working Paper.” http://www.sussex.ac.uk/spru/research/projects/lct. Leary, J., N. Scott, S. Sago, A. Minja, B. Batchelor, K. Chepkurui, E. Sawe, M. Leach, and E. Brown. 2019. “ECook Tanzania Cooking Diaries—October 2019 Working Paper.” REAM (Renewable Energy Association of Mynamar), Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid (FCDO) & Gamos Ltd. www.MECS.org.uk. Leary, J., N. Scott, N. Serenje, F. Mwila, S. Batchelor, M. Leach, E. Brown, and F. Yamba. 2019a. “ECook Zambia Cooking Diaries—October 2019 Working Paper.” CEEEZ (Centre for Energy, Environment and Engineering Zambia), Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid & Gamos Ltd. www.MECS.org.uk. ———. 2019b. “Opportunities & Challenges for ECook in Zambia—October 2019 Working Paper.” CEEEZ, Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid & Gamos Ltd. www.MECS.org.uk. Leary, J., N. Serenje, F. Mwila, S. Batchelor, M. Leach, E. Brown, N. Scott, and F. Yamba. 2019. “ECook Zambia Prototyping Report.” Implemented by CEEEZ, Gamos, Loughborough University, University of Surrey. Funded by FCDO, Innovate UK, Gamos. www.MECS.org.uk. Power Africa. 2015. “Development of Kenya’s Power Sector 2015–2020.” Nairobi, Kenya. ———. 2018. “Power Africa in Uganda.” 2018. https://www.usaid.gov/powerafrica/uganda. Scott, N., J. Leary, W. W. Hlaing, A. Myint, S. Sane, P. P. Win, T. M. Phyu, et al. 2019. “Opportunities & Challenges for ECook in Myanmar—October 2019 Working Paper.” REAM, Loughborough University, University of Surrey & Gamos Ltd. supported by Innovate UK, UK Aid & Gamos Ltd. www.MECS.org.uk. WHO (World Health Organisation). 2016. “Burning Opportunity: Clean Household Energy for Health, Sustainable Development, and Wellbeing of Women and Children.” Geneva, Switzerland. http://apps.who.int/iris/bitstream/10665/204717/1/9789241565233_eng.pdf?ua=1. Zubi, G., F. Spertino, M. Carvalho, R. S. Adhikari, and T. Khatib. 2017. “Development and Assessment of a Solar Home System to Cover Cooking and Lighting Needs in Developing Regions as a Better Alternative for Existing Practices.” Solar Energy 155: 7–17. https://doi.org/10.1016/j.solener.2017.05.077. Report Overview x x vii Ch apter 1 BACKGROUND 1.1.  State of Access to Clean Cooking and Electricity Sustainable Development Goal 7 (SDG7) seeks to ensure In order to align with the tracking of the SDG7 goals, this access to affordable, reliable, sustainable, and modern report uses the latest estimates on progress toward achiev- energy for all. It indicates that households require access ing SDG7, which indicate that the share of the population to both electricity and clean cooking. A new paradigm is with access to clean cooking increased to 61 percent in emerging that sees an opportunity to tackle both problems 2017, up from 57 percent in 2010 (ESMAP 2020a). However, by creating a symbiotic relationship in which actors from because population growth outpaced annual access gains, both sides can support each other in achieving universal the global access deficit remained stable, at about 2.9 billion. access to modern energy (Batchelor et al. 2019; Couture and Assuming the rate of increase in access of 0.5 percentage Jacobs 2019). points a year seen between 2010 and 2017, clean cooking solutions would reach only 68 percent of the global popu- The proportion of the global population with access to lation by 2030 (Figure 1.1). In 2010, it was estimated that an electricity increased from 83 percent in 2010 to 90 percent average annual increase of 2 percentage points would be in 2018 (ESMAP 2020a), with the number of people living necessary to achieve universal access to clean cooking. To without electricity dipping to 789 million, down from make up for slower progress than required over the period 1.2 billion in 2010. At the same time, an estimated 2.8 billion 2010–17, access would need to increase by a rate at least people still cook with biomass (ESMAP 2020a). Some of the 3 percentage points a year (ESMAP 2020a). 2 billion people that have access to electricity but still cook with biomass already have access to reliable electricity and Although there is evidence of some progress toward meet- could directly transition to cooking with electricity. Data from ing SDG7, it tends to be uneven across the globe (Figure 1.2). the Multi-Tier Framework (MTF)—used to measure the level In Sub-Saharan Africa, the annual average population growth and quality of energy access—provide insights at the country rate is about 2.7 percent (World Bank 2020). As a result, level on the proportion of households that are still cooking a large number of additional people increasingly rely on primarily with polluting fuels and technologies but that have biomass fuels for cooking. As a result of population growth, Tier 3, 4, or 5 electricity access (see Figure 1.3 for details).1 some countries are experiencing a decline in the share of the population with access to clean cooking solutions A forthcoming report by the World Bank’s Energy Sector (ESMAP 2020a). The rapid pace of urbanization also means Management Assistance Program (ESMAP) looks at a that households are often switching from collecting biomass 71-country sample of 5.3 billion people representing residue in rural areas to purchasing wood fuels (mainly 90 percent of lower- and lower-middle-income countries. It charcoal) from urban markets. Adam Smith International uses the Modern Energy Cooking Solutions (MECS) definition (2016) finds that a 1 percent rise in urbanization can increase of access.2 It finds that some 4 billion people—about half the charcoal consumption by 14 percent. In 2017, the average global population—lack the ability to cook efficiently, cleanly, annual rate of urbanization in Sub-Saharan Africa was about conveniently, reliably, safely, and affordably, suggesting 4.1 percent (in some countries as high as 5.7 percent) (World that the problem may be graver than previously thought. Bank 2020). At these rates, the population currently living Increasing the number of people who cook with electricity is in African cities—about 472 million people—is projected to one way of reducing this figure significantly. double by 2050 (CSIS 2018). Background 1 FIGURE 1.1 Actual and projected global access to electricity and clean cooking, 2000–30 100 Unelectrified without Clean Cooking 0.8 90 4% Short 80 king = 28% 32% Short 70 without Clean Coo 2.1 Global access (percent) Electrified but 60 Business as usual 2030 50 Predictions Electrified with Clean Cooking 40 30 4.6 20 10 0 2000 2005 2010 2015 2020 2025 2030 6.1 7.0 7.8 8.5 Population (billion people) Clean Cooking Electricity Linear (Clean Cooking) Linear (Electricity) Note: Energy access statistics from World Bank (2019b). Population forecasts from United Nations World Population Prospectus (UN 2017). Linear forecasting was used to project global access beyond 2016. Source: Adapted from Batchelor et al. (2019). FIGURE 1.2 Share of population with access to clean cooking fuels and technologies, by region, 2017 IBRD 44349 | APRIL 2019 POPULATION WITH ACCESS TO CLEAN COOKING FUELS AND TECHNOLOGIES (%) Less than 10 10-49 50-99 100 Data not available Top 20 Access Deficit Countries Source: ESMAP (2020). 2 ESMAP  |  Cooking with Electricity: A Cost Perspective Three-quarters of the 570 million people who gained access gas emissions.4 In addition, certain components of partic- to electricity since 2011 are concentrated in Asia (IEA 2018). ulate matter, collectively referred to as black carbon, are a However, clean cooking remains a challenge for many Asian powerful climate change forcing agent, as a result of their countries. China, for example, which has reportedly reached heat absorption characteristics. When black carbon settles 100 percent electrification, is one of the highest deficit coun- on otherwise reflective surfaces (such as snow or ice), the tries for access to clean cooking (Figure 1.2). forcing effect is compounded (World Bank 2013). Cooking with solid fuels is the largest source of black carbon emis- sions globally. Where fuel must be purchased, the increasing cost of char- 1.2.  The Burden of coal (and in some cases fuel wood) places a burden on poor and vulnerable families struggling to meet basic needs. In Cooking with Biomass Uganda, for example, charcoal prices increased by almost 30 percent in 2017 (inflation was less than 10 percent). On Fuels average, a household spends as much as $24 for a 75-kg bag of charcoal, which lasts about a month (Musoke 2017). This figure represents about 10 percent of the average The 2.9 billion people worldwide using biomass for cook- income of Kampala residents, almost 13 percent of average ing use a variety of fuels, including wood, charcoal, animal income in other urban areas, and 29 percent of average dung, crop waste, or other solid fuels, such as coal, in income in rural areas (UBOS 2017). The rising prices are open fires and traditional stoves, as the primary source of partly a result of the growing distances for transporting the cooking and heating energy. Four million people a year fuel from its source to urban areas. Households are thus die as a result of household air pollution; more than half of spending a significant and growing share of their monthly these deaths are among children under five (WHO 2018). In income on wood fuels. What is more, the poorest house- addition to the direct health burden from premature deaths holds often pay a premium for their daily fuel purchases and ill health, exposure to household air pollution is linked (45 percent on average for the urban poor), as a result of to low birthweight, which increases the risk of poor health cash flow constraints, and allocate a significant proportion outcomes throughout life. This avoidable, first-order public of their household expenditures to cooking fuels such as health problem has been an impetus behind recent initia- charcoal (World Bank 2015). tives for delivery and adoption of clean cooking solutions. Smith et al. (2014) estimate that household air pollution cost Where fuel is collected for self-consumption rather than low- and middle-income countries $1.5 trillion in 2013 in purchased, the time spent collecting it—mostly by women— welfare losses, primarily as a result of the health impacts, an could often be better spent on income generation, farming, amount equivalent to 3.3 percent of GDP (World Bank 2016). education, childcare, or leisure. Carrying heavy bundles of The public health crisis caused by Covid-19 is threatening firewood—often over long distances—can cause injuries, to exacerbate complications from exposure to household and foraging for firewood can expose girls and women to air pollution. A study conducted at the Harvard University gender-based violence. Access to modern energy cooking School of Public Health (Wu et al. 2020) suggests that services could address some of these problems and release there is significant overlap between exposure to particulate time into the labor market. matter and Covid-19 deaths. An increase of only 1μg/m3 in atmospheric particulate matter with a diameter of less than The biomass sustainability problem is worsening, as a result 2.5 micrometers (PM2.5) is associated with an 8 percent of population growth and rapid urbanization, accelerating increase in the Covid-19 death rate. charcoal consumption. Although cooking with charcoal may be cleaner than wood in terms of household air pollution In addition to the obvious impact on indoor air quality, impacts, the use of charcoal for cooking has approximately household combustion of solid fuels contributes to ambient four times the deforestation impact that cooking with wood air pollution. Investigators contributing to the analysis of the has, because approximately 75 percent of wood’s chemical latest global burden of disease estimated the global average energy is lost in the conversion to charcoal (Falcão 2008). In proportion of ambient PM2.5 attributable to household cook- several countries, governments have banned the production ing at 12 percent.3 In some places, unsustainable harvesting of charcoal, without offering a viable alternative. A transition of biomass also leads to degradation of landscapes, loss from solid fuels to clean and more sustainable fuels needs to of biodiversity and wildlife habitat, and net greenhouse accelerate considerably. Background 3 each context; the results are therefore relevant only for each 1.3.  Objective and Scope context. However, the report contextualizes these findings by drawing on broader datasets and undertaking sensitivity of the Report analysis to support more general conclusions that may be relevant to other developing countries that have large popu- lations without access to clean cooking solutions. This report identifies opportunities for cooking with electricity that are already cost-effective in developing The report concludes with recommendations for what regions and opportunities that are likely to open up in the governments and development organizations can do to near future. It aims to build the evidence base on whether support the transition to cooking with electricity in the right cooking with electricity could make a significant contri- contexts. Integrating planning and action on electrification bution to achieving SDG7 (ensuring access to afford- with the need to transition away from biomass cooking could able, reliable, sustainable, and modern energy for all) by add momentum to the mission of achieving SDG7 in particu- simultaneously enabling cost-effective access to modern lar. With appropriate support from governments, adoption of energy and clean cooking. eCooking can be accelerated, yielding substantial environ- mental, gender equity, and health benefits to some of the The report provides insights on the techno-economic viabil- world’s most disadvantaged people. ity of cooking with electricity in specific country contexts. It is based on new data on how people cook with electricity eCooking may not currently feature prominently in the in cultures where biomass cooking is prevalent. In each mindset of utility, mini grid, and off-grid developers, but that context, it compares the cost of cooking with traditional fuels may change soon, as commercial and political interest in with the cost of cooking with electricity across a range of eCooking is growing. ESMAP, Loughborough University, and system architectures (grid-connected, mini grid, and stand- their partners are collaborating on Modern Energy Cooking alone systems). Services (MECS), a major new UK Aid–funded program that aims to bring together the clean cooking and electricity The report presents case studies from Sub-Saharan Africa sectors to develop emerging opportunities for cooking with and Southeast Asia to explore the range of opportunities in electricity. 4 ESMAP  |  Cooking with Electricity: A Cost Perspective therefore attract private and government investment in a way 1.4.  The Case for Cooking that improved cookstoves have not. with Electricity The cost of biomass fuels is rising substantially in many contexts. Even in contexts with higher unit costs of electric- ity, higher biomass prices and lower consumption with new Until recently, the development community has not viewed energy-efficient appliances mean that the common percep- electricity as a viable option for enabling access to clean tion that electricity is too expensive for cooking is no longer cooking, because of reliability, safety, access, affordability, true in many contexts. Although biomass energy will likely and sustainability challenges. Blackouts and brownouts on continue to be the predominant fuel for cooking in many weak grids prevent people from cooking when they need to, parts of the developing world for some time, in particular in and collective usage causes peak loads on already strained Sub-Saharan Africa, and addressing the efficiency of its use grids to spike and exacerbate underlying problems. There is important, the development community needs to push the is also concern about poor-quality wiring, which could burn switch to clean fuels more actively, supported and driven by out and start a fire if high currents are drawn by inefficient the public and private sectors. cooking devices. Cooking with electricity can be a truly clean cooking solu- However, a growing community of actors is drawing atten- tion, in terms of direct emissions in the kitchen, which affect tion to the fact that through technological developments, both ambient and household air quality, as well as environ- these challenges can now be mental sustainability (assum- addressed in some contexts ing electricity is produced (Couture and Jacobs 2019; from low- or zero-emission Cooking with electricity can present a Batchelor 2013; Batchelor et sources). Electricity has the al. 2019). Electricity grids are transformative value proposition for potential to very quickly growing stronger and gaining households, allowing for more efficient switch entire urban and greater coverage (Kenya peri-urban communities that Power 2018; Power Africa and faster cooking times, multitasking, are already grid-connected 2018; Eberhard, Gratwick, and safer cooking, elimination of dangerous from traditional fuels, as the Kariuki 2018), and advance- indoor emissions, and a cleaner cooking power source is already ments in battery storage and available in people’s homes. solar photovoltaic (PV) can environment. Doing so would make tackling now enable access in even household air pollution more the remotest corners of the effective at the local level, globe. Where the grid is reliable, powering energy-­ efficient as outdoor air pollution from cooking with biomass has also electric cooking (eCooking) appliances draws much lower been shown to have significant health impacts for the entire current and places less strain than conventional eCooking community (Das et al. 2018). appliances; it also reduces costs to users. Where the grid is not reliable enough to cook directly, trickle-­ charging a Cooking with electricity can present a transformative value battery acts as a buffer and time-shifts electricity demand proposition for households, allowing for more efficient and away from busy peak times, enabling users to cook during faster cooking times, multitasking, safer cooking, elimina- blackouts or brownouts. tion of dangerous indoor emissions, and a cleaner cooking environment. Focus group discussions in the four countries Lease-to-own solutions have achieved significant uptake in studied in detail in this report—Kenya, Tanzania, Zambia, and Africa, mainly for solar lighting (Lighting Global 2018). Like Myanmar—emphasize the aspirational nature of eCooking many renewable energy technologies, eCooking solutions, (see Table 2.1). They reveal that consumers’ focus is on the in particular battery-supported models, tend to be capital cleanliness of the process (no soot, less spillage, less burnt expense (CapEx) heavy. For eCooking, these innovative food, and less sweat), which leaves clothes clean at the end business models enable direct substitution of daily/weekly/ of the process, rather than its environmental and health monthly charcoal expenditure and a reframing of the impacts. concept of the battery-supported cooking device not as an improved cookstove but as a repurposing of household The extent to which people adopt new technologies expenditure to support the roll-out of electrical infrastructure in their daily routine is considered a make or break (whether national grid, mini grid, or off-grid PV), which could point for programs promoting clean cooking solutions. Background 5 Many programs have seen technologies dispersed to particular fuel to prepare the meal. Fuel stacking may also households but then set aside after a few months. eCooking occur when fuel is seen as an energy security issue. From has already seen widespread uptake in several developing a household perspective, charcoal, even if not cheap, is countries. This report examines two examples: Zambia and reliable, in the sense that it is pervasive and provides assur- Myanmar, where 12 percent and 4 percent of the population, ance that a household can cook its next meal. In the case respectively, are already using electricity as the primary of electricity, households are likely to fear that blackouts or cooking fuel (WHO 2017). However, cooking is a highly brownouts will leave them unable to cook. Therefore, they culturally embedded practice; it is yet to be seen whether need to have a secondary cooking method, unless reliable the many benefits of eCooking will be sufficiently attractive energy storage is available within the eCooking subsystem. and apparent to households to sustain use and change As with any new fuel, reliability will thus need to be estab- behaviors across a broader range of contexts. lished and demonstrated over a period of time to assure households that electricity can be used as their primary, or There is a need to create awareness and build the capac- even their only, source of energy for cooking. ity of households to sustain the shift in cooking practices. Research suggests that adoption and sustained change are However, fuel stacking can also strengthen the value eminently reachable, at least in urban areas (Batchelor et al. proposition of specialized eCooking appliances, which can 2019; Leary et al. 2019b; Scott et al. 2019). The behavioral enable households to take their first step toward cook- implications for switching to cooking with electricity will vary ing with electricity without having to take a leap of faith. widely, however, as a result of diversity in cooking practices Traditional fuels used within fuel-stacking behaviors may and the multitude of eCooking appliances available. have mixed effects on the health benefits from cooking with clean fuels. It is therefore important to understand the In many clean cooking programs, “fuel stacking” often extent of stacking practices. Fuel stacking with other clean persists even when clean fuels are available (Gould et al. fuels, such as liquified petroleum gas (LPG), to create a 2018; WHO 2016). Households often use a primary and a clean stack can also offer a highly attractive value proposi- secondary fuel (and sometimes others as well) based on the tion to everyday cooks, drawing on the unique advantages type of meal prepared and the perceived value of using a of each energy source. 6 ESMAP  |  Cooking with Electricity: A Cost Perspective eCooking using renewable energy can offer a viable assurance and safety, in particular live cooking demonstra- pathway to achieving sustainable cooking. In fact, of all the tions, will be critical in increasing adoption. pathways investigated by Jacobs et al. (2016) in Beyond Fire, eCooking on mini grids and solar home systems yields the In tandem with these demand-side interventions, technical greatest co-benefits, as it simultaneously enables access to and economic feasibility studies will need to be carried electricity for other applications. out to find the most appropriate solutions. Once they are developed, they will need to be made available at scale at Gender dynamics within households also need to be well an affordable basis. understood. Early responses in communities that switched to electric appliances in Tanzania suggest that the quick Many more people have reliable access to electricity than nature of eCooking is more attractive to men, which could are cooking with clean fuels. In Zambia, for example, 28–36 provoke a shift in responsibilities in the kitchen (Chepkurui et percent of the population has access to electricity but just al. 2019). Battery-supported cooking devices are particularly 17 percent cook with electricity. And households already likely to be popular with all members of the family, as they cooking with electricity can benefit from efficiency improve- also enable access to reliable electricity for other purposes, ments (by adopting more efficient appliances, for example), such as TV and lighting. which would ease the load on power systems and improve load-shedding scenarios. In Myanmar, mini grids should As cases from pilot studies highlighted in this report show, be able to provide reliable access to electricity and help most of the demand-side barriers described above can be increase the share of the population cooking with electricity overcome by demonstrating the safety, convenience, and (currently at 24 percent). In Kenya, less than 1 percent of affordability aspects of eCooking for popular local foods. people cook with electricity, 16 percent cook primarily with Improving the perception of safety will be vital in increasing LPG, and the rest of the population cooks primarily with adoption rates, as some households may not fully trust effi- polluting fuels and technologies. These figures indicate a big cient eCooking appliances, in particular the electric pres- opportunity for uptake of eCooking, where 26 percent of the sure cooker (EPC), which is associated with the mixed track population has at least Tier 4 electricity access (Table 1.1 and record of regular stove-top pressure cookers. Active work Figure 1.3).5 on consumer-oriented communication highlighting quality TABLE 1.1 Types of electric cooking physically possible with each tier of electricity access TIER TYPE OF eCOOKING 0 Solar electric cooking system is only option. 1 Small solar home systems and solar lanterns;upgrade to dedicated solar eCooking system is essential. 2 Solar home systems. Energy (minimum 200Wh) may just be enough for very efficient eCooking using an electric pressure cooker once a day, but power may be a bigger restriction (minimum 50W). Upgrade to dedicated solar eCooking system is advised. 3 Voltage fluctuations may affect performance of stoves; energy may restrict 100 percent eCooking (minimum 1kWh/ day), and power (minimum 200W) may be too limited for even energy-efficient eCooking appliances. Informal grid connections with poor-quality wiring may be used, so battery is advisable in most cases. 4 Energy is sufficient, but power limitations (minimum 800W) and reliability (minimum 16 hours/day, with up to 14 disruptions a week) may prevent some households from using off-the-shelf AC eCooking appliances. More efficient appliances with lower power ratings and small batteries may be required in some cases. 5 Energy, power, reliability, and availability are sufficient for off-the-shelf AC eCooking appliances. Battery is not required. Note: See appendix H for descriptions of the tiers in the Multi-Tier Framework. Source: Author estimates. Background 7 FIGURE 1.3 Access to electricity and clean cooking in Zambia, Myanmar, and Kenya Population by Tier of Electricity Access 2% 2% Zambia 60% 8% 7% 21% Myanmar 30% 19% 12% 9% 17% 13% Kenya 46% 11% 5% 12% 12% 14% Tier 0 Tier 1 Tier 2 Tier 3 Tier 4 Tier 5 Population with access to Clean Cooking Zambia 83% 17% 1% Myanmar 75% 24% Kenya 84% 16% Polluting Fuels and Technologies LPG Electric Cooking Note: Clean fuels such as biogas or ethanol constituted less than 0.2 percent of the total and are not included. See appendix H for descriptions of the tiers in the Multi-Tier Framework. Source: Author estimates using data from MTF reports. 8 ESMAP  |  Cooking with Electricity: A Cost Perspective been able to support it, most appliances are rated at a 1.5.  A New Generation of relatively high power level. The power level is potentially a key constraint for battery-supported eCooking systems, Highly Efficient eCooking as the more rapidly a battery is discharged, the shorter its life. Mass-production of DC eCooking appliances has Appliances already begun (see, for example, Tesga Power 2019). They typically have power ratings of 25–50 percent of their AC equivalents. A new generation of energy-efficient eCooking appliances is available. Many of these devices are highly efficient at a Reducing overall energy consumption is also important, as specific task (for example, kettles for water boiling) and must it affects the cost of cooking. Simply preventing heat from therefore be combined with other appliances to cook the escaping from the cooking chamber using insulation can range of foods that make up a full menu. Induction stoves enable the same food to be cooked with a fraction of the are gaining in popularity as a result of their versatility for energy. This feature becomes very important when oper- a wide range of dishes and ability to heat a pot directly ating from battery storage with a limited capacity. Several through magnetic induction, making the heat source as groups of researchers (Batchelor et al. 2018; Watkins et responsive as gas (Parikh et al. 2020). Large-scale programs al, 2017) have already used this principle to create highly have been set up to facilitate the adoption of induction insulated environments and feed in a trickle of solar elec- stoves in a variety of developing country contexts, including tricity to create a very low-cost solar eCooking system (see India and Ecuador. In many of these trials, uptake has not Section 3.4 for details). been as promising as initially hoped (Banerjee et al. 2016; Gould et al. 2018). Although induction stoves are efficient at Figure 1.4 compares the eCooking appliances featured in this transferring heat to the pot, heat leaves the pot just as easily report, categorizing them as inefficient conventional, more as it does on a conventional stove. efficient, and most efficient modern appliances. A drawback of many of the highly efficient eCooking solutions is that In contrast, an EPC uses other mechanisms—insulation, most require specifically shaped and sized pots and pans, automatic control. and pressurization—to significantly reduce made of compatible materials. These standardized shapes, energy demand to an extent that is not possible with any sizes, and materials may present a challenge for households other fuel or appliance. As Batchelor et al. (2018, p. 1) note, that use different pots for different foods and want flexibility “It is temperature that cooks food,” not energy per se. in this sense when cooking. Therefore, raising the boiling point of water and stopping heat from escaping from the pot, rather than just raising the efficiency of converting energy from a fuel into heat in the pot, are extremely effective ways of improving the efficiency of the cooking process. The benefits of using EPCs are most significant when cooking dishes that require boiling for half an hour or more. However, the need to pressurize the pot to accelerate cooking times reduces the cook’s access to the dish to stir and check on progress. Initially, this drawback can present a significant psychological barrier for users. However, once overcome, this feature can actually be an asset, freeing up the cook’s time to perform other tasks while food is cooking, as the device’s automatic control mechanisms and sealed cooking chamber mean it requires minimal supervision. Other devices, such as rice cookers and slow cookers, also embody the insulation and automatic control elements, allowing cooks to multitask and occasion- ally check on the food if they want to (see appendix C for a discussion of assessing electricity demand for cooking). As households in developed economies (where marketing for eCooking devices has focused to date) have gener- ally prioritized fast cooking and electricity networks have Background 9 FIGURE 1.4 Assessment of eCooking appliances featured in this report HEAT HEAT TYPICAL TOTAL APPLIANCE TRANSFER TRANSFER POWER COOKING TIME VERSATILITY INTO POT OUT OF POT REQUIREMENTS (incl. preheating) Ine cient conventional appliances Electric oven Cooking chamber insulated, but not Baking, Convection sealed; whole oven 1–5kW Slow roasting, space around grilling only pot/dish heated Hotplate Convection and Any pot Conduction when radiation from 1–2kW per hotplate (round bottom pot in contact uninsulated pot; Average (DC: 300–700W) di cult); frying with element evaporation and boiling without lid More e cient modern appliances Induction/ Any Convection and infra-red stove flat-bottomed radiation from Induction/ Fast frying and (ferrous for uninsulated pot; 1–2kW per hob radiation bringing to boil induction) pot; evaporation frying and without lid boiling Most e cient modern appliances Rice cooker Insulation and Single deep Conduction fixed lid, but 300W–1kW pot only; via insulated Average not completely (DC: 200–400W) boiling and element sealed some frying Insulated electric Conduction Single shallow frying pan Insulation; via insulated Fast frying and pot only; evaporation 700W–1.5kW element bringing to boil frying and without lid stuck to pan boiling Electric pressure cooker Insulation and Single deep Conduction via Very fast fixed lid; 700W–1.2kW pot only; insulated (pressurized) completely (DC: 200–400W) boiling and element boiling sealed some frying Advantage over No particular advantage Disadvantage compared other appliances over other appliances with other appliances Note: For a broader range of appliances, see appendix D. 10 ESMAP  |  Cooking with Electricity: A Cost Perspective Increasing generating capacity alone will not solve the elec- 1.6.  Electrical trification problems. Substantial investment in transmission and distribution infrastructure will be needed to contribute Infrastructure in to grid electrification and to realize competitive electricity costs. In many Sub-Saharan Africa contexts, distribution and Sub-Saharan Africa transmission infrastructure remains poor, presenting a huge challenge for channeling increased generation capacity to end-users. The distribution segment of a power system The proposition that electricity could be used for clean is closest to the end-consumer; there are many stories of cooking is deeply integrated with progress on electrifica- communities “under the grid,” where transmission lines are tion. Conceptualizing cooking as a part of the investment in visible but communities remain unconnected. electrification and in modern energy more broadly, including investment in renewable energy, could spur more progress Although grid extension–based electrification has long been toward eCooking in the years to come. The case studies regarded as the reference model in developing economies, described in this report highlight examples in which grid and the private sector is spearheading the design of innovative electricity supply models based on off-grid technologies. off-grid capacity increased substantially in recent years. This section briefly overviews the broader state of investment in Beyond grid connections, decentralized generation and electrical infrastructure, particularly in Sub-Saharan Africa.“prosuming” (consumers producing their own electricity) where the majority of people without access reside. are two practices changing the landscape and increasing the pace of electrification in Providing universal access ways that grid extension has to affordable, reliable, not been able to do. Cost In many Sub-Saharan African sustainable, and modern ­ reductions of renewable energy for all remains an contexts, distribution and transmission technologies and improved ambitious goal. The popula- infrastructure remains poor, presenting a reliability make off-grid tion of Sub-Saharan Africa is technologies, notably stand- ­ expected to double by 2050; huge challenge for channeling increased alone systems and mini grids, by 2030, electricity supply generation capacity to end-users. reliable alternatives to grid across Africa will need to power infrastructure. triple to meet the demand from demographic growth in ESMAP (2019a) estimates that mini grids could these economies and their changing lifestyles and expec- cost-effectively supply half a billion people in Africa and Asia tations. Of the 840 million people remaining without access with electricity. These solutions hold potential in peri-urban to electricity globally, about 573 million or 68 percent are in and rural contexts characterized by limited, sparse demand Sub-Saharan Africa (ESMAP 2020a). and lower ability to pay. Fifty-seven percent of planned mini grids are based on solar-hybrid technologies. They Africa could meet a quarter of its energy demand by increas- aim to connect more than 27 million people globally, at an ing its renewable capacity to 310GW by 2030, up from investment cost of $12 billion (ESMAP 2019a). As of 2019, 42GW available in 2017 (IRENA 2020). The Africa Renewable cumulative mini grid investments in Sub-Saharan Africa and Energy Initiative aims to mobilize investment for at least South Asia were about $5 billion (ESMAP 2019a). Mini grid 300GW of renewable energy generation by 2030.6 Energy models are evolving, from providing only basic electricity trade through regional power pools is a core part of the long- services for households to providing electricity services for term strategy for increasing distributed generation and the income-generating activities. share of renewable energies. A simple calculation suggest that 1.5kWh consumption per household per day spread out In 2017, Lighting Global (2018) estimated that the off-grid over 10 hours of trickle-charging a battery would require just solar sector was providing electricity access to 73 million 32GW of additional peak load capacity to enable eCooking households worldwide. Most providers started with for Sub-Saharan Africa’s 900 million households currently basic lighting and phone-charging, increasingly using without access to clean fuels. If the battery could be charged prepaid mobile payments and other pay-as-you-go (PAYG) during off-peak hours, this generating capacity may already approaches, which are now also making larger systems exist. AC cooking without household batteries would require and DC-powered energy-efficient appliances more afford- 214GW of additional peak load capacity, assuming energy-­ able. The development of mobile money enabled many efficient appliances with a 1kWpeak demand per household. companies to reduce the costs associated with bill recovery Background 11 in remote rural areas while maximizing affordability and Kenya, Rwanda, Tanzania, and Uganda, and a handful of new responding to customers’ need to make small regular companies appear every year in neighboring countries. As payments. The Global Off-Grid Lighting Association (GOGLA) of December 2019, close to 1 million solar home system units reported that the off-grid solar industry sold 4.4 million had been sold in Kenya (GOGLA 2019). off-grid solar lighting products and 460,000 appliances in the first half of 2019 (GOGLA 2019). The number of PAYG It would be a missed opportunity not to consider integrating solar home systems sold in Kenya alone is about to reach eCooking into the planning for each of these electrification 300,000 kits per year—about equivalent to the annual modes, all of which are gaining momentum, to make faster growth in rural households. More than 30 PAYG solar compa- progress toward addressing both the clean cooking and elec- nies are now operating in the peri-urban and rural areas of trification challenges of the SDG7 goals. 12 ESMAP  |  Cooking with Electricity: A Cost Perspective Ch apter 2 METHODOLOGY 2.1.  Techno-Economic Modelling A key ambition of this report is to explore the opportunities Table 2.1 breaks down the activities carried out in each for cooking with electricity using different technological c ­ ountry, highlighting the most relevant output data from approaches and in a variety of contexts. It uses a tech- each. Appendix C outlines the key findings. For the full no-economic model to estimate the monthly costs of cooking analyses, see the MECS Working Papers indicated in ­ in various scenarios. Table 2.1 or the summaries of the key opportunities and ­ challenges in each country (Batchelor et al. 2019; Leary et al. Leach and Oduro (2015) developed a numerical simulation 2019; Scott et al. 2019). model of cooking by a household linked to a system design model for a battery-supported eCooking device (either a The daily energy demand figures from the cooking diaries stand-alone solar-powered or a grid-connected device).7 were used to calculate the eCooking system size and any Their model is a simple proof of concept simulation, charac- additional fuel use needed. The model follows a traditional terizing the technologies and cooking energy requirements approach of discounted cashflow analysis, accounting from secondary data. It included a single eCooking appli- for initial capital costs of cooking appliances (and where ance, the hot plate. applicable, supporting energy storage and power genera- tion system components); their replacement costs at end This model was recently updated to model a wider range of component life; and operating costs, with the output the of eCooking appliances and to reflect current cost trends levelized monthly cost of cooking.8 of the major components. An empirical model for battery degradation was also added, capturing the high current The model was applied in different scenarios, to explore drain of cooking and the likely high ambient temperatures of opportunities for cost-effective eCooking in a range of the system in use. Appendix E outlines the model (for more contexts. It was also used to explore assumptions about detail, see Leach et al. 2019). technology performance and cost, the business model employed, and two points in time (2020 and 2025). The techno-economic model of the eCooking system was applied to represent the daily cooking requirements The results are used to explore the economic viability of of a household based on data on meals, cooking habits, eCooking compared with use of traditional fuels. They are and fuel use collected in 2017 and 2018 through a series presented in the following sections as case studies, clus- of studies undertaken as part of a project funded by tered by access to electricity. Although the analysis includes Innovate UK, Gamos, and UK Aid designed to assess a wide range of parameters and assumptions, and wider the opportunities and challenges that lay ahead for generalization of the results is discussed, it still explores only eCooking in high-impact potential markets. The country a subset of the possibilities, as an initial scoping. Section 6.2 studies were conducted in Kenya, Zambia, Tanzania, and suggests areas for further development. Myanmar. Methodology 13 TABLE 2.1 Summary of data collection methodologies used in this report NUMBER CARRIED DATA OUT IN NUMBER OF COLLECTION EACH PARTICIPANTS METHODLOGY COUNTRY PER ACTIVITY KEY OUTPUT DATA MECS WORKING PAPER Cooking diary 1 20 Energy demand for cooking, Leary, Scott, Sago, et al. (2019); studies compatibility of local cooking Leary, Scott, Numi, et al. (2019); practices with eCooking appliances, Leary, Scott, Hlaing, et al. (2019); fuel prices Leary, Scott, Serenje, et al. (2019a) Household 1 200 Fuel prices, fuel choices, verification Scott, Leary, Serenje, et al. (2019); surveys of monthly expenditures Scott, Leary, Sago, et al. (2019); Scott, Leary, Hlaing, et al. (2019); Scott, Batchelor and Jones (2019) Focus groups 4 5–15 Identification and exploration Leary, Serenje, Mwila, et al. (2019); of opportunities for eCooking, Leary, Win, Myint, et al. (2019); Leary, compatibility of local cooking Scott, Sago et al. (2019); Chepkurui, practices with eCooking appliances, Leary, Numi, et al. (2019) fuel choices, verification of monthly expenditures Stakeholder 1–2 20–60 Identification and exploration of REAM et al. (2018); Leary, Mwila, workshops opportunities for eCooking Serenje, et al. (2019); Villema et al. (2018); Chepkemoi et al. (2019) ESMAP  |  Cooking with Electricity: A Cost Perspective Throughout the report references to “AC cooking” describe 2.2.  System Architectures the first category (cooking on an AC grid using appliances without battery support). for eCooking in Strong- The report focuses on electricity stored in chemical batter- Grid, Weak-Grid, and ies. Other technologies, including the use of phase-change materials for thermal storage, are possible. A number of Off-Grid Contexts groups are working on prototyping thermal storage for a low-cost stand-alone system. These efforts are at an early stage of development, however, and most details are not in Table 2.2 categorizes the range of system architectures that the public domain. They are therefore not included in the can enable eCooking in strong, weak, and off-grid contexts. modelling. This report focuses on three architectures: ● off-the-shelf AC eCooking appliances that can be connected directly to strong grids STRONG GRIDS ● hybrid battery-supported appliances that can run on AC cooking is possible on strong grids, mainly national grids both AC and on DC (via the battery), suitable for use on but potentially also larger hydro-powered grids or solar weak grids hybrid mini grids with significant battery storage. When feasi- ● DC appliances that can be connected directly to ble, it is likely to be the most cost-effective solution (unless a battery charged by a solar panel in an off-grid the grid tariff is extremely high), as it involves simply plug- system. ging in off-the-shelf AC appliances. TABLE 2.2 Simplified typology of eCooking devices for strong, weak, and off-grid settings USE OF BATTERY GRID OR MINI GRID SOLAR HOME SYSTEM Without battery Strong grid Off-grid AC grid eCooking DC solar eCooking + + Battery- Weak grid Off-grid supported DC grid battery-powered eCooking DC solar battery–powered eCooking + + + + Note: For a more complete set of definitions, see appendix B. For the system architectures modelled in this report, see table 2.4. Methodology 15 WEAK GRIDS (INCLUDING SOME MINI GRIDS) 2.3.  Cost Trends Battery-supported cooking can enable eCooking in weak grid and other mini grid contexts, with a charger used to recharge the battery. Off-the-shelf AC appliances can be Although other power generation and energy storage used via an inverter or DC appliances that can be connected technologies exist, PV panels and lithium-ion batteries are directly to the battery (using the built-in battery management particularly important for expanding access to eCooking. system). Appliances can be configured to run solely from Solar PV is the most readily deployable technology for both the battery or from the grid when available and the battery stand-alone systems and mini grids across the global South. when not (in a configuration similar to that of an uninterrupt- Lithium-ion batteries are particularly well suited for eCook- able power supply). Alternatively, a hybrid AC/DC appliance ing, because they offer a longer cycle life and greater depth can eliminate the need for an inverter, the configuration of discharge and can tolerate more rapid discharge than lead modelled in this report. acid batteries, the standard off-grid electricity storage for over 100 years (Batchelor 2015; Zubi et al. 2017). Lead acid batteries and their variants have been improved significantly in recent years and continue to be the most cost-effective STAND-ALONE SOLAR SYSTEMS solution for many off-grid applications. However, they are not generally recommended for continuous high-drain applica- For remote off-grid regions, this report focuses on tions such as cooking, because such conditions reduce both solar-powered battery-supported eCooking. For isolated their lifetime and usable capacity (Battery University 2019). off-grid households, stand-alone systems are the only option. They are likely to be powered by solar, although other gener- Figure 2.1 shows the price trends for PV modules and ation sources, such as small-scale wind or pico-hydro, could lithium-ion battery packs. For PV, the World Bank’s Off-Grid also be employed. DC cooking appliances can be connected Solar Market Trends Report 2018 anticipates some price directly to the solar panels to enable cooking during sunny stabilization, but the International Renewable Energy Agency periods (this option is not modelled in this study but is (IRENA 2019) projects continuation of the recent price explored in Section 3.4. reduction trends. The price of a lithium-ion battery pack fell FIGURE 2.1 Actual and projected prices for PV modules and lithium-ion battery storage, 2010–22 a. Crystalline silicon solar module b. Lithium-ion battery pack 2.5 2.39 1,000 1,000 –20% 2.08 –25% 2.0 800 800 689 Price ($/kWh) 599 Price ($/W) 1.5 600 540 1.06 350 1.0 400 0.73 0.78 273 0.61 0.55 209 200 0.5 0.46 0.33 0.33 200 125 0 0 0 12 0 12 13 13 14 15 16 (f ) 11 (f ) 11 14 15 16 (f ) (f ) 17 17 1 1 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 22 20 20 22 20 20 20 20 Note: These figures are for commercialized PV and battery technologies and chemistries. Different technologies may well take over in future, but they would likely do so mainly because of lower cost, so the cost trajectories should be largely technology independent. (f) = forecast. Source: Lighting Global (2018). 16 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 2.3 Parameter values used in high- and low-cost scenarios for eCooking systems 2020 2025 LOW-COST HIGH-COST LOW-COST HIGH-COST PARAMETER VALUE VALUE VALUE VALUE Battery price (lithium-ion, $/kWh) 280 350 180 220 Usable maximum capacity remaining at 80 90 80 90 replacement (%) Battery life (cycles) 3,000 2,000 3,000 2,000 PV-battery roundtrip efficiency (%) 90 85 90 85 Price of fuel 2/3 of 2018a mean 4/3 of 2018 mean 2018 low value + 2018 high value + value value 3% a year 3% a year Note: All financial values are in 2018 U.S. dollar prices. a. Some values are from late 2017 or early 2019. by an average rate of 20 percent a year between 2010 and Battery Price Survey (Goldie-Scot 2019), which includes Bloomberg New Energy Finance’s forecasts to 2030 for 2017, and reductions at a similar rate are forecast for the next five years (Goldie-Scot 2019). However, these projections are lithium-ion pack prices, which are assumed to be at scale for for PV panels and battery packs alone; they do not include electric vehicle applications. A 51 percent premium is applied transport, installation, and supply chain margins, which to reflect the higher costs of packaging stationary batteries are unlikely to decline as rapidly as the core technology, and the economies of scale seen in the very large electric suggesting that the rate of cost reduction for PV and vehicle market, as Frith (2017) suggests. Another 20 percent is added to reflect the costs for transportation to and impor- batteries in use will be lower. (See appendix E for analysis of cost trends and the detailed assumptions and calculations tation into Africa. These assumptions lead to the low-cost made for the modelling. See appendix F for a summary of value in Table 2.3. The high-cost value reflects a more the data inputs for each of the cases.) pessimistic view of the stationary battery market, either a battery price reduction profile This report presents a that lags by about two years comparison between the or an additional 25 percent costs of eCooking in 2020 premium on battery costs for The price of a lithium-ion battery pack and 2025, exploring two implementation in-country. expected trends: reducing fell by an average rate of 20 percent The other parameters for costs for battery-supported a year between 2010 and 2017, and battery and system efficiency eCooking through technical reflect technical uncertainty and organizational learning; reductions at a similar rate are forecast about the real-life perfor- and increasing charcoal, LPG, for the next five years. mance of these systems. firewood and kerosene prices. Product sales, marketing, and The analysis incorporates local distribution costs are not some simple parameter value uncertainty, reflecting more included in this modelling. Throughout the modelling, finan- optimistic and more pessimistic outlooks for the key vari- cial values are in 2019 dollars. Appendix E provides other ables in performance and cost (table 2.3). technical and cost assumptions. Assumptions about battery price and performance strongly Fuel prices for each country are based on the results from influence the overall costs of battery-supported eCooking household surveys carried out alongside the cooking diary systems. Battery prices are from The 2019 Lithium-Ion studies in 2017–19 (Scott, Leary, Hlaing, Myint, Sane, Win, Methodology 17 Phyu, Moe, Batchelor, et al. 2019; Batchelor, Leary, et al. 2019; Leary, Scott, Serenje, et al. 2019b; Leary, Scott, Numi, 2.4.  Summary of et al. 2019), plus an assumption of a 3 percent price increase each year thereafter. A high/low range is then applied Contexts, Systems, and around these values by adding/subtracting one third. The price range assumptions are based on observations during Fuel Prices the cooking diary and related studies in East Africa and Myanmar. The price trends reflect an assumption that there will be increasing pressure on charcoal production through By inputting local fuel/electricity prices and the quantity policy initiatives for environmental protection (see case study of each that is required to cook local foods, the tech- 1) and upward pressure on fossil oil prices internationally, no-economic model can compare the relative costs of affecting LPG. These assumptions are not the result of a cooking with each type of fuel across a range of system comprehensive analysis. Some sensitivity analysis to key architectures. assumptions is reported in the case study results, but further exploration of fuel price trends will be important in future Table 2.4 summarizes the contexts and the system architec- work. tures modelled. For the national and mini grid architectures, the tariff is the key cost variable; for the stand-alone archi- tecture, the price of PV modules and the solar resource are fundamental. Battery price and performance are important for all battery-supported cases. 18 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 2.4 System architectures and modelling parameters in each case study context Methodology MODELLED APPLIANCES MODELLED FUEL PRICES 50 ELECTRICITY CASE SYSTEM 100 PERCENT PERCENT TARIFF ($/ FIREWOOD CHARCOAL KEROSENE DATA STUDY ARCHITECTURE ECOOKINGa ECOOKING KWH) ($/KG) ($/KG) ($/LITER) LPG ($/KG) SOURCE Urban 1: Nairobi, National AC and Appliance stack Electric 0.17 n.a. 0.49 1.1 1.08 Cooking Kenya grid battery- of efficient pressure diaries (n = 20) supported and inefficient cooker (EPC) and utility DC appliances only (KPLC 2019a) 2: Lusaka, National AC and Appliance stack EPC only 0.01 n.a. 0.21 n.a. 2.07 Household Zambia grid battery- of efficient survey supported and inefficient (n=200), DC appliances cooking Baseline only: diaries (n=20) Integrated four-plate and utility cooker with oven (ZESCO 2019) Rural 3: Shan State, Mini grid AC and Appliance stack EPC only 0.16 0.13 0.15 0.82 1.08 Household Myanmar battery- of efficient survey (n = 200), supported and inefficient interview DC appliances with mini grid developer 4: Kibindu, Mini grid AC and Appliance stack EPC only 1.35 0.04 0.13 0.82 1.16 Household Tanzania solar of efficient survey (n = 200), battery– and inefficient interview supported appliances with mini grid DC developer 5: Echariria, Off-grid Solar Appliance stack EPC only n.a. 0.13 0.31 1.18 1.33 Interviews with Kenya battery– of efficient community supported and inefficient members and DC appliances household survey (n = 200) Note: a. See table 2.5 for details on appliances tested in each country context. The blend of inefficient and efficient appliances was modelled as two appliances per household: an EPC and a hot plate. See appendix E for details. n.a.: Not applicable. 19 intensity and frequency, and the compatibility of local 2.5.  Demand for cooking practices with the broad range of energy-efficient eCooking appliances available. Electricity for Cooking The cooking diary studies tracked the energy use and cooking practices of 80 households in 4 countries over 6 As cooking is a highly culturally specific practice, the type weeks. They generated qualitative data on the compatibility of appliances and the amount of power required varies of various electric appliances with local cooking practices significantly. Very little information is available on how much and quantitative data on energy demand for cooking. Tables energy is needed to cook a meal, in particular how much 2.5 and 2.6 list the figures used as inputs for the modelling electricity and how it varies across cultures. Apart from the in this report. Further details on the cooking diary studies availability of electricity, the viability of eCooking solutions can be found in appendix C and the cooking diary country depends on the cooking processes employed, the cooking reports (see Table 2.1). 20 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 2.5 Measured energy consumption for eCooking and modelling assumptions PROPORTION OF ENERGY CONSUMED BY ELECTRIC 100 PERCENT ELECTRICITY PRESSURE COOKER (EPC) INPUT DATA FOR CASE (MEASURED DURING COOKING COOKING 50 PERCENT OF STUDY MODELLING DIARY STUDIES) MEALS (MODELLED) (4.2 PEOPLE PER HH)c Median Median daily daily per Median daily Median daily 100 percent 50 percent Number of energy per Mean HH capita energy per per capita eCooking eCooking COUNTRY APPLIANCESa daysb HH size energy HH (kWh) energy (kWh) (kWh) (kWh) Kenya EPC, rice 431 1.4kWh 3.1 0.46kWh 0.47 0.15 1.92 0.64 cooker, hot (5.1 MJ) (1.65 MJ) plate Myanmar Rice 476 1.02kWh 4 0.26kWh 0.34 0.09 1.08 0.36 cooker, (3.7 MJ) (0.93 MJ) induction stove, infra-red stove, EPC, thermo-pot Tanzania Rice 423 2.06kWh 4.2 0.49kWh 0.69 0.16 2.06 0.68 cooker, (7.4 MJ) (1.76 MJ) induction stove, hot plate, EPC, thermo- pot, kettle Zambia EPC, hot 99 1.63kWh 7.9 0.21kWh 0.55 0.07 0.87 0.29 (efficient and plate (5.9 MJ) (0.75 MJ) inefficient appliances) Average for efficient and inefficient appliances 1.48 0.49 Zambia (very Integrated 494 2.47kWh 3.3 0.75kWh 0.82 0.25 3.15 1.05 inefficient four-plate (8.9MJ) (2.71MJ) appliances cooker and with oven practices) Note: Tables shows measured energy consumption for 100 percent eCooking on a mixture of inefficient and efficient appliances and modelled energy consumption for 50 percent eCooking on EPCs (assuming the other 50 percent is met by traditional fuels). HH = household. a. Households may also have used other eCooking appliances they already owned. b. Number of days refers to number of days of data using only that fuel. c. Figure refers to number of household members cooked for (mean of means). Fuel stacking is a complex behavior that allows households measured median daily energy consumption figures for to optimize cost, usability, and other factors by using multiple 100 percent eCooking presented in Table 2.5 by one third. cooking fuels in their household. Appendix C provides a This proportion is derived from analysis of the cooking diary detailed explanation of how it was modelled in this study. data, which show that with minimal training, participants The electricity demand of a typical household stacking an with a hot plate and an EPC chose to cook 50 percent of EPC with its traditional fuel is estimated by multiplying the the menu on an EPC. Across the range of foods households Methodology 21 TABLE 2.6 Normalized energy consumption cooking with traditional fuel, by fuel type FIREWOOD (KG) CHARCOAL (KG) KEROSENE LPG (KG) 100 50 100 50 100 50 100 50 COUNTRY PERCENT PERCENT PERCENT PERCENT PERCENT PERCENT PERCENT PERCENT Kenya 3.50a 1.75a 1.75b 0.87b 0.25 kg 0.12 kg 0.23 0.11 (0.31 liters) (0.15 liters) Myanmar 1.54 0.77 n.a. n.a. n.a. n.a. 0.20 0.10 Tanzania 3.50a 1.75a 1.75 0.87 n.a. n.a. 0.33 0.16 Zambia n.a. n.a. 1.04 0.52 n.a. n.a. 0.17c 0.08c Note: Data are from cooking diary periods for traditional fuel use only. Values are normalized to a 4.2 person household. For measured values of median energy consumption for each fuel, see table C.2. n.a.: Not available. a. Firewood data were not available in the Kenya or Tanzania cooking diary datasets, so consumption data were estimated using the ratio of firewood to charcoal energy consumption (approximately 1:1) from Myanmar and the ratio of firewood to charcoal energy density (approximately 1:2) from appendix C. b. Insufficient records were available for charcoal cooking from the cooking diary study in Kenya to make a reliable measurement of charcoal consumption. As cook- ing practices in Kenya are similar to practices in Tanzania and electricity consumption was measured to be similar in the two counties, the values for charcoal cooking in Tanzania were used as model inputs for Kenya. c. LPG data were not available in the Zambia cooking diary datasets, so consumption data were estimated using the average of the Tanzania to Zambia ratios for charcoal and eCooking with energy-efficient appliances (approximately 2:1). chose to cook using it, the EPC used an average of Figure 2.2 shows the three fuel-/appliance-stacking 50 percent of the energy of the hot plate. The accompa- scenarios modelled in this report: nying energy for traditional fuels in fuel-stacking scenarios is therefore simply half of the measured median figures ● Household cooking: 100 percent electric, stacking of presented in Table 2.5. inefficient appliances (for example, hotplate); more effi- cient appliances (for example, electric frying pan); and In this report, a household is modelled as comprising most efficient appliances (for example, EPC) 4.2 people, the average size in the cooking diaries. The ● Household cooking: 50 percent electric, stacking most charcoal, LPG, firewood, kerosene, and electrical energy use efficient appliances (for example, EPC), with baseline data for each country relate to households of different sizes fuels for remaining 50 percent and are scaled linearly from the median per capita energy values. ● Cooking as a microenterprise: Use of EPCs for boiling “heavy foods” only.9 Case Studies 1–5 model the first two options. Part 2 of case study 4 explores the third option, the most efficient form of eCooking. Case study 2 also models 100 percent eCooking with very inefficient appliances and practices. 22 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE 2.2  Energy storage required to support eCooking with different appliances, practices, and fuel-stacking options CASE STUDY 2 Very ine cient appliances & practices 3 kWh NOT MODELLED IN THIS REPORT Ine cient appliances only 2 kWh HOUSEHOLD COOKING CASE STUDIES 1 5 Appliance stack: e cient & ine cient 1.5 kWh Fuel stack: EPC with other fuels 0.5 kWh + CASE STUDY 4 PART 2 EPC for boiling heavy foods once a day 0.15 kWh COOKING AS A MICRO ENTERPRISE Methodology 23 In grid-connected scenarios, the plots of modelling results 2.6.  Business Models and in the next section show both battery-supported eCooking, with the battery charged from the grid/mini grid, and direct Financing Horizons eCooking, where the only capital cost to be financed is the cost of the appliances. Each of the two business models could be applied to any of the battery-supported and AC This study examines both utility and lease-to-own business eCooking scenarios. However, the 20-year financing option models. For utility, or energy service, business models, for AC eCooking seems highly unlikely. To reduce complex- the costs of cooking on each system are calculated over a ity, it was therefore omitted from the results charts. For AC repayment period spanning an expected 20-year system eCooking, financing a $50 EPC even over a period as long lifetime. For the lease-to-own business model, cooking costs as five years seems unnecessary, but this business model are calculated for recovery of capital costs over a five-year assumption was applied to make the cost results comparable financing period. At the end of five years, users could be left to those for battery-supported cooking. With the exception with the device. However, for the battery-supported solu- of Zambia (case study 2), where tariffs are exceptionally low, tions, they would face ongoing costs for the replacement of the effect of moving to a one- or two-year financing period the battery and other components. Creative business models for direct cooking should not be significant, as the major cost will therefore be needed (see Chapter 4). The real discount will be electricity units. rate applied for both business models is 9.6 percent.10 24 ESMAP  |  Cooking with Electricity: A Cost Perspective Ch apter 3 ECOOKING IN GRID-CONNECTED AND OFF-GRID SYSTEMS: MODELLING RESULTS AND DISCUSSION 3.1.  Overview of Case Studies Five case studies were chosen to illustrate the range of structure of tariff bands. Kenya Power, for example, now opportunities available for eCooking in both urban and rural has surplus generation capacity and is looking to increase areas utilizing power from national grids, mini grids, and solar demand for electricity, which is currently barely used for home systems (Figure 3.1). Leary et al. (2018) carried out a cooking. At the same time, the government of Kenya issued global market assessment to identify high-potential markets a logging ban in 2019 to protect the country’s dwindling for battery-supported eCooking. Cooking diary studies were forest reserves, causing charcoal prices to double overnight. carried out in four high-scoring countries to gather empiri- In contrast, although electricity is already the aspirational cal data on electricity demand for cooking local foods with cooking fuel in Zambia, the national utility (ZESCO) has efficient and inefficient appliances. repeatedly been forced to carry out load shedding over the past few years, as late rainfall has severely limited genera- The case studies draw attention to opportunities identi- tion capacity on its hydropower-dominated grid. fied during the course of the cooking diary research. They feature a range of system architectures from Table 2.2, illus- The first case study explores an opportunity for East Africans trating the broad range of solutions that have the potential to to transition completely away from biomass by fuel stacking achieve social impact by enabling eCooking on stand-alone LPG with an EPC. LPG is currently the aspirational fuel across systems, mini grids, weak grids, and reliable grids. Sections most of East Africa yet many households with an LPG stove 3.2–3.4 present the results of the case study analyses; still purchase charcoal to cook “heavy foods” such as tripe, Section 3.5 cuts across the case study results, drawing out which they believe is cheaper (Leary, Fodio Todd et al. 2019). generalizable findings on the cost-effectiveness of eCooking compared with current practice. The second case study illustrates an opportunity for coun- tries with significant populations already cooking with The section on eCooking on national grids explores the electricity but using inefficient appliances, to optimize dynamics of cooking with grid electricity in Nairobi, Kenya loading on their grids. The evidence from the cooking and Lusaka, Zambia. The motivations for encouraging or diaries shows that for households currently cooking on managing the uptake of eCooking on urban grids in African highly inefficient appliances (such as integrated four-plate cities vary substantially, as do electricity tariffs and the cookers with ovens), energy-efficient eCooking appliances eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 25 FIGURE 3.1 Comparison of the five case studies and rationale for selection DEMAND SIDE: KEY OPPORTUNITY ENERGY CASE CONTEXT SUPPLY SIDE BASELINE FUELS/ TO ENABLE 100% STORAGE LOCATION APPLIANCES CLEAN COOKING CONSIDERED LPG, charcoal Clean fuel stack: LPG and kerosene and most e cient Urban, electric appliances national grid 1 Stimulate demand (EPCs) for surplus national None Nairobi, grid electricity Kenya Ine cient electric Most e cient (EPCs) Urban, Household Mitigate load appliances (hotplates, and minimal use of national grid battery 2 shedding on oven) and charcoal less e cient appliances national grids (hotplates, oven) Lusaka, with energy Zambia storage Rural, micro- Firewood and e cient Only e cient electric hydro mini-grid electric appliances appliances (induction Mitigate peak (induction stove, rice stove, rice cooker and Household cooker and insulated insulated electric battery 3 loading constraints on electric frying pan) frying pan) Shan State, micro hydro Myanmar mini-grids with energy storage Rural, solar hybrid mini-grid Clean fuel stack: LPG Charcoal and firewood and most e cient Centralized electric appliances battery bank 4 Stimulate demand for electricity in rapidly (EPCs) Kibindu village, growing solar-hybrid Tanzania mini-grid sector Charcoal, kerosene Clean fuel stack: LPG Rural, o -grid LPG and firewood and most e cient Household electric appliances battery 5 Enable electricity access and clean (EPCs) Echariria cooking with solar village, Kenya systems 26 ESMAP  |  Cooking with Electricity: A Cost Perspective and practices can reduce electricity demand for cooking by two-thirds (see Table 2.5). Supporting eCooking appli- ances with a battery can also time shift energy demand for cooking and reduce peak loading, as well as enable custom- ers to cook during blackouts or load shedding. The section on eCooking on mini grids contrasts a micro-­ hydro mini grid with a relatively low tariff in Myanmar (where users are already cooking with electricity at off-peak times) with a solar hybrid mini grid in Tanzania with a tariff an order of magnitude higher. However, with the rapidly falling prices of batteries and solar PV, new opportunities are opening up for integrating energy-efficient eCooking into a broader range of systems. Mini grids are usually installed in areas where biomass fuels can be collected or purchased at very low cost. However, urbanization is causing many people who used to collect fuel to start paying for it, creating an opportunity to translate expenditures on biomass fuels into electricity units, which could drive down the tariff for the mini grid as a whole. Peak loading is a major concern for eCook- ing on power-limited mini grids, but many are already using a variety of time-shifting techniques, which could decouple cooking from overall electricity demands on the mini grid, smoothing out the load profile and further reducing the unit cost.11 The third case study, on Myanmar, highlights the opportu- nity for micro-hydro mini grid developers who have already enabled cooking on their systems to allow their customers to do all of their cooking with electricity. At peak times, grids often reach capacity and the voltage dips. Some innovative mini grid developers have been able to enable off-peak eCooking by getting users to agree to cook with electricity only when the voltage is high enough (indicated by a volt- meter installed by the mini grid developer in every kitchen). This case study explores the potential role of battery storage in overcoming the supply constraints to enable eCooking at with high-performance battery storage and a suitably sized any time. solar panel to create a solar home system capable of deliv- ering cooking services. Until recently, such a device would The fourth case study looks at the prospects for cooking have been unrealistically expensive for most households in on a 20kW solar/biomass hybrid mini grid in Tanzania that developing countries. However, this is no longer the case, connects 58 households that currently cook with firewood or due to the falling prices of the two main cost components, charcoal, both of which are available at very low cost. There PV and batteries. This approach offers important co-benefits, is interest in eCooking among connected households and in the form of access to electricity for other applications. the grid infrastructure can support cooking loads. However, the tariff is very high. The case study explores two eCooking The fifth case study describes a Kenyan village, where cook- scenarios: (a) regular household cooking and (b) a concept ing was previously dominated by collected firewood, but tailored to maximize cooking efficiency for users operating dwindling forest resources and increasing livelihood oppor- microenterprises to precook beans for sale using an EPC. tunities have led many residents to start paying for firewood (or adopt charcoal, kerosene, or LPG). It explores whether The section on eCooking with stand-alone systems explores pairing a DC EPC with lithium-ion battery storage and a suit- the opportunity for eCooking powered by solar home ably sized solar panel may be able to offer a cost-effective systems. Energy-efficient eCooking appliances can be paired off-grid eCooking solution. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 27 by enabling on-grid eCooking. Case study 1 explores the 3.2.  eCooking on National opportunity for fuel stacking an EPC with LPG in Kenya (and neighboring East African countries), where 0 percent of the Grids population currently uses electricity as its primary cooking fuel (WHO 2017). Figure 3.2 shows that electricity is not a mainstream cooking solution in many Sub-Saharan African Grid electricity in much of Sub-Saharan Africa and South/ countries. Southern Africa is a notable exception to this Southeast Asia has capacity, reliability, access, and afford- trend, as electricity is already widely adopted there. In fact, ability challenges that generally preclude cooking with of the 58 million people in Sub-Saharan Africa who already electricity. Strong grids can enable a wide range of energy cook primarily with electricity, 41 million reside in South Africa services; however, on weak grids, load shedding, voltage (WHO 2017). However, in many countries, reliability is still fluctuations (brownouts), blackouts, and other inconsisten- a challenge, so case study 2 explores the opportunity for cies in the power supply limit the range of energy services mitigating load shedding in Zambia with energy-efficient and on offer and frequently disrupt the delivery of the services battery-supported appliances. that are available. Historically, cooking with electricity has not been encouraged in these contexts, because eCooking Table 3.1 compares statistics on the affordability, reliability, appliances, especially older and more inefficient models, access to, and renewable fraction of grid electricity in the draw much more energy and power than basic appliances countries studied in detail in this section of the report. The such as lights, radios, and TVs. Not all grids are strong following section explores some of the trends, with particular enough to support eCooking. However, emerging trends in reference to achievements in Southern Africa and emerging many places suggest that new opportunities are arising. opportunities in East Africa. The case studies in this section highlight two contexts in which there are clear opportunities to achieve social impact Percentage of households cooking primarily with electricity in Sub-Saharan African and FIGURE 3.2  South/Southeast Asian countries Cooking primarily with electricity >60 % 40 - 60 % 20 - 40 % 0 - 20 % No data 1 million people IBRD 45267 | AUGUST 2020 Note: Populations of more than 1 million people cooking primarily with electricity are indicated by black circles, the size of which is proportional to the number of people. See Figure 3.5 for a breakdown of the fuel mix in selected Sub-Saharan African countries. Source: Data from WHO (2017). 28 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 3.1 Electricity supply factors in Kenya, Tanzania, Uganda, and Zambia ANNUAL SHARE OF PERCENT OF NUMBER AND POPULATION POPULATION DURATION OF LIFELINE TARIFF GENERATION COOKING WITH ACCESS BLACKOUTS ELECTRICITY AND MONTHLY MIX (PERCENT PRIMARILY WITH CASE STUDY TO ELECTRICITY (SAIFI, SAIDI)a TARIFF ($/KWH) ALLOWANCE RENEWABLE) ELECTRICITY (%) Case study 1 Kenya Total: 64 Number: 13 0.23 Lifeline tariff: 87 0 Urban: 81 Average $0.17/kWh duration: 60 Allowance: hours 100kWh Tanzania Total: 33 Number: 47 0.15 Lifeline tariff: 34 1 Urban: 65 Average $0.04/kWh duration: 21 Allowance: hours 75kWh Uganda Total: 22 Number: 42 0.20 Lifeline tariff: 93 0 Urban: 57 Average $0.06/kWh duration: 59 Allowance: 15kWh hours Case study 2 Zambia Total: 40 Number: 5 0.09 Lifeline tariff: 97 12 Urban: 75 Average $0.02/kWh duration: 50 Allowance: hours 200kWh Note: a. SAIFI (the System Average Interruption Frequency Index) is the average number of service interruptions experienced by a customer in a year. SAIDI (the System Average Interruption Duration Index) is the average duration of outages experienced by a customer over the course of a year, measured in hours. Source: World Bank (2019a, 2019c); ZESCO (2019); KPLC (2019a); TANESCO (2019); Umeme (2019); WHO (2017). AFFORDABILITY Cowan (2008) studied the informal settlement of Imizamo Yethu on the outskirts of Cape Town. He showed that even South Africa is the only Sub-Saharan African country in with inefficient hot plates, cooking with electricity was by which the majority of the population (73 percent [WHO 2017]) far the cheapest option. However, the upfront cost of the already cook primarily using electricity. A household survey appliance was a barrier, as the most popular appliance (the conducted by the Republic of South Africa (2012) showed double hot plate) was considerably more expensive than a a steady increase in eCooking with income levels, with paraffin stove (although less expensive than an LPG stove). 55 percent of households in the poorest quintile using elec- He combined participatory field methods with laboratory tricity and up to 90 percent of households in the richest quin- testing to evaluate the energy requirements for the prepara- tile doing so. Even electrified low-income households were tion of local meals using the most common energy sources: more likely to rely on firewood than electricity, however, electricity, LPG, ethanol gel, and paraffin. The results which “suggests the existence of barriers or practices clearly showed that other than collected fuelwood (which amongst poorer households that inhibit a fuller transition” is assumed to be free), electricity was by far the cheapest (Republic of South Africa 2012, p. 25). Sebitosi and Pillay option for all types of dish.12 (2005), Bekker et al. (2008), and Cowan (2008) all suggest that cultural resistance to change plays a significant part, To counter the false perception that electricity is too expen- which is cemented by the perceived high cost of electricity. sive for cooking and incentivize the transition away from paraffin and firewood among the poor, the South African eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 29 government launched the Free Basic Electricity policy in prices in Zambia are still below $0.10/kWh, and although 2003. It entitled poorer households to 50kWh/month free standard residential tariffs in Kenya, Tanzania, and Uganda of charge (Bekker et al. 2008; Lemaire 2011). As a result, the are higher, they are all still below $0.25/kWh (see Table 3.2). proportion of households cooking with electricity grew from 59 percent in 2003 to 73 percent in 2011 (WHO 2017). Figure 3.3 compares the cost of cooking with charcoal and electricity across a range of typical charcoal prices and Figure 3.3 shows that South Africa has one of the lowest electricity tariffs in Sub-Saharan African cities. In cities with tariffs on the continent ($0.07/kWh). However, ener- high charcoal prices, such as Nairobi ($0.49/kg), cooking gy-efficient appliances and rising charcoal prices have now diary demand data suggest that consumers are likely to opened up opportunities for affordable eCooking in other be paying around $27/month to cook with charcoal. At this African countries as well. Electricity tariffs have remained price, it would be more cost-effective for consumers who relatively affordable, while charcoal prices have risen are connected to 38 of the 39 Sub-Saharan African utilities significantly in many East and Southern African countries studied by AFREA and ESMAP (2016) to cook all their food (Batchelor 2015), meaning that it is now cost-effective for with electricity. Even in cities with lower charcoal prices, such many urban charcoal users to switch to eCooking. Electricity as Lusaka ($0.21/kg), eCooking would still be cost-effective Sensitivity analysis comparing the cost of eCooking with the cost of cooking with charcoal across FIGURE 3.3  Sub-Saharan Africa 35 30 $0.5/kg Cost of Cooking ($/month) 25 Typical Charcoal 20 price range in Sub-Saharan 15 African Cities $0.2/kg 10 5 0 0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Ethiopia, Sudan Sierra Leone Seychelles Liberia Botswana, Burundi, Guinea, Benin, Burkina Faso, Gabon, Lesotho, Malawi, Mali, Rwanda, Senegal, Togo Cape Verde Mozambique, Nigeria, Rep. of Congo, South Africa, Gambia, Kenya, Mauritania, Electricity tari ($/kWh) Zambia, Zimbabwe Mauritius, Uganda Cameroon, Central African Republic, Comoros, Côte d’Ivoire, Ghana, Madagascar, Niger, São Tomé & Principe, Swaziland, Tanzania At 0.5$/kg: eCooking cheaper At 0.2$/kg: eCooking cheaper than Charcoal than Charcoal for 38/39 Sub-Saharan for 23/39 Sub-Saharan African Utilities African Utilities 100% electricity 100% Charcoal Note: Grid tariffs are the average cash collected per kWh reported by AFREA and ESMAP (2016). They therefore differ from the tariffs used as modelling inputs in the case studies, which used the latest available retail tariff for the first 100kWh/month. Cooking demand values are based on Tanzania cooking diaries data. All electric solutions are AC, modelled with a five-year financing horizon. Modelled charcoal prices represent the range of prices recorded by the study team in the Sub-Saharan African cities studied in this section: Lusaka ($0.21/kg), Dar es Salaam ($0.32/kg), Kampala ($0.28/kg), and Nairobi ($0.49/kg). 30 ESMAP  |  Cooking with Electricity: A Cost Perspective for grid-connected charcoal users in over half (23 out of 39) neighborhoods, vastly reducing the lifeline allowance per of the surveyed countries. user. Although South Africa’s Free Basic Electricity program is AFREA and ESMAP (2016) note that lifeline tariffs are usually exceptionally generous, a number of utilities offer discounted cross-subsidized by revenue from regular retail tariffs and social or lifeline tariffs, designed to enable affordable access that utilities in only two Sub-Saharan African countries to basic energy services for lower-income households (Seychelles and Uganda) currently have cost-reflective tariffs. (AFREA and ESMAP 2016). These tariffs offer a fixed amount Utilities may be reluctant to encourage greater utilization of of electricity each month at a reduced price, ranging from lifeline tariff allowances without further subsidization (from 15kWh to 500kWh, with discounts of 10–100 percent off of national budgets, official development assistance, or further regular retail tariffs. Pricing structures for electricity tariffs can cross-subsidization by increasing the regular household, be complex, making direct commercial, or industrial comparisons between them tariffs). As a result, although difficult. For example, many the modelling in Case Studies utilities use a rising block tariff 1 and 2 is carried out with In some countries, lifeline tariff (19 out of 39 surveyed), where the lifeline tariff, a sensitivity the lower consumption blocks allowances are already enough to enable analysis is presented to show are discounted at several poorer households to access affordable the effect of using the regular different rates, gradually rising retail tariff or future potential to the regular tariff. However, modern energy cooking services. tariff increases. 10 of the 39 utilities surveyed by AFREA and ESMAP had a tariff in which the first block (which is the only block if only a lifeline and regular retail tariff are offered) had an allowance of 100kWh/month or GENERATING CAPACITY AND RENEWABLE ENERGY more; 24 offered 50kWh/month or more. Howells et al. (2006) criticized South Africa’s focus on In some countries, lifeline tariff allowances are already electricity as a replacement for basic cooking fuels on the enough to enable poorer households to access afford- grounds that demand for electricity for cooking was likely able modern energy cooking services. Since AFREA and to be highest in the evening, when electricity demand for ESMAP’s (2016) study, Kenya Power increased its lifeline other applications was also highest, putting additional strain tariff allowance to 100kWh/month. It is unlikely that poorer on South Africa’s already overloaded coal-driven power households own a wide range of appliances, in particular grid. Instead, they advocated for exchangeable credits that energy-intensive appliances. Analysis of customer data could be used to purchase LPG for cooking, as both options from Kenya Power’s (2018) annual report shows that its depend on fossil fuels. domestic customers spent an average of just $6.22/month, which if divided by the lifeline tariff of $0.17/kWh suggests In contrast, Zambia’s grid is dominated by hydropower an average consumption of less than 37kWh/month.13 The (97 percent), offering a renewable alternative to unsustain- average household in the Kenya cooking diaries study able charcoal production. But supply is extremely vulnera- used 0.46kWh/capita/day, or 14kWh/capita/month, to cook ble to seasonal shortfalls in energy production as a result all its meals (see Table 2.5). These figures imply that the of limited rainfall. Case study 2 explores the opportunities average Kenyan household (modelled in this study as 4.2 for mitigating load shedding on Zambia’s overloaded people) could cook all its food with electricity and maintain grid. On grids that reach their limits at peak times but still its regular electricity consumption within the lifeline tariff have spare capacity, utilities could strategically incentivize allowance. households to prepare foods that are not time critical or trickle-charge a household battery by using price signaling In Zambia, the lifeline tariff allowance is 200kWh/month, (for example, off-peak tariffs). and the cooking demand figures in Table 2.5 are consider- ably lower, giving even more flexibility. However, the focus In contrast, East African grids, driven by long-term economic groups carried out in parallel with the cooking diaries studies growth ambitions, have been increasing their generation indicated that shared meters (one metered connection capacity and now have surplus generation, presenting an supporting multiple households) are prevalent in poorer opportunity to expand electrical demand into new sectors, eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 31 such as cooking. With the notable exception of Tanzania, the Fuel stacking is an alternative strategy to enable households majority of electricity in the region is generated from renew- with unreliable grid connections to cook some of their food able sources (see Table 3.1). Although utilities in the region with electricity. Case study 1 explores a clean fuel stack of have historically shied away from stimulating demand as a LPG and electricity. The survey conducted by the Republic of result of shortfalls in supply, this is now changing. Recent South Africa (2012) showed that 42 percent of South African installations in Uganda have increased generating capacity households relied entirely on electricity for cooking, indicat- to 950 MW (Akena and Wanless 2020), creating a surplus of ing that grid connections for those users must be sufficiently predominantly renewable generation (93 percent). What is reliable. However, where grid reliability is lower, fuel stacking more, Power Africa has identified another 1,900 MW of proj- can be a highly effective way of achieving a more resilient ects for completion by 2030 (Power Africa 2018). Generating household energy system. If the fuel stacked with electricity capacity in Tanzania was roughly 1,500 MW in 2017 (EWURA is a clean fuel, such as LPG, health outcomes can be compa- 2017), but with a further 1,600 MW planned, this capacity was rable to a fully electric solution. projected to double shortly (Eberhard, Gratwick, and Kariuki 2018). In addition, the Stiegler’s Gorge hydropower project will bring an additional 2,100 MW online (Eberhard, Gratwick, and Kariuki 2018), so the government’s targets of reaching 5,000 ENERGY ACCESS MW by 2020 (Export.gov 2019) and 10,000 MW by 2025 (EEG 2016) appear feasible. This increase in generating capacity The high rates of eCooking uptake in South Africa have been is encouraging, but many transmission and infrastructure enabled by energy policy that focused heavily on expanding challenges and management issues within the private and electricity access (Beute 2012). Just a third of the population public sectors remain. These problems notwithstanding, in had access to electricity before 1990 (Bekker et al. 2008); some countries eCooking could now be considered as a tool by 2012, the figure had grown to more than 80 percent to stimulate demand for excess electricity. (Republic of South Africa 2012). South Africa increased access by both extending transmission lines into rural areas formerly far away from the grid and by densifying connec- tions in urban communities lying “under the grid.” RELIABILITY In almost every country in Sub-Saharan African and South/ The reliability of grid electricity in urban East Africa is Southeast Asian, a sizable population now has access to now relatively high: the System Average Interruption electricity but still cooks primarily with other fuels, indicating Duration Index (SAIDI) and the System Average Interruption considerable untapped potential (Figure 3.4). Case study 1 Frequency Index (SAIFI) from each country’s economic explores how Kenya—which raised electricity access rates center indicates that in all four countries, power outages from 19 percent in 2010 to 75 percent in 2018 (ESMAP 2020) total less than five hours/month (see Table 3.1). Although user but still has 0 percent of the population using electricity as its experience varies significantly within any city (for example, primary cooking fuel (WHO 2017)—could increase the use of users with illegal connections will surely have less reliable electricity for cooking.14 access to electricity than the statistics suggest), the data suggest that the reliability of grid electricity in major cities is Of course, many of the people who do not use their electricity already sufficient for many people to consider eCooking. connection for cooking may well be cooking with other clean fuels, such as LPG, or fuels that can be collected for free, such Drawing power for cooking activities can have a detrimental as firewood. In India, for example, where the electrification impact on weak electrical systems where load shedding and rate is 79 percent, just 4 percent of the population cooks with load limitations are common. In systems with low voltage commercialized polluting fuels and technologies (charcoal, (for example, less than 150V on a grid designed for 220V), coal, and kerosene). In contrast, in many countries in East even the most efficient electrical cooking appliances would Africa, West Africa, and South East Asia, many households not be able to function. Voltage stabilization and/or stor- cook primarily with charcoal, coal, or kerosene (Figure 3.5)— age-based reinforcement could help overcome this problem. all fuels that the World Health Organization has declared as If used in a grid-connected context, such systems would be responsible for premature deaths from respiratory illnesses particularly effective if the built-in battery capacity can be (WHO 2016). In virtually all contexts, people are purchasing controlled by the utility. For instance, a smart-charge control- these fuels. This spending could be redirected into purchas- ler that allowed users to charge their batteries only at night ing electricity units for cooking. In Kenya, 29 percent of the could increase grid utilization ratios and avoid peak loading population cooks with commercialized polluting fuels and constraints. technologies; in Zambia, the figure is 37 percent. 32 ESMAP  |  Cooking with Electricity: A Cost Perspective Percentage of households with grid connections that still cook primarily with fuels other FIGURE 3.4  than electricity in Sub-Saharan Africa and South/Southeast Asia Grid-Connected, Cooking primarily with other fuels 80 - 100 % 60 - 80 % 40 - 60 % 20 - 40 % 0 - 20 % No data IBRD 45265 | AUGUST 2020 Source: Data from WHO (2017) and World Bank (2019c). Percentage of households cooking primarily with commercialized polluting fuels and technologies FIGURE 3.5  (charcoal, coal, or kerosene) in Sub-Saharan Africa and South/Southeast Asia Cooking primarily with commercialized polluting fuels and technologies (charcoal, coal, or karosene) 80 - 100 % 60 - 80 % 40 - 60 % 20 - 40 % 0 - 20 % No Data IBRD 4526 6 | AUGUST 2020 Source: Data from WHO (2017). Urbanization is spreading rapidly across the African are exhausted, charcoal has to be brought from farther continent; many areas that were previously rural are and farther away, pushing up the price in urban centers becoming peri-urban, meaning that many people who (Adam Smith International 2016). “Another Nigeria” will used to collect firewood are now forced to pay for it be added to the continent’s total urban population by (or to purchase charcoal instead). As nearby forests 2025, and urban centers are set to double in size over eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 33 FIGURE 3.6 Primary cooking fuel used in selected countries in East and Southern Africa Urban Total 100 90 Percentage of Population using as 80 70 Primary Cooking Fuel 60 50 40 30 20 10 0 Kenya Tanzania Uganda Zambia Ethiopia South Africa Kenya Tanzania Uganda Zambia Ethiopia South Africa CASE STUDY 1 CASE CASE STUDY 1 CASE STUDY 2 STUDY 2 Electricity Generator LPG Natural Gas Biogas Kerosene Alcohol Cool Charcoal Solar Wood Dung Crop Waste Source: Adapted from Leary et al. (2018), based on data from WHO (2017). the next 25 years, reaching 1 billion people by 2040 (Lall, Collectively, Ethiopia, Kenya, Tanzania, Uganda, and Zambia Henderson, and Venables 2017). are home to 38 million people who have a grid connection but choose to cook with commercialized polluting fuels and Charcoal still dominates cooking in urban areas of East and technologies, charcoal, or kerosene (Leary et al. 2018). There Southern Africa (Figure 3.6). Ironically, despite having surplus is thus considerable latent opportunity for expanding the use electricity and high access rates in urban areas (57 percent of electricity for clean cooking. in Uganda, 65 percent in Tanzania, and 81 percent in Kenya), less than 1 percent of the urban population in all three coun- LPG has become the fuel of choice for urban elites in Kenya tries uses electricity as a primary cooking fuel (WHO 2017). In and Tanzania, with 24 percent and 8 percent of the urban contrast, a substantial proportion of the urban population in population, respectively, using it as their primary cooking Zambia (27 percent) and Ethiopia (18 percent) already cooks fuel. However, many of these households still use charcoal with electricity, most likely facilitated by the low unit cost of for certain foods, creating an opportunity for a clean fuel electricity (under $0.02/kWh in both countries for the first stack that can enable households to move completely away 200kWh/month). from biomass (Case Study 1). 34 ESMAP  |  Cooking with Electricity: A Cost Perspective CASE STUDY 1 Building on the Success of LPG to Displace Charcoal in Urban East African Kitchens with a Clean Fuel Stack on the supply (reliability, access, poor-quality wiring, partic- SUMMARY ularly in informal connections) and demand (perception of Power generation source: National grid (Kenya Power cost, taste, behavioral change) sides. However supply-side and Lighting Company [KPLC]) (with references to barriers are decreasing and energy-efficient appliances are TANESCO and Umeme) offering a new opportunity to overcome many demand-side challenges. Tariff: $0.17/kWh Baseline fuels: Kenya is rapidly expanding and strengthening its national • charcoal ($0.49/kg) electricity grid. Like its neighbors Uganda and Tanzania, Kenya now generates substantially more electricity than it needs. In • LPG ($1.08/kg) 2018, generation capacity stood at 2,351MW, although peak • kerosene ($1.1/liter) demand was just 1,802MW, creating a power generation Future scenarios: surplus of 549MW. Kenya Power reported over 500,000 new customers in its 2018 annual report, bringing its customer • AC electricity and battery-supported DC electricity base to 6.7 million in 2018. The Last Mile Connectivity Project15 • most efficient appliances (rice cookers, EPCs) and has seen massive expansion of the grid into rural areas, hot plates raising the national electricity access rate from both grid and Location: Nairobi, Kenya (with references to Dar es off-grid solutions from 29 percent to 73 percent in just five Salaam, Tanzania and Kampala, Uganda) years (Kenya Power 2018). Many of these new customers have very low demand, however, bringing in limited extra revenue The results of this case study show that in urban for the utility, and the costs of connecting and maintaining the contexts with relatively high traditional fuel prices infrastructure in rural areas, where transmission and distribu- and average electricity prices, both AC and tion lines are substantially longer, has been a challenge. battery-supported DC eCooking options already offer considerable cost savings (charcoal: $23–$34/month; To increase demand, Kenya Power’s demand stimulation team AC: $12/month; AC/battery-supported DC hybrid: has been showcasing what households can do with electricity, $15–$17/month). These options will become more competitive if, as expected, traditional fuel prices continue to increase. LPG currently offers the lowest- Fuel stacking using LPG for manual FIGURE 3.7  cost cooking ($6–$11/month), but many users currently control and an electric pressure stack LPG with charcoal for cooking heavier foods. cooker for automatic control In urban areas with well-established LPG markets, stacking LPG with efficient eCooking appliances can offer an affordable and desirable pathway to moving completely away from biomass ($7–$10/month). Introduction East Africa presents a strategic opportunity, because it contains many of the world’s largest charcoal markets and electricity grids are becoming stronger and reaching more people than ever before. eCooking has seen limited uptake in East Africa, because of intertwined challenges eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 35 including demonstrating eCooking with its Pika na Power reduce the cost of cooking even further by fuel stacking (Cook with Electricity) program (KPLC 2019b). The program LPG with an EPC (see Figure 3.7). LPG offers quick lighting, has aired as a prime-time TV show and set up a demonstra- high maximum heat output, and manual heat control, and it tion kitchen open to the public with live cooking classes and can heat any shape utensil, making it ideal for dishes such the option to take home the appliances used on the show. as chapati. EPCs offer automatic control of an insulated and Any of Kenya Power’s 10,000 employees can pay for appli- pressurized cooking pot, making them ideal for heavy foods ances in installments deducted from their salary, and on-bill like beans. financing is under consideration as a way of extending this option to all 6.7 million customers. Pika na Power had focused After the cooking diaries studies in Nairobi and Dar es on induction and infra-red stoves, but it recently added a new Salaam were completed, participants were able to keep the appliance to the program, the EPC. appliances they tested. As many already had LPG stoves, this fuel-stacking configuration of LPG with an EPC is how LPG is the aspirational fuel across East Africa, with wide- the majority of these cooks choose to prepare their food, as spread uptake across Nairobi and Dar es Salaam and an both manual and automatic control are important for different emerging market in Kampala. LPG is distributed in pressurized cooking processes (Leary, Scott, Numi, et al. 2019; Leary et cylinders, connected by a regulator and hose to an LPG stove al. 2019). or, alternatively, with a cylinder-top burner. Prices have fallen considerably as markets have become more established and Nairobi, Dar es Salaam, and Kampala are all regional defor- supply chain efficiency improved. However, many house- estation hotspots. As a result, the price of charcoal has been holds with an LPG stove still purchase charcoal to cook heavy steadily increasing (World Bank 2015), as the forests around foods, such as beans, because they believe that it is cheaper. these urban centers are stripped bare and charcoal has to Although this may have been the case 10 years ago, today be brought from farther and farther away. Deforestation is a the relative price points of LPG and charcoal have reversed, major issue in Kenya, where an estimated 64 percent of the and the eCookBook and cooking diary studies show that LPG biomass harvested each year for household fuel is classi- is now competitive for all food types (Leary, Scott, Numi, et al. fied as nonrenewable (Drigo et al. 2014). In 2018, a logging 2019; Leary et al. 2019; Leary et al. 2019). ban was put in place to protect the nation’s dwindling forest reserves, causing the price of wood fuels to double over- This case study explores an opportunity for East Africans night. Although prices have now settled down, they are still who have already partially adopted modern energy in the the highest in the region, causing many people to consider form of LPG to transition completely away from biomass and other options. 36 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 3.2 Fuel prices in Nairobi, Kampala, and Dar es Salaam in 2020 and 2025 used in modelling 2020 2025 FUEL PRICE NAIROBI KAMPALA DAR ES SALAAM NAIROBI (MODELLED) Charcoal ($/kg) 0.49 0.28 0.32 0.62 LPG ($/kg) 1.08 1.41 0.77 1.42 Kerosene ($/liter) 1.1 n/a 0.82 1.55 Note: All financial values are in 2019 dollars. Fuel prices were obtained from household surveys and interviews with local practitioners. This case study focuses on Nairobi, extrapolating the results Results to compare opportunities in Kampala and Dar es Salaam, where electricity prices are similar but charcoal is much In Nairobi, charcoal is already by far the most expensive fuel. cheaper (Table 3.2). Charcoal forms a much larger part of Although there is a widespread perception that electricity the fuel mix in urban East Africa; in Uganda, two-thirds of the is too expensive for cooking, it is actually cheaper than urban population use it as their primary cooking fuel. The both charcoal and kerosene (Figure 3.8). Supporting part statistics are similar in Tanzania. In urban Kenya, kerosene of the daily cooking demand with a battery is already far and charcoal are equally popular (see Figure 3.6). more cost-effective than using charcoal, regardless of the eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 37 FIGURE 3.8 Monthly cost of cooking with main fuels in Nairobi, 2020 and 2025 Direct AC eCooking Fuel Stacking: AC eCooking / Battery-supported DC eCooking Battery-supported DC eCooking Fuel Stacking: Direct AC eCooking / Charcoal Charcoal Fuel Stacking: Battery-supported DC eCooking / Charcoal LPG Fuel Stacking: Battery-supported DC eCooking / LPG Kerosene Fuel Stacking: LPG / AC eCooking Fuel Stacking: Kerosene / Battery-supported DC eCooking Fuel Stacking: Kerosene / AC eCooking a. 2020 40 35 30 Cost of Cooking ($/month) 25 20 15 AC and Charcoal and battery DC Electricity 10 Kerosene and 5 LPG and Electricity Electricity 0 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 50% Bat DC 50% Bat DC Bat DC Bat DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG Kerosene 50% Kero 50% Kero 50% Kero 5-yr 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Electricity-National Grid b. 2025 40 35 30 Cost of Cooking ($/month) 25 20 15 Charcoal and Electricity AC and 10 battery DC Kerosene and 5 LPG and Electricity Electricity 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 50% Bat DC 50% Bat DC Bat DC Bat DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG Kerosene 50% Kero 50% Kero 50% Kero 5-yr 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Electricity-National Grid Note: Grid tariff of $0.17/kWh corresponds to price of the first 100kWh/month from KPLC in 2019. Where applicable, batteries are LiFePO4 sized for 50 percent (0.93kWh) or 100 percent (2.78kWh) of daily cooking load. 38 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 3.3 Electricity tariffs in Kenya, Tanzania, and Uganda ELECTRICITY TARIFF LIFELINE TARIFF ($/ LIFELINE ALLOWANCE COUNTRY UTILITY ($/KWH) KWH) (KWH/MONTH) Kenya Kenya Power 0.23 0.17 100 Tanzania TANESCO 0.15 0.04 75 Uganda Umeme 0.20 0.06 15 Source: KPLC (2019a); TANESCO (2019); Umeme (2019). repayment horizon. In Nairobi, charcoal is so expensive that discount is just 15kWh/month). The assumptions about even sizing a battery bank large enough to meet 100 percent eCooking within a lifeline tariff band inevitably involve wider of daily cooking demand is already cost-effective under a complexities about the economics and politics of cross-sub- utility model (20-year horizon); by 2025, it is projected to be sidies, as discussed above. cost-effective with lease-to-own business models (5-year horizon) as well. Fuel stacking with electricity decreases the cost of cooking for both charcoal and kerosene users. Fuel stacking with battery-supported devices is already a cost-effective option for charcoal users. When trees were more abundant and the LPG market was nascent, cooking with LPG was relatively expensive. The huge disparity in the cost of cooking with LPG and charcoal illustrates just how much the tables have turned. Many households that cook primarily with LPG still buy charcoal to cook heavy foods, however, because of the persistent but now false perception that it is the cheapest way of cooking them. In 2020, the cheapest way to cook in Nairobi is either stacking LPG with efficient eCooking appliances or using LPG alone.16 By 2025, the projected rise in LPG prices will mean that stacking LPG with electricity is likely to become the cheapest option. Although the LPG bar dips lower than the LPG/electricity bar in Figure 3.8, it does so only because of the broad range of uncertainty over LPG prices (+/–33 percent). At $0.23/kWh, Kenya’s regular electricity tariff is the highest in the region (Table 3.3). However, at 100kWh/month, the lifeline tariff offers the most generous allowance and is likely to be sufficient for cooking on top of other applications. However, Kenya Power’s lifeline tariff does not reflect a very deep discount; in Tanzania and Uganda, standard tariffs bracket this rate, at $0.15 and $0.20/kWh, respectively. Both countries’ utilities offer more generous discounts, but the monthly allowance is more limited (in Uganda, Umeme’s eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 39 FIGURE 3.9 Sensitivity of modelling results to charcoal price in Nairobi, Dar es Salaam, and Kampala, 2020 45 Kampala: Dar es Salaam: Nairobi: $0.28/kg $0.32/kg $0.49/kg 40 Charcoal 35 50% Battery DC 50% Charcoal 30 Cost of Cooking ($/month) 100% Battery DC 25 50% Direct AC 50% Charcoal 20 50% AC 50% DC 15 Direct AC 10 5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Price of Charcoal ($/kg) Note: Lines follow the midpoints of the range of costs for each scenario (the middle of the bars in Figure 3.8). All scenarios are modelled with a 20-year financing horizon. Where applicable, batteries are LiFePO4 sized for 50 percent (0.93kWh) or (2.78kWh) of daily cooking load. Electricity tariffs and popular foods are similar across the the black charcoal line offers lower cooking costs. For three countries. In contrast, the cost of charcoal varies Nairobi, where charcoal prices are highest, even sizing a widely. Figure 3.9 shows the sensitivity of cooking costs battery to support 100 percent of cooking demand is cost to the price of charcoal, for both AC and battery-sup- comparable with charcoal, although all other options are ported DC eCooking in 2020, using the demand values significantly cheaper. With lower charcoal prices in Dar es and electricity price for the Nairobi case modelled above. Salaam, stacking battery-supported DC eCooking devices The vertical blue lines indicate current charcoal prices in with charcoal and 100 percent eCooking with a battery the largest city of each country (Nairobi, Dar es Salaam, sized to support 50 percent of daily demand are on a par and Kampala). The other lines show the relationship with charcoal; all AC options are still cheaper. With even between cooking cost and the price of charcoal for a lower charcoal prices in Kampala, 100 percent AC cook- variety of cooking approaches. At the charcoal price ing and stacking charcoal with AC electricity are the only relevant to a locality, any cooking approach that is below cost-effective options in 2020. 40 ESMAP  |  Cooking with Electricity: A Cost Perspective CASE STUDY 2 Tackling Load Shedding in Lusaka, Zambia, by Time Shifting and Reducing Electricity Demand for Cooking SUMMARY Introduction Power generation source: National grid (ZESCO) This case study illustrates an opportunity for countries Tariff: $0.01/kWh with significant populations already cooking on electricity Baseline fuels: to optimize the loading on their grid. Efficient eCooking appliances and practices can significantly reduce energy • AC electricity (hotplates and ovens) demand and peak loading for households currently cooking • charcoal ($0.21/kg) on inefficient appliances, such as hot plates and ovens. Future scenario: Supporting eCooking appliances with a battery can time- shift energy demand for cooking and reduce peak loading; • more efficient appliances, battery-supported DC it can also enable customers to cook during blackouts or electricity, LPG ($2.07/kg) load shedding. Location: Lusaka, Zambia The results of this case study show that in contexts with low electricity tariffs, both AC and battery-sup- Charcoal market in Lusaka, Zambia, FIGURE 3.10  ported eCooking can already offer considerable cost alongside electricity distribution savings for charcoal users even if charcoal is relatively infrastructure cheap (charcoal: $5–$10/month; AC: $2/month; AC/ battery-DC hybrid: $3–$5/month; battery-DC: $6–$9/ month). However, supporting inefficient eCooking appliances and practices with batteries is not cost-effective ($19–$28/month versus $6–$9/month), as the battery bank must be three times larger. In contexts with emerging LPG markets and low electric- ity tariffs, electricity is the cheapest option for moving away from biomass, even if load shedding is severe and the entire day’s cooking load has to be supported by a battery, as long as energy-efficient appliances and practices are employed (LPG: $8–$13/month; battery-DC: $6–$9/month). eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 41 Zambia’s national grid is 97 percent dependent on hydro- deforestation (Dlamini et al. 2016). Two-thirds of urban power. When the rains came late in 2015 and hydropower Zambians use charcoal as their primary fuel, but many fuel capacity was vastly reduced, the national utility (ZESCO) stack electricity and charcoal. Charcoal production increased was left with no choice but to implement load shedding. The dramatically to meet the growing demand during load shed- situation improved in 2016 and 2017, when load shedding ding, stepping up the pressure on Zambia’s already strained seemed like a thing of the past. However, water levels at natural resources (Dlamini et al. 2016). key hydropower installations were low in 2019, leading to 4 hours of scheduled blackouts per day initially, rising to The opportunity to dramatically reduce energy consumption 8 hours/day (14 hours/day in residential areas) in 2019. with energy-efficient cooking appliances is a highly attrac- tive proposition (Figure 3.11). eCooking is the aspirational Twenty-seven percent of urban Zambians already cook with solution for most people in Zambia, but the legacy of old and electricity as their primary fuel. Charcoal and electricity are inefficient equipment makes cooking with electricity unnec- the fuels of choice in urban Zambia; even when grid electric- essarily slow and expensive, despite extremely low tariffs. ity is available, many people still chose charcoal (Figure 3.10). As a result, ZESCO, is looking for ways to manage electricity ZESCO’s recent load shedding caused a significant number demand more sustainably. Finding a more efficient alterna- of users to revert back to charcoal, rapidly accelerating tive to hot plates is vitally important. FIGURE 3.11 Comparison of mbaula, hot plate, and electric pressure cooker Mbaula • Ubiquitous across urban Zambia • Inefficient, requires expensive fuel, unhealthy, environmentally destructive • Popular for “long boilers” Hotplate • Aspirational • Popular for quicker-cooking dishes • Efficient for quick dishes, healthier and less environmentally destructive than charcoal. • Still slow, expensive, and unpopular for “long boilers” Electric pressure cooker • Available but not yet popularized • Far more energy efficient, quicker, and easy for “long boilers” • Can also cook “medium boiler/friers” and “quick friers” 42 ESMAP  |  Cooking with Electricity: A Cost Perspective Household energy storage in the form of battery-sup- system is power limited, adding battery storage can help ported eCooking systems is also an attractive proposition reduce peak demand by time-shifting electricity demand for users, although it may exacerbate ZESCO’s problems. for cooking. Detailed power system modelling is needed to Battery-supported eCooking systems can enable cook- explore the effects of adding battery-supported eCooking ing during load shedding. Solar suppliers in Lusaka now onto ZESCO’s overloaded grid. report selling more battery/inverter systems for back-up power supplies in grid-connected households than off-grid Results solar systems. Battery-supported eCooking systems also enable access to other low-power energy services, such as Comparing figures 3.8 and 3.12 reveals that the outlook lighting, at no additional cost. However, depending on how in Lusaka is very different from the outlook in Nairobi. In much hydropower generation is run-of-the-river and how Lusaka, cooking with energy-efficient appliances is already much has reservoir storage, ZESCO may well already have by far the cheapest option ($2/month) and LPG is the most the ability to schedule generation at scale. If high levels of expensive ($8–$13/month). Cooking with charcoal is cheap: water storage are already built into the system, the power Evidence from cooking diaries shows that households supply is likely to be limited by energy, not peak power. typically pay just $5–$10/month (Leary, Scott, Serenje, et al. Therefore, introducing more loads, even battery-supported 2019a). At only $0.21/kg, the price of charcoal in Lusaka is loads, may be detrimental, as some energy is lost during less than half the price in Nairobi. However, at $0.014/kWh, the charge/discharge cycle, which may further reduce the ZESCO’s generous lifeline tariff is 1/10th that of KPLC. amount of energy available on the grid. In contrast, if the eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 43 FIGURE 3.12 Monthly cost of cooking using main fuels in Lusaka, Zambia, 2020 and 2025 Direct AC eCooking Fuel Stacking: AC eCooking / Battery-supported DC eCooking Battery-supported DC eCooking Fuel Stacking: Direct AC eCooking / Charcoal Charcoal Fuel Stacking: Battery-supported DC eCooking / Charcoal LPG Fuel Stacking: Battery-supported DC eCooking / LPG Kerosene Fuel Stacking: LPG / AC eCooking Fuel Stacking: Kerosene / AC eCooking Fuel Stacking: Kerosene / Battery-supported DC eCooking a. 2020 Electricity- ational rid 40 35 30 Cost of Cooking ($/month) 25 20 15 Charcoal and LPG and Electricity Electricity AC and 10 AC and Battery DC Battery DC 5 0 100% AC 50% AC 50% AC 100% 100% 100% AC 50% AC 50% AC 100% Bat 100% Bat 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC 5-yr 50% Bat DC 50% Bat DC DC DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG 20-yr 5-yr 20-yr 5-yr 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Ine cient Appliances Energy-e cient Appliances Charcoal LPG and Practices and Practices b. 2025 Electricity- ational rid 40 35 30 Cost of Cooking ($/month) 25 20 LPG and Electricity 15 Charcoal and Electricity 10 AC and AC and Battery DC Battery DC 5 0 100% AC 50% AC 50% AC 100% 100% 100% AC 50% AC 50% AC 100% Bat 100% Bat 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC 5-yr 50% Bat DC 50% Bat DC DC DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG 20-yr 5-yr 20-yr 5-yr 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Ine cient Appliances Energy-e cient Appliances Charcoal LPG and Practices and Practices Note: Figure uses energy demand data from neighboring Tanzania, where cooking practices are similar, because insufficient cooking diary data were available for LPG in Zambia. Where applicable, batteries are LiFePO4 sized for 50 percent (0.42kWh) or 100 percent (1.46kWh) of daily cooking load. Grid tariff used is $0.014/ kWh, which corresponds to the 2019 price of the first 200kWh per month from ZESCO. 44 ESMAP  |  Cooking with Electricity: A Cost Perspective The evidence from the cooking diaries shows that Zambian households with inefficient appliances and practices use three times as much energy as households with energy-efficient appliances and practices (95kWh/month versus 26kWh/month). However, even inefficient cooking still falls well within the 200kWh/month lifeline allowance. Zambia has historically had one of the lowest electricity tariffs in the world. As a result, many Zambians have adopted highly inefficient eCooking appliances (such as integrated four-plate hot plate and ovens) and have had little incentive to adopt energy-efficient practices. The modelling results show that cooking with inefficient appliances ($3/month) is more expensive than cooking with efficient appliances ($2/ month), although the majority of the cost in this scenario is in the appliance financing rather than the electricity. However, inefficient cooking puts a much bigger load on Zambia’s already overstretched national grid. Although the cost savings for switching to AC efficient appliances may be small, the difference is magnified many times over when the cooker must be supported by a battery. Supporting energy-efficient appliances and practices with a battery sized to support 100 percent of cooking on a five-year business model costs $10–$12/month. Supporting inefficient appliances and practices is estimated to cost $32–$40/month. ZESCO is actively encouraging users to switch to LPG , but LPG prices are currently high, as the market has yet to develop. LPG adoption could reduce the severity of load shedding by reducing demand for electricity and offer customers an alternative to charcoal when load shedding does happen. However, safety concerns among potential users and a long overland supply chain into this landlocked Southern African country present significant barriers (Leary et al. 2019). As a result, at $2.07/kg, LPG is twice as expen- sive in Lusaka as in Nairobi, putting the cost of cooking with gas at $8–$13/month. still cheaper than charcoal. If load shedding became more severe and the battery had to be sized for a full day’s cook- The modelling suggests that battery-supported cooking ing (1.26kWh), the cost of battery-supported cooking would with energy-efficient appliances and practices is already the rise to $6–$9/month for the utility model (20-year), which cheapest way for users to mitigate load shedding in Lusaka. is already cost comparable with charcoal in 2020. For the With four-hour blackouts, a battery sized to meet half the private sector (five-year) model, the costs in 2020 would rise daily demand (0.42kWh) could allow ZESCO’s customers slightly above charcoal, to $10–$12/month, but still sit far to cook whenever they wanted to. If ZESCO were able to below LPG. develop an on-bill financing mechanism to break down the high upfront cost, their customers could cook for just $3–$5/ By 2025, the opportunities open up even farther, with even month in 2020 (assuming a 20-year repayment horizon, batteries sized to meet a full day’s cooking demand from including battery and other equipment replacement). Even efficient appliances and practices paid back over five years under a private sector initiative that required a five-year resulting in cooking costs on a par with charcoal. Unless horizon, the price would be just $5–$6/month, which is the LPG market develops and prices fall significantly, LPG eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 45 FIGURE 3.13 Sensitivity of modelling results to potential tariff increases by ZESCO, 2020 and 2025 a. 2020 14 100% Battery DC 12 Cost of Cooking ($/month) 10 50% Direct AC 50% Battery DC 50% Battery DC 50% Charcoal 8 100% Charcoal 100% Direct AC 6 50% Direct AC 50% Charcoal 4 2 0 0 0.04 0.08 0.12 0.16 0.20 Lifeline >200 kWh/mth Grid electricity tari ($/kWh) Previous <200 kWh/mth ZESCO tari s Lifeline Current <200 kWh/mth >200 kWh/mth ZESCO tari s b. 2025 14 12 100% Battery DC Cost of Cooking ($/month) 10 50% Direct AC 50% Battery DC 50% Battery DC 50% Charcoal 8 100% Charcoal 50% Direct AC 50% Charcoal 6 100% Direct AC 4 2 0 0 0.04 0.08 0.12 0.16 0.20 Grid electricity tari ($/kWh) Lifeline Current <200 kWh/mth) >200 kWh/mth ZESCO tari s Lifeline Proposed <100 kWh/mth) >100, <300 kWh/mth >300 kWh/mth ZESCO tari s Note: Where applicable, batteries are LiFePO4 sized for 50 percent (0.42kWh) or 100 percent (1.3kWh) of daily cooking load. All scenarios are modelled with a 20-year financing horizon. 46 ESMAP  |  Cooking with Electricity: A Cost Perspective is likely to remain a luxury fuel for the elite, as costs are are cost-effective in 2020. For the upper band (more than projected to rise to $9–$16 a month. 300kWh/month), a battery sized for 50 percent of daily cook- ing would have a monthly cost similar to charcoal ($8–$9/ Electricity prices in Zambia have historically been very low, month), regardless of whether AC electricity or charcoal as ZESCO has been heavily subsidized by the Zambian were used for the other 50 percent. Even at this highest government. However, tariffs have been gradually increasing proposed tariff, AC eCooking is still the cheapest option, at toward cost-reflective levels. In 2017, tariffs were raised from just $6/month. K 0.61 ($0.05)/kWh to K 1.06 ($0.08)/kWh, but a lifeline tariff of K 0.18 ($0.014)/kWh still applied for consumption below It may seem unlikely that poorer households would ever 200kWh/month (19 percent tax included for all). ZESCO reach consumption levels of 300kWh/month. But focus recently applied to the Energy Regulatory Board to raise and groups revealed that it is possible, because of shared meter- restructure tariffs.17 Public outcry led the president to step ing, which may pose a significant barrier for eCooking (Leary in and veto any further tariff increases until after the next et al. 2019). Poorer households are more likely to share a election, in 2021. connection with their landlord/lady, because connection fees and monthly standing charges are high. Two tenants The results of the sensitivity analysis show that even if tariffs sharing a meter with their landlord/lady could use only are raised after the next election, eCooking will still be the 100kWh/month before they slipped into the highest bracket; cheapest option. Figure 3.13 explores the sensitivity of the five tenants would have just 50kWh/month. What is more, cooking cost to the grid tariff. Any cooking approach that many tenants with shared meters pay a fixed monthly fee is below the charcoal cost line at the relevant tariff level for utilities to their landlord/lady. Although electricity is very offers cost savings compared with cooking with charcoal cheap in Zambia, cooking heavy foods on a hot plate uses alone. The proposed tariffs had two monthly consumption many units. For this reason, many landlords/ladies ban their thresholds, 100kWh and 300kWh. For a household whose tenants from cooking heavy foods with electricity. Targeting consumption fits into the lowest tariff bracket, even a battery social marketing campaigns at landlords/ladies to make them sized for a full day’s cooking is cheaper than charcoal in aware of just how little electricity EPCs consume (and there- 2020. For the middle band (100–300kWh/month), this fore why they should allow their tenants to use them) could option becomes cost-effective by 2025; all other options be a key enabler for eCooking. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 47 demand, and many developing regions are rapidly urbaniz- 3.3.  eCooking on Mini ing, creating an opportunity to stimulate demand by substi- tuting expenditures on biomass fuels with expenditures on Grids electricity. Without significant subsidy or other financial assistance, mini Electricity is the most versatile energy vector, and trans- grid operators often struggle to bring in enough revenue forming this “high-grade” energy into “lower- grade heat is to cover the cost of establishing the infrastructure. Tariffs often seen as wasteful. This perception is accentuated in are often very high, because demand is typically low, as many mini grid contexts, where the unit cost of electricity is newly connected rural households often use only low-power relatively high compared with national grid tariffs and cook- energy services, such as lighting. A study by the Institute ing fuel expenditure relatively low, because it can often be for Transformative Technology (ITT 2016) notes the “chicken collected for free or purchased at a lower cost than in urban and egg problem”: Until most customers adopt higher-power centers. As a result, many mini grids prohibit the use of high- appliances and increase total demand on the grid, the tariff power electrical appliances (for example, appliances that must be prohibitively high to be break even, but customers produce heat) in order to avoid overloading the system and will not want to adopt higher-power appliances as long as conserve limited power and energy for other applications. the tariff is high. Cabling is a major cost for mini grid developers. To reduce Cooking offers a valuable tool for stimulating demand and costs, many developers install cables and distribution boards bringing mini grid consumption above the break-even point. that are rated for lower power usage. However, the unit Figure 3.14 illustrates the break-even tariff for a typical 30kW cost of electricity on mini grids is strongly influenced by solar hybrid mini grid in India supplying a village of 1,000 Break-even tariffs for typical solar hybrid mini grid in India at different levels of energy FIGURE 3.14  consumption 3.50 LED plus phone 3.00 charge, 2–5 kWh 2.50 Break-even Tari (kWh) Break-even consumption at current cost 2.00 Plus Fan 3–5 kWh 1.50 Plus Small TV 5 kWh 1.00 Plus Cooking Plus Irrigation pump, 30–60 kWh Plus Refrigerator, 40–50 kWh 0.50 20–30 kWh Current cost ($0.40/kWh) 1 10 20 30 40 50 60 70 80 90 100 Paid consumption (KWh/month/household) Current consumption Source: Adapted from ITT (2016). 48 ESMAP  |  Cooking with Electricity: A Cost Perspective people. ITT (2016) notes that to break even at current levels the capabilities of existing mini grids, cooking loads can be of consumption (LED lights and phone-charging only), the considered at the design stage. Doing so can enable the mini grid operator would have to charge significantly more cost of building higher-capacity centralized infrastructure than current typical tariffs of $0.40/kWh. However, if house- to support AC cooking to be objectively compared with holds consumed 30kWh each month, the mini grid could the costs of enabling DC eCooking with battery-supported break even. The cooking diary studies show that house- devices, adding load management devices, or any of the holds in Kenya, Myanmar, Tanzania, and Zambia consumed other options discussed here. 30–60kWh/month to cook all their food (see Table 2.5). Cooking offers a rare opportunity to capture an existing expenditure and divert it into the revenue of mini grid devel- ECOOKING ON POWER-LIMITED MINI GRIDS opers. Lighting (by kerosene, candles, or dry-cell batteries) is usually an existing household expenditure; other appli- Cooking on mini grids is not a new idea: Many consumers cations (TV, radio, refrigeration) usually are not. In order in South and Southeast Asia are already using power-lim- to create more demand and use the infrastructure more ited mini grids for cooking. The abundance of hydropower effectively (and therefore bring down the unit cost), there is resources has enabled the establishment of mini grids with often a strong drive to add productive applications (such as very low unit costs. Rice is the major staple across much irrigation), which can both increase demand and simultane- of Asia. It presents a particularly attractive opportunity, as ously create/enhance the ability to pay. electric rice cookers are very easy to use and relatively low powered, and energy-efficient insulated appliances are Cooking is also a productive application (for restaurants, already available. However, further development of eCook- street food vendors, and so forth), and in many contexts, it ing on power-limited micro-hydro mini grids is often held already requires expenditure. In many rural areas, cooking back by peak loading constraints, as generating capacity is fuel is often collected; however in urban, peri-urban, and limited by the size of the turbine installed and/or the head/ an increasing number of rural areas, people have to pay flow characteristics of the watercourse. for it. The potential for time saving for fuel collectors and/ or expenditure substitution for fuel purchasers presents a Peak loading is a major issue for eCooking on power-­ valuable opportunity for mini grid developers to boost their limited mini grids, but a variety of time-shifting techniques revenue and lower the unit cost for all consumers by increas- can decouple electricity demand from supply, smoothing ing demand on the grid. In addition, as communities urban- out the load profile and bringing down the levelized cost ize, the demand for quick, easy, and clean cooking tends to electricity (LCoE) (Figure 3.15).18 In systems with low load of ­ increase, as people want more time for their paid employ- factors, the additional demand generated from cooking ment and households aspire to modernize. Urbanization also could increase this ratio and generate more income for reduces access to forest resources, driving people who used the mini grid developer/operator, by using energy that to collect biomass fuel for cooking to start paying for it and may otherwise be wasted. For example, run-of-the-river people who already pay to pay more. hydropower systems usually do not have any storage; they simply divert the flow around the turbine when demand is Mini grids can be broadly categorized into those that are low. This energy could be captured and stored with battery power limited (for example, run-of-the-river micro-hydro) storage, either centrally or at the household level as part of a and those that are energy limited (for example, PV power- battery-supported eCooking system. ing small DC loads with oversized cables). Most mini grids exhibit characteristics of both (for example, PV with AC loads Smart metering, distributed load control, and collabora- is power limited by the power rating of the inverter, although tive agreements can also mitigate peak loading issues. micro-hydro systems with reservoir storage are also energy Many mini grid developers are already implementing such limited), but it is useful to differentiate the two, as each pres- approaches, which they can simply apply to eCooking. Smart ents different opportunities and challenges for eCooking. metering can empower users to decide when they want to cook based on price and other signals sent by mini grid The case studies in this section explore the cost of adding operators. Distributed load controllers can allow mini grid household batteries to support eCooking on a power-limited developers to restrict users to activating cooking appliances micro-hydro mini grid to mitigate peak loading constraints only when excess power and energy are available (Gammon, (Case Study 3) and AC eCooking on an energy-limited solar Boait, and Advani 2016) and ensure that cooking appliances hybrid mini grid with spare energy and power available do not switch on simultaneously (instead cycling on and off (Case Study 4). Of course, rather than simply expanding alternately). Case Study 3, on Myanmar, illustrates another eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 49 FIGURE 3.15 Effect of increasing load factor on levelized cost of electricity of power-limited mini grids 100 LCOE = $0.35/kWh 80 Percentage of peak load LCOE = $0.42/kWh 60 40 20 LCOE = $0.55/kWh 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 22% load factor 40% load factor 80% load factor Source: ESMAP (2019a). option: developing agreements among customers regarding storage. Although the former requires the mini grid devel- the use of eCooking appliances, either by specifying variable oper to plan and implement the expansion, household times or getting them to agree to cook only when the volt- batteries can allow users to add storage without having to age is high enough, indicating spare capacity. wait for the mini grid developer to upgrade the system. In solar hybrid mini grids, there is also the option of adding additional dispatchable generation to increase capacity at peak times. ECOOKING ON ENERGY-LIMITED MINI GRIDS Lombardi et al. (2019) investigated the cost implications of New opportunities are opening up for the integration of upgrading the centralized infrastructure on an energy-limited energy-efficient eCooking in a broader range of systems, in solar mini grid to meet eCooking demand. They found mini- particular solar and solar/diesel hybrid mini grids. Costs in mal change in the LCoE. They modelled two representative the mini grid sector have dropped significantly as a result of cases from Tanzania, focusing on induction cookers. They both dramatic declines in the prices of key components (PV found that the levelized cost of cooking a meal with induc- panels and battery storage) and a number of nontechno- tion cookers ($0.17–$0.38) was similar to that of cooking with logical drivers, such as bundling, standardization, reduced charcoal and less than the cost of cooking with kerosene or regulatory uncertainty, and reduced costs of capital (ESMAP LPG. Assuming an average of 2.5 meals per day, the average 2019). In addition, the availability of new energy-efficient monthly cost would be $13–$29, which is comparable to the eCooking appliances greatly reduce the amount of electricity values obtained for optimized solar hybrid mini grids in Case required to deliver eCooking services. As a result, eCooking Study 4 of $17–$25/ month. Although the total capital costs on mini grids should no longer be confined to micro-hydro of fully transitioning to eCooking in residential scenarios systems with very low tariffs. would be would almost three times the costs associated with the base load, the LCoE would increase only slightly, as a Battery storage and generation on solar mini grids can be result of a parallel increase in the load demand. In the case added in a centralized or decentralized manner—by, for of community centers with high-load appliances, the incre- example, adding additional storage to the main battery mental capital costs would be only 17 percent higher. bank in a solar mini grid or adding household battery 50 ESMAP  |  Cooking with Electricity: A Cost Perspective CASE STUDY 3 Enabling 24-Hour eCooking on Micro-Hydro Mini Grids in Myanmar able to cook when they want to rather than when the energy SUMMARY level in the mini grid allows them to. Power generation source: 80kW micro-hydro mini grid Myanmar’s Shan State has an estimated hydropower poten- tial of 100GW, yet most of the region is not connected to the Tariff: $0.16/kWh national grid. During the 60 years of military rule, the country Baseline fuels: was cut off from the rest of the world. As a result, local devel- • AC electricity (rice cookers, insulated electric frying opers and communities took matters into their own hands, plans, induction stoves) setting up over 5,000 micro-hydro-powered mini grids to • firewood ($0.12/kg) satisfy the region’s growing demand for electricity. Future scenarios: These systems can be classified into three categories: • battery-supported DC electricity, LPG ($1.08/kg) • rice cookers, insulated electric frying plans, ● Household pico-hydro systems are used largely for induction stoves lighting, because they generate only up to 2kW. They could be used for eCooking, however. Location: Shan State, Myanmar ● Community-owned micro-hydro systems are systems in which the peak load demand often surpasses the The results of this case study show that adding battery system output, forcing them to load-shed during the dry storage can enable customers of power-constrained season. The systems usually have a flat tariff structure. mini grids to do all of their cooking with electricity. eCooking presents a strong value proposition for these Cooking entirely with AC electricity is the cheapest systems, which could benefit by increasing their load option (AC: $7/month; firewood: $5–$10), but doing factors during off-peak hours. Many communities with so would overload the mini grid. Supporting the entire these system are now within reach of the national grid, cooking load with a battery would not be cost effec- but because the national grid is not reliable, many tive ($15–$18/month). A battery is required only when the grid is overloaded, however, so a much smaller (and cheaper) battery could enable 100 percent Powerhouse at one of the many FIGURE 3.16  eCooking (AC/battery-DC hybrid: $9–$10/month). The small community-owned micro-hydro results suggest that by 2025, such approaches will systems in Shan State, Myanmar be cost competitive with firewood (firewood: $6–$12/ month; AC/battery-DC hybrid: $8–$9/month). Introduction This case study highlights the opportunity for mini grid developers who have already enabled eCooking, to allow their customers to carry out all of their cooking with electric- ity. Peak loading is often the major constraint on mini grids; time-shifting electricity demand for cooking into off-peak periods can increase the load factor of the mini grid. It also increases customer satisfaction, by enhancing the quality of one of the most important energy services, as customers are eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 51 communities seek to upgrade their micro-hydro systems Voltmeter installed in kitchen in FIGURE 3.18  instead of using the national grid. Myanmar ● Cooperative-owned mini-hydro systems have been built by local decentralized renewable energy developers since the early 1990s (Figure 3.16). They often feature multiple productive end-uses. One such system is the 80kW mini grid at Naung Pain Lay, Pyn Oo Lwin. It extends to 11 villages, reaching over 600 households with more than 65 kilometers of distribution and transmission lines, supplying 36,000kWh a month and enabling a range of household and enterprise energy services, including cooking. The system is developed and owned by a local cooperative that includes consumers from each village. The motivation for the cooperative to develop the mini grid has been to provide energy access and to establish a locally owned Many mini grids do not allow users to plug in eCooking appli- electricity-based enterprises that can uplift the local ances, out of fear of overloading the system. Several coop- economy and meet some of its development needs. erative-owned mini-hydro systems have empowered their customers to partially enable eCooking without overloading A typical tariff for a cooperative-owned system in Shan State the grid. Lighting for households is not profitable enough to is K 250 ($0.16)/kWh. This tariff is used to model the compar- pay back the cost of infrastructure on its own, so developers ative costs of eCooking in this case study. From an interna- had to look for other energy services to increase revenue. tional perspective, this tariff seems affordable. However, the Traditionally, almost all households in Myanmar cooked cooperative is competing with a highly subsidized national on firewood, with many households paying other people grid that sells electricity at just K 30 ($0.02)/kWh). However, to collect it for them. In rural areas, firewood is usually the government grid sacrifices quality for cost (Figure 3.17). purchased by the bullock cart load or in bundles weighed by During the rainy season, blackouts often last several days. viss (1.63 kgs). Household surveys carried out in parallel with Many cooperatives have chosen to develop medium-quality the cooking diary studies revealed an average price of K 300 infrastructure that balances affordability with reliability. As ($0.12)/viss. Many cooperative-owned mini-hydro system a result, many customers choose to stay connected to mini customers have now switched to electricity for cooking. Most grids even when the government grid arrives, installing volt- of them already use energy-efficient appliances, such as rice age stabilizers if they want to connect delicate loads such as cookers or electric frying pans. However, at peak times, grids TVs or refrigerators (Figure 3.17). often reach capacity, causing the voltage to dip. Volt meters have been installed in kitchens (Figure 3.18) and collabora- tive agreements negotiated with users to allow eCooking FIGURE 3.17 Voltage stabiliser in Myanmar appliances to be plugged in only when users can see that the grid is not overloaded. Above 180V, eCooking appliances can be plugged in. This case study models two future scenarios— battery-supported cooking and LPG—that could enable a transition to 100 percent clean cooking. Supporting the efficient eCooking appliances with a battery would enable users to cook whenever they wanted, potentially enabling them to cook all their food using electricity, smoothing out the load profile, and freeing up capacity at peak time. Although Myanmar is a gas-producing country, the domes- tic LPG market is only just emerging, as during the military government, only government officials were allowed to buy LPG. According to household surveys, where it is available, LPG in rural Myanmar is already affordable, at an average price of $1.08/kg. 52 ESMAP  |  Cooking with Electricity: A Cost Perspective Rice is the major staple in Myanmar, and rice cookers are cooking load with a battery would not be cost-effective, already widely adopted. The insulated rice cooker and but it is also not necessary, as a battery is required only electric frying pan are two of the most popular eCooking when the grid is overloaded. Therefore, a much smaller appliances in Myanmar. They are inexpensive (less than $20) (and cheaper) battery can enable 100 percent eCooking. and widely available, and their insulation and low power This option is already cost comparable with firewood in draw make them highly compatible with the power and 2020 with a 20-year financing horizon and by 2025 with a energy constraints inherent with cooking on mini grids. Rice 5-year horizon. From a technical point of view, this option cookers are also commonly used to prepare soup, one of could be performed manually, with users switching to DC the main components of a typical Myanmar meal. Insulated battery-supported appliances when the voltage drops electric frying pans are also widely used for cooking the third below 180V. A simple safeguard device (already widely component in Myanmar cuisine, curries, for which induction used in Myanmar) could also be reconfigured to switch from and infra-red stoves are also gaining popularity. Kettles and the AC supply to the DC battery–supported supply at the thermo-pots (insulated kettles) are often used to boil water. same 180V threshold. Hybrid AC/DC appliances are already EPCs are starting to enter the market, but they have not yet on the market, meaning that the user would barely notice become standard issue. Cooking diary participants used the transition from one to the other. a blend of all of these appliances, so the modelling below represents cooking with a range of energy-efficient appli- Although LPG is not yet available and is more expensive ances (rather than hot plates and efficient appliances as in than firewood and electricity, the overlapping bars in the other case studies). Figure 3.19 show that it would be cost competitive in 2020. Fuel stacking between battery-supported electricity and Results LPG or firewood pushes up the price in 2020, but by 2025, increasing fuel costs mean the reverse is true for LPG. In In 2020, AC eCooking was cheaper than cooking with fact, by 2025, even 100 percent battery-supported cooking firewood in Myanmar (Figure 3.19). As direct cooking on would be cost competitive with LPG on a 20-year financing the mini grid involves no use of traditional fuels, there is plan, should LPG price trends be toward the upper end of no cost range (the bar is just a line) Supporting the entire the modelled range. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 53 FIGURE 3.19 Monthly cost of cooking using main fuels in Shan State, Myanmar, 2020 and 2025 Direct AC eCooking Fuel Stacking: AC eCooking / Battery-supported DC eCooking Battery-supported DC eCooking Fuel Stacking: Battery-supported DC eCooking / LPG LPG Fuel Stacking: LPG / AC eCooking Firewood Fuel Stacking: Firewood / AC eCooking Fuel Stacking: Firewood / Battery-supported DC eCooking a. 2020 25 20 Cost of Cooking ($/month) 15 10 AC and battery DC 5 LPG and Electricity Firewood and Electricity 0 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC LPG 50% LPG 50% LPG 50% LPG Wood 50% Wood 50% Wood 50% Wood 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Electricity: Micro-Hydro Mini Grid b. 2025 25 20 Cost of Cooking ($/month) 15 10 AC and Battery DC 5 LPG and Electricity Firewood and Electricity 0 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC LPG 50% LPG 50% LPG 50% LPG Wood 50% Wood 50% Wood 50% Wood 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Electricity: Micro-Hydro Mini Grid Note: Mini grid tariff is $0.16/kWh. Where applicable, batteries are LiFePO4 sized for 50 percent (0.52kWh) or 100 percent (1.56kWh) of daily cooking load. 54 ESMAP  |  Cooking with Electricity: A Cost Perspective At just $7 a month, cooking with micro-hydro-generated electricity is already one of the cheapest options in 2020, but many mini grids are not currently able to support cooking at peak times. Figure 3.19 shows that at the current mini grid tariff of $0.16/kWh, the most popular current cooking of stacking AC electricity with firewood is the cheapest viable option, at $6–$8/month. However, supporting the cooker with a battery during peak hours (50 percent battery DC, 50 percent AC) is only marginally more expensive, at $9–$11/ month, and is comparable to using firewood alone for house- holds that are paying for it. Sizing the battery to cover 100 percent of the daily cooking load would be significantly more expensive, at $14–$22/month. By 2025, battery-supported cooking is projected to become one of the cheapest viable options, at $8/month. Falling battery storage costs and rising fuel prices mean that all of the fuel-stacking options for battery-supported eCooking are cost comparable with the use of that fuel alone. battery becomes cost-effective with tariffs below $0.13/kWh. For mini grids with spare capacity at peak times, AC eCook- Figure 3.20 explores the sensitivity of cooking costs to the ing is cheaper than stacking firewood/electricity (50 percent mini grid tariff. Supporting the eCooker with a battery at peak firewood 50 percent AC) with tariffs below $0.21/kWh in times becomes cheaper than stacking firewood/electricity 2020 and $0.20/kWh in 2025. AC eCooking is cheaper than below $0.11/kWh in 2020 and $0.13/kWh in 2025. In fact, firewood below tariffs of $0.27/kWh in 2020 and $0.32/kWh by 2025, even supporting the whole day’s cooking with a in 2025. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 55 FIGURE 3.20 Sensitivity of modelling results to mini grid tariffs in Myanmar, 2020 and 2025 a. 2020 25 100% Battery DC Current tari 0.16 $/kWh 20 50% Direct AC 50% Battery DC 100% Direct AC Cost of Cooking ($/month) 15 50% Direct AC 50% Firewood 10 Firewood 5 0 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Minigrid Tari ($/kWh) b. 2025 25 100% Battery DC Current tari 0.16 $/kWh 20 50% Direct AC 50% Battery DC 100% Direct AC Cost of Cooking ($/month) 15 50% Direct AC 50% Firewood 10 Firewood 5 0 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Minigrid Tari ($/kWh) Note: Figure shows when each eCooking scenario could become cost-effective. It therefore uses the upper bound of the traditional fuel price ranges and the lower- bound assumptions for eCooking costs, with the 20-year financing model. Where applicable, batteries are LiFePO4 sized for 50 percent (0.52kWh) or 100 percent (1.56kWh) of daily cooking load. 56 ESMAP  |  Cooking with Electricity: A Cost Perspective CASE STUDY 4 Exploring the Range of Opportunities for eCooking on Solar Hybrid Mini Grids Introduction: Household cooking SUMMARY Power generation source: 20kW solar/biomass hybrid Odarno et al. (2017) describe how the emerging solar mini mini grid grid industry in Tanzania owes its success to a number of factors, including progressive, light-handed regula- Tariff: $1.35/kWh (with comparisons to typical tariffs tion and the falling costs of solar PV and battery storage. from the mini grid sector in 2018 and 2025) The Tanzania Electric Supply Company (TANESCO) operates Baseline fuels: a number of large fossil fuel–based mini grids, which sell • charcoal ($0.13/kg) power at the same tariff as the national grid ($0.15/kWh, with • firewood $(0.04/kg) a lifeline tariff of $0.04/kWh for the first 75kWh/month). It is highly subsidized, as the cost of diesel-generation is much Future scenarios: higher. Many hydro-powered mini grids offer similar tariffs • AC electricity, most efficient appliances (EPCs) ($0.10–$0.20/kWh) without subsidization. However, much of • LPG ($1.16/kg) the population lives in more arid areas of the country without access to a suitable watercourse for hydropower generation. Location: Kibindu District, Tanzania For these people, solar and solar hybrid mini grids represent the most readily deployable technology available today. This two-part case study considers household cook- ing on a solar mini grid and a microenterprise that The solar hybrid mini grid sector is developing rapidly, precooks beans. driving down costs and opening up further opportunities for affordable eCooking. The case study presents a sensi- Part 1: Household cooking. In regions with low tivity analysis to reveal the forms of eCooking that become biomass prices and typical mini grid tariffs, cooking cost-effective at typical tariffs today and those that become with electricity would currently be expensive (char- so in the near future. ESMAP’s (2019a) comprehensive anal- coal: $6–$12/month; AC: $36+/month; AC/battery-DC ysis of the mini grid sector reveals that in 2018, solar hybrid hybrid: $45+/month). However, mini grid tariffs are expected to fall. As a result, by 2025, in peri-urban regions, where biomass fuels are more expensive, Kibindu village residents experimenting FIGURE 3.21  fuel-stacking electricity with charcoal ($9–$15/month) with range of efficient eCooking appli- becomes cost-effective for some charcoal users ances during a focus group session ($6–$12/month). A clean fuel stack of electricity and LPG may also become an attractive option for some users ($13–$21/month). Part 2: Microenterprises. Precooking beans in an EPC with typical mini grid tariffs would already be much cheaper ($2–$4/month) than using charcoal ($6–$13/ month). Even in rural areas with low charcoal prices, such as Kibindu, falling mini grid tariffs would make eCooking for these heavy foods cost competitive by 2025. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 57 mini grid tariffs typically ranged from $0.55–$0.85/kWh. With Kibindu village ($0.13/kg). The price of charcoal is so low that a combination of increased load factor, streamlined planning, even when modelling with the considerably lower typical further declines in component costs, and other measures, tariffs from ESMAP’s (2019a) study ($0.55–$0.83/kWh, as tariffs are projected to fall to $0.25–$0.38/kWh by 2025. opposed to $1.35/kWh), the cost–viability gap in Kibindu is As tariffs decline, cooking with electricity is becoming an still considerable in 2020. increasingly affordable option for households connected to solar hybrid mini grids in Tanzania. However, by 2025, charcoal prices are projected to have risen an average of 3 percent per year, as a result of the The Tanzania Traditional Energy Development increasing scarcity of forest resources, and ESMAP (2019a) Organisation (TaTEDO) has been championing the use of projects tariffs to have fallen by 55 percent by optimizing efficient eCooking appliances in Dar es Salaam. It hopes solar hybrid mini grid design and deployment. Figure 3.22 to be able to enable its mini grid customers to cook with shows that at $9–$15/month, fuel stacking electricity with electricity shortly. The 20kW solar-biomass (maize cob) charcoal becomes cost-effective for some charcoal users, hybrid mini grid in Kibindu is a partnership between who will be paying $6–$12/month. What is more, at $13–$21/ TaTEDO’s social enterprise, Sustainable Energy Services month, a clean fuel stack of electricity and LPG may become Company (SESCOM), and Husk Power, financed by Power an attractive option for some. Africa. As of 2020, 58 households were connected, with plans to increase to 100 in the next phase. The mini grid Although it is unlikely that mini grid users will be paying already has centralized battery storage and distributes the high charcoal prices typical of urban areas (more than in 230V AC. The connections are not load limited, so $0.4/kg), the household surveys conducted by the study efficient eCooking appliances can be plugged in directly. team suggest that users of mini grids installed in peri-urban The tariff is currently very high (T Sh 3,100 ($1.35)/kWh), areas of East Africa are likely to be paying $0.2–$-0.4/kg. as it is a small-scale pilot project with innovative gener- Figure 3.23 shows that with ESMAP’s (2019a) typical tariffs ation technologies. Tariffs will be regularly reviewed as from 2018 ($0.55–$0.83/kWh), even cooking 50 percent the load factor increases as more customers connect and of the Tanzanian menu with electricity was unlikely to consumption per customer increases. be cost-effective for most solar hybrid mini grid custom- ers. For peri-urban mini grid customers paying tariffs at the An initial focus group in Kibindu village revealed significant bottom end of this range ($0.55/kWh) and charcoal prices interest in eCooking (Figure 3.21). Household surveys carried at the top end ($0.40/kg), fuel-stacking electric appliances out by the study team revealed the relative price points of with charcoal (at a cost of $22/month) is less expensive cooking fuels. No one in the village had ever cooked with than cooking solely with charcoal ($23/month). However, electricity, LPG is not available, and kerosene is not used for for all other customers, charcoal is more cost-effective. As cooking. Firewood is bought in bundles, with a small bundle a result, there is still a cost–viability gap of up to $11/month. (estimated at about 6 kg) for cooking one meal going for This gap represents the maximum difference between the T Sh 500 ($0.22) and a large one (estimated at 15 kg) for a cost of cooking with charcoal and the cost of fuel-stacking whole day’s cooking selling for T Sh 1,000 ($0.43). Charcoal charcoal and electricity (that is, the cost when the mini grid costs T Sh 1,500 ($0.65) for a 20-litre bucket (estimated to tariff is highest [$0.83/kWh] and the cost of charcoal lowest contain 5 kg of charcoal) during the dry season and TSh [$0.2/kg]). 2,000 ($0.87) during the rainy season. In a household survey undertaken for the project, average fuel prices were found to By 2025, tariffs for optimized solar hybrid mini grids are be T Sh 88 ($0.04)/kg) for firewood and T Sh 292 ($0.13)/kg projected to be 55 percent lower than in 2018 (ESMAP for charcoal. LPG in Dar es Salaam sells for about T Sh 1,779 2019a), opening up a broader range of opportunities for ($0.77)/kg). The cost is estimated to increase by 50 percent if cost-effective eCooking. By 2025, optimized mini grid charcoal is transported to Kibindu. tariffs are projected to fall to $0.25–$0.38/kWh (ESMAP 2019a), reducing the monthly cost of fuel stacking to $11–$19 Results: Household cooking (Figure 3.23). At these tariffs, it would be cost-effective for most peri-urban charcoal users to switch, as they would Figure 3.22 illustrates the cost–viability gaps for the 50 be paying $12–$23/month. The cost of cooking solely with percent and 100 percent household eCooking scenarios electricity would drop to $18–$28/month, making it cost- by comparing the monthly cost of cooking using ESMAP’s effective for consumers on mini grids in peri-urban areas (2019a) typical and projected mini grid tariffs (for 2020 with tariffs at the lower end of the range to switch to fully and 2025, respectively) and typical charcoal prices in rural eCooking solutions. 58 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE 3.22 Monthly cost of cooking using main fuels in Kibindu, Tanzania, 2020 and 2025 Direct AC eCooking Fuel Stacking: AC eCooking / Battery-supported DC eCooking Battery-supported DC eCooking Fuel Stacking: Direct AC eCooking / Charcoal Charcoal Fuel Stacking: Battery-supported DC eCooking / Charcoal LPG Fuel Stacking: Battery-supported DC eCooking / LPG Firewood Fuel Stacking: LPG / AC eCooking Fuel Stacking: Firewood / Battery-supported DC eCooking Fuel Stacking: Firewood / AC eCooking a. 2020 50 45 40 AC and Battery DC Cost of Cooking ($/month) 35 30 25 20 15 LPG and Electricity Charcoal and Electricity Firewood and 10 Electricity 5 0 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG Firewood 50% 50% 50% 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Firewood Firewood Firewood 5-yr 20-yr 5-yr Electricity—Solar Hybrid Mini Grid b. 2025 50 45 40 35 Cost of Cooking ($/month) 30 25 20 AC and 15 Battery DC 10 LPG and Electricity Charcoal and 5 Electricity Firewood and Electricity 0 100% AC 50% AC 50% AC 100% 100% 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 100% 50% AC 50% Bat DC 50% Bat DC 5-yr 50% Bat DC 50% Bat DC Bat DC Bat DC Charcoal 50% Char 50% Char 50% Char LPG 50% LPG 50% LPG 50% LPG Firewood 50% 50% 50% 20-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr 5-yr 20-yr 5-yr Firewood Firewood Firewood 5-yr 20-yr 5-yr Electricity—Solar Hybrid Mini Grid Note: ESMAP’s (2019a) typical solar hybrid mini grid tariffs of $0.55–$0.85/kWh from 2018 are assumed applicable in 2020; its optimized solar hybrid mini grid tariffs of $0.25–$0.38/kWh are applied in 2025. Where applicable, batteries are LiFePO4 sized for 50 percent (0.97kWh) or 100 percent (2.98kWh) of daily cooking load. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 59 FIGURE 3.23 Break-even analysis for mini grid tariffs for household cooking, 2020 and 2025 100% Electricity Fuel Stacking: Electricity and Charcoal 100% Charcoal A. 35 $ /kg 0.4 30 $/kg 0.2 25 0.4 $/kg Cost of Cooking ($/month) 20 Typical charcoal price range in East African peri-urban contexts: 15 0.2–0.4 $/kg 0.2 $/kg 10 5 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Electricity Tari ($/kWh) Typical Tari Range for Solar Hybrid Fuel stacking cheaper for some Mini-grids in 2018 peri-urban mini-grid customers B. 35 $ /kg 0.4 30 $/kg 0.2 25 0.4 $/kg Cost of Cooking ($/month) 20 Typical charcoal price range in East African peri-urban contexts: 15 0.2–0.4 $/kg 0.2 $/kg 10 5 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Electricity Tari ($/kWh) Predicted Tari Range for Optimized 100% eCooking cost-e ective for some and fuel stacking Solar Hybrid Mini-grids in 2025 cheaper for most peri-urban mini-grid customers Note: Cooking demand values are based on Tanzania cooking diary data. All electric solutions are AC modelled with a five-year financing horizon. Typical tariff for 2018 and 2025 optimized tariff ranges are from ESMAP (2019a). Modelled charcoal prices are $0.2–$0.4/kg. 60 ESMAP  |  Cooking with Electricity: A Cost Perspective Introduction: Cooking as a microenterprise close by also sell precooked foods, in particular foods that are time-consuming to prepare, such as beans. Customers The second part of this case study highlights what is can take the precooked beans home and quickly prepare a expected to be the first step for eCooking in the most tasty meal by frying the ingredients for the sauce and stirring economically challenging contexts. In a context such as in the softened cereals. Many households use charcoal in Kibindu, with very expensive electricity and very cheap the same way—boiling heavy foods in bulk and then frying biomass, only the most efficient eCooking solutions will be portions on a different fuel at a later date (Leary, Fodio Todd cost-effective. et al. 2019). The EPC is the most efficient cooking appliance, leveraging Results: Cooking as a microenterprise efficiency gains that are possible only with electricity (insu- lation and automation) and combining them with pressuriza- For typical solar hybrid mini grids, the most efficient tion. It is most efficient at boiling heavy foods. By combining forms of eCooking was already cheaper than charcoal in it with energy-efficient practices, the EPC offers the most peri-urban areas in 2018 (Figure 3.24). A microenterprise efficient eCooking solution that has the greatest impact precooking cereals once a day would spend $2–$4/month on the foods that require the most energy—namely, heavy with an EPC or $6–$13/month with charcoal. However, foods. Controlled cooking tests carried out for the eCook- charcoal prices in Kibindu are very low ($0.13/kg) and Book (Leary et al. 2019) showed that boiling half a kilogram the tariff very high ($1.35/kWh), so in 2020 there is still a of yellow beans requires almost 1 kg of charcoal or 0.3. kg of cost–viability gap of $1/month, even for this most efficient LPG but just 0.15kWh of power using an EPC.19 form of eCooking ($4/month for charcoal versus $5/month for an EPC). However, even if the tariff in Kibindu does not Precooking (parboiling) beans is a growing microenterprise fall any further, by 2025, the projected 3 percent annual activity in East Africa. It involves boiling cereals (or other increase in charcoal prices is likely to make the two options heavy foods) to the point at which they become soft, with the cost comparable. In contrast, with optimized solar hybrid expectation that frying will be carried out later to make the mini grids, the cost of precooking cereals once a day is final dish as tasty as possible. Many street vendors who sell projected to drop far below the cost of cooking with char- vegetables and charcoal in small quantities to people living coal, to just $1–$2/month. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 61 FIGURE 3.24 Break-even analysis for mini grids tariffs for microenterprise cooking, 2020 and 2025 100% Electricity 100% Charcoal A. 14 0.4 $/kg 12 10 Typical charcoal price range in East African Cost of Cooking ($/month) peri-urban contexts: 0.2–0.4 $/kg 8 0.2 $/kg 6 o Cost of eCooking at Kibindu in 2020 4 Cost of cooking with charcoal in Kibindu in 2020 2 0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Typical Solar Hybrid 100% eCooking cheaper Electricity tari ($/kWh) Mini-grids Tari s for all peri-urban mini-grid Kibindu experimental customers in 2018 Solar Hybrid Mini-grid Tari in 2020 B. 14 0.4 $/kg 12 Typical charcoal price 10 range in East African peri-urban contexts: Cost of Cooking ($/month) 0.2–0.4 $/kg 8 0.2 $/kg 6 4 2 0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Electricity Tari ($/kWh) Optimized Solar Hybrid 100% eCooking cheaper for all peri-urban mini-grid Mini-grid tari s in 2025 customers Note: Cooking demand values are based on controlled cooking tests for the eCookBook (Leary, Fodio Todd, et al. 2019). All electric solutions are AC modelled with a five-year financing horizon. Typical tariffs for 2018 and 2025 optimized tariff ranges are from ESMAP (2019a). Modelled charcoal prices are $0.2–$0.4/kg. 62 ESMAP  |  Cooking with Electricity: A Cost Perspective 3.4.  eCooking with Stand-alone Systems Stand-alone systems are the only way of supplying elec- tricity in remote off-grid locations. In the last decade, solar home systems have become the default solution for off-grid programs. Although most systems are designed to support low-power energy services, such as lighting and mobile phone-charging, they can also support energy-efficient eCooking appliances paired with high-performance battery storage and a suitably sized solar panel, to create a fully electric household cooking system that has the poten- tial to meet all of a household’s everyday cooking needs. Until recently, such a device would have been unrealisti- cally expensive for most families in developing countries. However, over the last decade, the price of the two main cost components, PV and batteries, has fallen considerably, and highly efficient eCooking appliances, such as the EPC, are now available on the mass market. Case Study 5 shows that in some markets, a solar home system sized for highly efficient eCooking can already be cost-effective. This finding is supported by a growing body of evidence from academics and practitioners (Leach and Oduro 2015; Jacobs et al. 2016; Couture and Jacobs 2019; Zubi et al. 2017). However, the size of the initial investment required for a PV-battery system for cooking (some hundreds people who currently have access to low- or no-cost wood of dollars) puts it outside the ability and willingness of most fuel. First, with increasing incomes, increasing population, customers to purchase directly. Appropriate consumer and decreasing resources, the cost of wood fuel is likely to financing models will therefore be essential, as discussed in continue to rise. Second, with continuing declines in input detail in Chapter 4. technology costs and further innovations in eCooking system efficiencies, eCooking holds the potential for substantial cost Solar eCooking solutions are likely to be most valuable in declines. Case Study 5 demonstrates that even before these rural areas. However, rural households may produce or two trends take effect, eCooking can already be the most gather their fuel with their own labor and no cash expendi- cost-effective cooking solution in some markets, which hold ture (Buskirk 2019). In contrast, access to wood fuel in urban the potential of becoming early adopters. centers is usually highly constrained, meaning that charcoal use is generally more prevalent than firewood and that the prices of biomass fuels are usually higher than in peri-ur- ban and rural areas. As a result, for hundreds of millions of COMPARING THE COSTS OF DIFFERENT MODELS households that have access to self-produced or low-cost wood fuel, it is much harder for eCooking to compete, Several studies have modelled the cost of cooking with because the increased efficiency of using wood directly for solar-powered battery-supported eCooking systems. Recent cooking fuel makes it about five times less expensive than studies highlight the potential of EPCs to dramatically charcoal on a per energy unit basis, as the efficiency of char- reduce costs (Leach and Oduro 2015; Jacobs et al. 2016; coal production in Sub-Saharan African contexts is typically Couture and Jacobs 2019; Zubi et al. 2017). Each model 10–25 percent (Falcão 2008). takes a slightly different approach and is built on its own set of assumptions and input parameters. Table 3.4 summa- However, two long-term trends will likely make eCooking rizes the key parameters from each study (for details, see competitive in the not too distant future even for many of the appendix G). eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 63 TABLE 3.4 Key parameters of selected studies modelling the costs of solar eCooking systems HOUSE- UPFRONT FINANCING AC HOLD SIZE COST ($) MODELED HORIZON OR (NUMBER BATTERY SOURCE YEAR (YEARS) DC APPLIANCE OF PEOPLE) STORAGE PV (W) LOW HIGH Case Study 5 2020 5 DC EPC and 4.2 2.2kWh 630 1,162 1,342 hot plate LiFePO4 2020 5 DC EPC and 4.2 0.74kWh 220 453 513 LPG LiFePO4 2025 5 DC EPC and 4.2 2.2kWh 630 869 976 hot plate LiFePO4 2025 5 DC EPC and 4.2 0.74kWh 220 351 387 LPG LiFePO4 Beyond 2019 3 DC Hotplate 5 1.5kWh 400 1,526 1,799 Fire Electric lithium-ion Cooking (Couture and 2019 3 DC Induction 5 1.2kWh 300 1,390 1,635 Jacobs 2019) lithium-ion 2019 3 DC Slow 5 0.45kWh 100 491 572 cooker lithium-ion 2019 3 DC EPC 5 0.36kWh 80 600 681 lithium-ion Zubi et al. 2020 10 DC EPC 6 2.1kWh 420 2,266 2,266 (2017) LiFePO4 2025 12 DC EPC 6 2.1kWh 420 1,926 1,926 LiFePO4 2030 14 DC EPC 6 2.1kWh 420 1,644 1,644 LiFePO4 2035 15 DC EPC 6 2.1kWh 420 1,426 1,426 LiFePO4 Beyond Fire 2016 20 AC Hotplate 5 Not stated Not 1,032 6,202 (Jacobs et al. stated 2016) 2016 20 AC Induction 5 Not stated Not 1,008 6,060 stated Leach and 2015 20 AC Hotplate 4 2.2–9.8kWh 367–1,331 1,032 6,202 Oduro (2015) LiFePO4 2025 20 AC Hotplate 4 2.2-8.7kWh 367–13,31 718 3,550 LiFePO4 In 2015, Leach and Oduro modelled the cost of cooking on that it offered the most viable pathway, as it would require a solar home system. They concluded that at the time it was minimal behavior change from charcoal. By 2020, they not cost-effective compared with charcoal or LPG but that by projected costs of $7–$70/household/month, depending on 2020 it could be. Leach and Oduro’s original model focused a wide variety of input parameters, including household size solely on a 500W hot plate as, at the time, it was believed 64 ESMAP  |  Cooking with Electricity: A Cost Perspective and uncertainty about the performance, cost, and lifetime of competitive with kerosene and LPG. They did not include key components. the cost of financing, but the net present cost and system lifetime can be used to estimate monthly costs of cooking, In 2016, the first Beyond Fire study (Jacobs et al. 2016) which by 2020 were $19/household/month, falling to $14 by projected that there was still a long way to go before cook- 2025 and $9 by 2035. ing on a solar home system could be cost competitive. The sequel, published just three years later (Couture and Jacobs Figure 3.25 compares the projected monthly cost of the 2019), found that it was already cost competitive. Couture cooking service from each model. It shows that cooking and Jacobs (2019) reported an 82 percent reduction in PV with an EPC powered by a solar battery/electric system costs and a 76 percent reduction in battery storage costs built between 2020 and 2035 is likely to cost the consumer since 2010. More importantly, they expanded their modelling $9–$30/month. However, it is unlikely that most households of demand to include the most efficient appliances, dramat- would be able to cook all their food on an EPC (or a slow ically reducing their estimates for the cost of cooking on cooker) without significantly changing their menu. As a solar eCooking systems. Their 2016 report estimated the result, this study builds on the work of Zubi et al. (2017) and cost of solar eCooking with a hot plate or induction stove Couture and Jacobs (2019) to incorporate the cost of using at $56–$162/ household/month financed on a three-year an additional appliance or fuel for foods that are incompat- PAYG contract. Their updated 2019 report estimated that this ible with EPCs. This study predicts that solar home systems cost had dropped to $44–$59 and found that replacing the designed for 100 percent eCooking with a hot plate and an induction stove or hot plate with an EPC reduced the cost to EPC would cost $20–$29/household/month in 2020. Smaller just $20–$23. solar home systems powering just the EPC and paired with LPG to offer a clean fuel stack are projected to be the Zubi et al. (2017) modelled a solar home system designed more cost-effective, at $13–$17/household/month, falling to to power a DC EPC. They concluded that it was already cost $11–$15 by 2025. FIGURE 3.25 Monthly cost of cooking with different fuel options projected by various models Fuel Stacking: Battery-supported DC eCooking / LPG Battery-supported DC eCooking Battery-supported AC eCooking 80 70 60 Cost of Cooking ($/month) 50 40 30 20 10 0 EPC and EPC and EPC and EPC and Hotplate Induction Slow EPC in EPC in EPC in EPC in EPC in Hotplate Induction Hotplate Hotplate hotplate LPG in hotplate LPG in in 2019 in 2019 cooker 2019 2020 2025 2030 2035 in in in in in 2020 2020 in 2025 2025 in 2019 2016 2016 2015 2020 Beyond Fire: Electric Cooking Beyond Fire Leach and Case Study 5 Zubi et al. (2017) (Couture and Jacobs 2019) (Jacobs et al. Oduro 2016) (2015) Note: Table 3.4 shows additional parameters. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 65 ALTERNATIVE SOLUTIONS spent on fuel collection and tending fires into income-gener- ating activities. This additional income can substitute for the The stand-alone systems section of this report focuses on lack of existing expenditure on cooking fuels in rural commu- substituting existing expenditures on biomass fuels with nities where firewood is collected for free. payments for a modern energy cooking service supplied by a solar-powered battery-supported system. Other solutions— SunCulture are using this model by experimenting with including directly driven DC appliances, alternative energy eCooking as an additional service for their RainMaker2 with storage technologies, alternative power generation sources, ClimateSmart Battery™ solar home system, which features and income generation with productive appliances—can also a 310W PV panel and 444Wh lithium-ion battery origi- enable eCooking in off-grid regions. nally designed to power the RainMaker2 irrigation pump. Low-power appliances such as LED light bulbs and TVs are Income generation with productive appliances already packaged with the system, offering significant extra value to users. Initial pilots are underway to explore the As an alternative to designing customized systems to viability of extending the system’s range of energy services power an eCooking appliance, it is possible to simply plug to include eCooking by offering a DC EPC that plugs into an eCooking appliance into a larger solar home system the same port as the pump. Both appliances have similar designed for other purposes. Although smaller systems peak power ratings (of about 300W), so they can be used designed for lighting and other low-power applications interchangeably. Demand for irrigation is lower in the rainy would be overloaded by eCooking appliances, many larger season, when wood fuel becomes harder to access. Usage systems designed for productive applications could support of the DC EPC is therefore likely to peak when usage of the them. Such systems also open up new markets, as the pump is likely to dip. productive applications enable a direct repurposing of time 66 ESMAP  |  Cooking with Electricity: A Cost Perspective Direct-drive DC appliances Early prototypes of DC cooking appliances that can be connected directly to solar panels to enable cooking during sunny periods are under development (Batchelor et al. 2018; Gius et al. 2019; Watkins et al. 2017). Omitting the battery enables the development of very low-cost solutions, with capital costs below $100. To make maximum use of solar energy, these appliances are highly insulated, which also offers thermal energy storage. With a standard resistive heat- ing element, a load controller is required, however. Gius et al. (2019) show how a chain of diodes connected directly to a PV panel can offer an even cheaper solution by acting as both a heating element and a voltage controller. In sunny locations where a significant proportion of cook- ing involves boiling, which typically takes place during the daytime, this option can provide much more efficient use of solar energy, as energy storage inherently involves energy loss (Buskirk 2019). However, as a result of the low power input (typically 100–300W), frying is challenging, and cook- ing in the evening, after the sun has set, or in the morning, before it has risen, will require fuel stacking, additional small electrolyzers to produce hydrogen for cooking exist, generation sources, or energy storage. but proof of concept is still required for these technologies at household scale. Further conversion to other energy carriers Alternative energy storage technologies (which might be easier to handle than hydrogen) would be implausible at small scale. Despite its relatively high capital Energy storage options for solar electricity for the end-use of cost, the main advantages of electro-­ chemical storage is cooking include the following: that batteries are modular and can be deployed at any scale, maintenance requirements are low or zero, and the stored ● thermal (highly insulated appliances, a hot fluid, phase- energy can easily be used for other applications (as it can change materials) quickly and efficiently be converted back into electricity). ● mechanical (using a micro-flywheel) ● chemical (power to gas, storing as hydrogen or with Alternative power generation sources further conversion to some gaseous or liquid fuel) ● electro-chemical (in the form of batteries). Other renewable generation sources, such as small-scale wind or pico-hydro, could also be employed in place of or as Thermal energy storage could be cheapest, but losses part of a hybrid system alongside solar. With the exception can be considerable without very careful use by the cook. of diesel generators, additional generation sources are often Automatic control of the cooking process is much easier to very site specific. This report focuses on solar eCooking achieve with an electric appliance, but conversion of stored as the most universally deployable stand-alone renewable thermal energy back to electricity is unrealistic at small scale. energy solution currently available. Chemical energy storage is attractive, as the cooking experi­ ence would be similar to cooking with LPG. Propositions for eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 67 CASE STUDY 5 The Next Generation of Cooking-Enabled Solar Home Systems Introduction SUMMARY Power generation source: Solar PV Kenya is East Africa’s commercial hub. It has a strong track record for innovation in the energy for development space. Solar resource: 3.85kWh/day/kWp in lowest insolation The M-Pesa mobile money system has reached scale in month Kenya, enabling innovative energy service companies to Baseline fuels: roll out PAYG solar solutions in the mass market. In the first • charcoal ($0.30/kg) half of 2019, 974,000 pico-solar products and solar home • LPG ($1.33/kg) systems were sold in Kenya, making it the biggest solar • kerosene ($1.18/liter) home system market in the world, overtaking India, which • firewood ($0.13/kg) has 27 times more people, for the first time (GOGLA 2019). Most viable future scenarios: This case study shows that eCooking is now a possibility • Clean fuel stack: LPG and PV-powered battery- for people who live beyond the reach of electricity grids. supported DC EPCs The solar revolution that has enabled access to low-power energy services for millions of people provides an ideal Location: Echariria, Nakuru County, Kenya platform to build a solar eCooking industry to cater to Kenya’s vast off-grid population and pave the way for a The case study results show that with suitable busi- similar transformation across Africa. LED made solar lighting ness models, high biomass fuel  prices in some heav- systems affordable by reducing energy demand by an order ily deforested contexts can already make fuel stacking of magnitude; the EPC may well hold an equally transforma- solar eCooking cost-effective for some biomass tive potential for solar eCooking. users (charcoal: $12–$22/month; charcoal/solar electric fuel stack: $12-18/month; firewood: $10–$19/ Although the village center in Echariria, located in the month; firewood/solar electric fuel stack: $12–$17/ Kenyan highlands, has been grid connected for several month). The cheapest option is currently LPG ($6–$11/ years, the connection fee is too high for many people living month). However, although fuel stacking LPG with a on the periphery, who remain off-grid. A community solar battery-supported solar-powered DC EPC increases the cost ($8–$13/month), it yields important co-ben- efits by enabling electricity access for other uses. Early prototype of a battery- FIGURE 3.26  By 2025, fuel stacking LPG with a DC EPC ($8–$13/ supported DC electric pressure month) starts to become cost comparable with cook- cooker designed by SCODE ing all food on LPG ($7–$13/month), meaning that the low-power energy services typically enabled by solar home systems (lighting and so forth) will be available at marginal extra cost. This benefit is likely to be a key purchasing trigger, as it offers value to everyone in the household, not just the cook. Hundreds of thou- sands of Kenyan households are already paying $10/ month or more to PAYG solar providers for solar home systems. 68 ESMAP  |  Cooking with Electricity: A Cost Perspective Participatory cooking session with FIGURE 3.27  Residents of Echariria, Kenya at FIGURE 3.28  prototype of DC electric pressure a community meeting with a DC cooker in Echariria, Kenya electric pressure cooker Charcoal stove and battery that FIGURE 3.29  is regularly charged at Echariria’s solar hub Note: Two previously unrelated energy services—electricity for entertainment and for cooking—could be united into a single product that can make cooking cleaner, faster, and easier. The battery could be charged directly from each household’s rooftop PV to power SCODE’s DC EPC. project begun in 2016 enabled approximately 40 households Although biomass fuel prices have stabilized, many people to charge a small (480Wh) lead acid battery at a central hub have now experienced modern cooking with LPG and are equipped with a 3kW PV array. Productive energy services, reluctant to go back to biomass, at least not for all their such as egg incubation, are also available at the solar hub. cooking needs. Interviews with local residents revealed that However, the household systems can support only basic a tin of charcoal currently sells for K Sh 50 (K Sh 30 [$0.30]/ energy services, such as lighting and TV. kg); a sack typically costs K Sh 1,000–K Sh 1,300 and weighs approximately 40–50 kg (K Sh 20–K Sh 32/kg). Firewood Just a few years ago, firewood dominated cooking in is typically bought in bundles for K Sh 50, K Sh 100, or K Sh Echariria. As the pace of life has slowly increased, charcoal, 250. The largest is a 20-kg bundle carried overhead that kerosene, and LPG have crept into kitchens throughout the costs an estimated K Sh 13/kg. Most people use kerosene village. In fact, most households now fuel stack several of for lighting, but some also use it for cooking. To prevent these options. Access to firewood has become more and the adulteration of vehicle fuel with cheaper kerosene, the more difficult, as the village expands and people have to government recently increased prices; kerosene now sells walk farther and farther to collect firewood. Instead of doing for K Sh 115–K Sh 120/liter. LPG refills are available at K Sh so, many people pay others to collect it for them or buy char- 750–K Sh 800 for 6 kg and K Sh 1,750–K Sh 1,800 for 13 kg coal, which has a higher energy density and can therefore (K Sh 125–K Sh 138/kg). be transported from further away. Kenya’s 2018 logging ban caused the price of wood fuels to spike. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 69 Supported by the MECS program’s challenge funds, SCODE with much shorter repayment horizons (typically one to two (Sustainable COmmunity DEvelopment) is developing and years) are currently standard in the solar lighting industry. testing an innovative solar eCooking system consisting of Fuel stacking the battery-supported solar-powered DC EPC solar panels, battery storage, and a DC EPC (Figure 3.26). being developed by SCODE is already cost competitive for SCODE brings several decades of experience with a range charcoal users in Echariria in 2020 on a five-year repay- of clean cooking and off-grid solutions in Nakuru County, ment horizon. including biogas, improved charcoal stoves, and solar home systems. Until now, however, clean cooking and off-grid The least expensive option is currently LPG. However, electricity access were two very different activities. SCODE although stacking LPG with SCODE’s battery-supported has developed an early prototype, and participatory cooking solar-powered DC EPC increases costs, it yields important sessions have enabled community members to try out this co-benefits. Diverting existing expenditure on charcoal or new technology (figures 3.27, 3.28, and 3.29). LPG into a solar eCooking system also embeds electric power generation into households, enabling them to charge Results directly from their own rooftop PV rather than having to carry a heavy battery to the solar hub and wait for it to charge. The modelling results in Figure 3.30 show that high The energy left over in the battery after cooking is likely to biomass fuel prices already make fuel stacking solar be sufficient to run the low-power appliances that house- eCooking cost-effective for charcoal and firewood users in holds currently use (see Figure 3.29). Further analysis of Echariria. In fact, if SCODE were able to develop an energy time-of-day usage will be needed to determine which addi- service business model with a 20-year financing horizon, tional energy services can be supported. Of course, these a 100 percent solar eCooking solution would already be co-benefits would be even greater for villagers who do not cost-effective. However, lease-to-own or business models currently have electricity access. 70 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE 3.30 Monthly cost of cooking using main fuels in Echariria, Kenya, 2020 and 2025 Battery-supported DC eCooking Fuel Stacking: Battery-supported DC eCooking / Charcoal Charcoal Fuel Stacking: Battery-supported DC eCooking / LPG LPG Fuel Stacking: LPG / AC eCooking Firewood Fuel Stacking: Firewood / Battery-supported DC eCooking Kerosene Fuel Stacking: Kerosene / Battery-supported DC eCooking Fuel Stacking: Kerosene / AC eCooking a. 2020 30 25 Cost of Cooking ($/month) 20 15 10 Charcoal and PV Kerosene and PV Firewood and LPG and PV PV 5 0 100% PV Bat 100% PV Bat Charcoal 50% PV Bat 50% PV Bat LPG 50% PV Bat 50% PV Bat Kerosene 50% PV Bat 50% PV Bat Firewood 50% PV Bat 50% PV Bat DC DC DC DC DC DC DC DC DC DC 20-yr 5-yr 50% Charcoal 50% Charcoal 50% LPG 50% LPG 50% Kero 50% Kero 50% Firewood 50% Firewood 20-yr 5-yr 20-yr 25-yr 20-yr 5-yr 20-yr 5-yr Electricity— Solar Home System b. 2025 30 25 Cost of Cooking ($/month) 20 15 10 Charcoal and PV Kerosene and PV Firewood and PV LPG and PV 5 0 100% PV Bat 100% PV Bat Charcoal 50% PV Bat 50% PV Bat LPG 50% PV Bat 50% PV Bat Kerosene 50% PV Bat 50% PV Bat Firewood 50% PV Bat 50% PV Bat DC DC DC DC DC DC DC DC DC DC 20-yr 5-yr 50% Charcoal 50% Charcoal 50% LPG 50% LPG 50% Kero 50% Kero 50% Firewood 50% Firewood 20-yr 5-yr 20-yr 25-yr 20-yr 5-yr 20-yr 5-yr Electricity— Solar Home System Note: Where applicable, batteries are LiFePO4 sized for 50 percent of the menu on a DC PV-powered battery-supported EPC (0.74kWh capacity, 220Wpeak PV) and stacking with charcoal, LPG, or kerosene, or cooking 100 percent of the menu ( 2.2kWh battery and 630Wpeak PV.) eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 71 By 2025, traditional fuel costs are projected to have the co-benefits of access to electricity for other purposes increased by 15 percent and the price of solar eCooking may entice some users to fuel stack (at a cost of $13–$14/ system costs to have fallen by 23 percent. As a result, fuel month). In 2020, fuel stacking SCODE’s solar-powered stacking LPG with SCODE’s DC EPC using a five-year repay- DC EPC starts to become cost comparable with charcoal ment horizon starts to become cost comparable with cook- when prices hit $0.35/kg; for 100 percent solar eCooking, it ing all food with LPG, meaning that at marginal extra cost, becomes cost comparable at $0.45/kg. LPG users can gain access to the low-power energy services typically enabled by solar home systems. What is more, the The battery is the main cost component in battery-supported modelling results show that 100 percent solar eCooking solar eCooking systems, making the optimization of energy system with a five-year repayment horizon would be at cost demand with energy-efficient appliances and practices parity with charcoal and only marginally more costly than critical. Figure 3.32 shows the breakdown of the compo- purchased firewood. nent costs for each of the modelled scenarios. It reveals that solar panels make up a relatively small fraction of the Figure 3.31 explores the opportunities in 2020 for overall system cost. In the 50 percent fuel-stacking case, cost-effective eCooking on sites with similar levels of the system comprises a lithium iron phosphate (LiFePO4) solar irradiation (3.85kWh/day/kWp in lowest insolation battery of 0.74kWh rated capacity (with up to 80 percent month) but different charcoal prices. SCODE’s solar-pow- usable capacity assumed), charged by a 220Wpeak PV panel. ered DC EPC is not yet cost-effective for charcoal users in This battery meets half of the daily cooking requirements Echariria. But in Nairobi, where charcoal prices are consid- with an EPC; the other half is met by kerosene, charcoal, or erably higher, a solar home system sized for a full day’s LPG. In this scenario, cost recovery occurs over five years, so cooking (via a hot plate and EPC, at a cost of $26–$29/ the battery and balance of system components should not month) would already be cost comparable with charcoal require replacement (battery life is 3,000 cycles in this lower- ($27/month). In Kibindu, Tanzania (Case Study 4), char- bound scenario, equivalent to eight years of daily use).20 In coal prices are much lower, so 100 percent eCooking is the 100 percent eCooking case, a 2.2kWh LiFEPO4 battery not at all competitive with charcoal ($6/month). However, pack is charged by a 630Wpeak PV panel. Sensitivity of solar battery–eCooking and fuel-stacking scenarios to charcoal price with a FIGURE 3.31  five-year repayment horizon, 2020 35 PV (high cost) 30 100% Charcoal PV (low cost) 25 50% PV Ctso of Cooking ($/month) 50% Charcoal (upper bound) 20 50% PV 50% Charcoal (lower bound) 15 10 5 Kibindu, Tanzania Echariria Nairobi 0.11 $/kg 0.31 $/kg 0.49 $/kg 0 0.1 0.2 0.3 0.4 0.5 Charcoal Price ($/kg) 72 ESMAP  |  Cooking with Electricity: A Cost Perspective Breakdown of solar eCooking and fuel costs for systems sized to meet needs of average Kenyan FIGURE 3.32  household in 2025 20 18 Fuel tacking with 16 Solar Battery-Supported DC EPC Charcoal 14 Cost of Cooking ($/month) Kerosene 12 LPG 10 8 6 4 2 0 100% PV 50% PV 50% PV 50% PV 100% Charcoal 1% LPG 100% Kerosene 50% Charcoal 50% LPG 50% Kero Replacements Cooking appliances Balance of system Controls/charger Battery PV Note: Figure is based on lower-bound costs in 2025 with a five-year financing horizon. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 73 3.5.  Implications for eCooking in Off-Grid and Grid‑Connected Contexts CROSS-COMPARISON OF CASE STUDIES firewood/EPC: $8–$12/month; charcoal: $6–$12/month; charcoal/EPC: $10-16/month). The most cost-effective Figure 3.33 compares the results from the five case studies. clean cooking solution is a clean fuel stack of LPG and The main results include the following: an EPC ($12–$21/month). ● Case Study 1 illustrates an urban context with high ● Case Study 5 illustrates an off-grid rural area with charcoal prices ($0.49/kg), low LPG prices ($1.08/kg), moderate biomass fuel prices (charcoal: $0.30/kg; and average electricity prices (0.17/kWh). By 2025, both firewood: $0.13/kg) and moderate kerosene and LPG LPG ($7–$12/month) and a clean fuel stack of LPG and prices ($1.18/liter, $1.33/kg). By 2025, a solar home an AC EPC ($8–$10/month) are likely to the lowest-cost system designed to support both a hot plate and an options; both will be substantially cheaper than charcoal EPC to cook all foods ($19–$21/month) is expected to ($27–$39/month). The EPC offers a particularly desirable be cost comparable with charcoal ($14–$24/month) and solution for cooking heavy foods that can encourage marginally more expensive than firewood ($10–$19/ households to move completely away from biomass. month). The cheapest option is expected to be LPG ($8–$12/month). However, a clean fuel stack of LPG ● Case Study 2 illustrates an urban context with lower with a solar home system powering a DC EPC ($11–$14/ charcoal prices ($0.21/kg), high LPG prices ($2.07/kg), month) can offer valuable co-benefits by enabling access and low electricity prices ($0.01/kWh) but recurring load to electricity for other purposes at marginal extra cost. shedding that prevents households from cooking when they want to. The findings show that by 2025, a hybrid AC/DC eCooking system with a battery sized for half the day’s cooking using energy-efficient appliances and practices will be the cheapest option ($7–$8/month), substantially cheaper than charcoal ($6–$12/month) or fuel stacking charcoal or LPG with electricity ($4–$7/ month and $11–$17/month, respectively). Even if load shedding is more severe and the battery needs to be sized for an entire day’s cooking load ($10–$12/month), eCooking would still be cheaper than charcoal. ● Case Study 3 shows a rural area, with moderate firewood prices ($0.12/kg) and electricity access from a micro- hydro mini grid with a low tariff ($0.16/kWh). By 2025, fuel stacking electricity with firewood is likely to remain the cheapest option ($6–$9/month), unless the generating capacity of the mini grid is upgraded to enable 24-hour AC cooking. However, for marginal additional cost, a battery sized to support half the day’s cooking load could enable 24-hour eCooking ($9–$10/month), the cost of which would be on a par with firewood ($6–$11/month). ● Case Study 4 depicts a rural area with low-cost biomass fuels available (firewood: $0.04/kg, charcoal: $0.13/kg) and access to electricity via a mini grid with a very high tariff ($1.35/kWh). By 2025, tariffs in the solar hybrid mini grid sector are expected to have fallen considerably (to $0.25–$0.38/kWh), enabling eCooking at marginal extra cost by fuel stacking an EPC (firewood: $4–$6/month; 74 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE 3.33 Emerging opportunities for cost-effective eCooking identified in each of the five case studies Current practice New opportunities for eCooking AC eCooking Fuel Stacking: AC eCooking / Charcoal Battery-supported Fuel Stacking: Battery-supported DC eCooking / LPG CASE STUDY 1 Kenya Grid DC eCooking 40 Charcoal Fuel Stacking: LPG / AC eCooking LPG Fuel Stacking: AC eCooking / Battery-supported 30 Firewood DC eCooking Kerosene Fuel Stacking: Firewood / AC eCooking 20 10 0 CASE STUDY 2 Zambia Grid 20 15 10 5 0 CASE STUDY 3 Myanmar Micro-Hydro Mini-Grid Cost of Cooking ($/month) 12 8 4 0 CASE STUDY 4 Tanzania Solar Hybrid Mini-Grid 25 20 15 10 5 0 CASE STUDY 5 Kenya O -Grid—Solar Home System 30 20 10 0 Note: Reference year is 2025; all solutions are modelled with a 5-year financing horizon, except grid-connected battery-supported systems, which are modelled with a 20-year horizon. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 75 BEYOND THE CASE STUDY RESULTS and electricity tariff. An array of results was created for each system architecture and fuel-stacking combination. Charcoal The following section interpolates and extrapolates the was used as the baseline fuel of reference. Prices from $0 results obtained from the five case studies to compare to $0.6/kg and electricity tariffs from $0 to $1.4/kWh were the costs of eCooking with grid-connected (both AC and modelled, reflecting the range of values seen in the case battery-supported) and stand-alone system architectures in a studies (only tariffs up to $0.6/kWh are shown in the figures broader range of contexts. below). Figure 3.34 models households cooking using the energy demand data from the cooking diary study carried out Figures 3.34 and 3.35 show which system architectures are in Tanzania (see Table 2.6). Figure 3.35 models the productive most cost-effective at each combination of charcoal price use case for Tanzania described in Case Study 4 part 2. Optimal-system diagrams for household cooking, based on electricity/charcoal price FIGURE 3.34  combination and quality of the grid, 2025 A. Reliable Grid B. Unreliable/Weak Grid Grid-eCooking 0.6 0.6 Solar-Battery-eCooking Charcoal 0.5 0.5 50 percent AC/ 50 percent DC- Charcoal Price ($/kg) Charcoal Price ($/kg) Grid-Battery- 0.4 0.4 eCooking Fuel Stacking: 0.3 0.3 Grid-eCooking and Charcoal 0.2 0.2 Fuel Stacking: Solar-Battery-eCooking and Charcoal 0.1 0.1 0 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 Electricity Tari ($/kWh) Electricity Tari ($/kWh) C. Very Unreliable/Weak Grid 0.6 0.5 Charcoal Price ($/kg) 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 Electricity Tari ($/kWh) Note: All scenarios are based on data from the Tanzania cooking diaries. Areas show which system architecture is most cost-effective at each electricity/charcoal price combination. For reliable grids (panel a), blackouts/voltage instability are assumed to have a negligible effect on grid-connected eCooking, so no battery is required for grid-connected architectures. For unreliable/weak grids (panel b), blackouts/voltage instability are assumed to affect 50 percent of grid-connected eCooking, during which a hybrid appliance switches from AC power to DC battery-supported mode. Batteries for grid/battery cooking are therefore sized at 1.0kWh to meet half the daily load. 76 ESMAP  |  Cooking with Electricity: A Cost Perspective The results of this analysis suggest that by 2025, eCook- For very unreliable/weak grids (panel c), blackouts/voltage ing is likely to be cheaper than charcoal in most contexts. instability are assumed to affect all grid-connected eCook- Charcoal is cheaper than eCooking only when charcoal ing. Batteries for grid/battery cooking are therefore sized at prices are low and grid tariffs are moderate or high (bottom 3.0kWh to meet all of the daily load. right of figures 3.34 and 3.35). For household cooking, Figure 3.34 shows that when charcoal prices exceed In all scenarios, batteries for solar battery–powered ­eCooking $0.3/kg, eCooking is always the cheapest solution, regard- are sized at 2.4kWh to allow the system to meet 100 percent less of the grid tariff. Above this threshold, solar battery– of daily cooking load, assuming that 20 percent of the load powered eCooking is cheaper than charcoal, however can be met directly by PV. Grid-powered ­ eCooking and grid-­ connected eCooking offers an even more cost-effective solar battery–powered eCooking system architectures were option when the grid tariff is low to medium. Below this modelled with a five-year financing horizon; grid/battery-­ threshold, the quality of grid electricity becomes the key powered eCooking was modelled as a utility (20-year horizon). factor. For reliable grids (panel a), only a small triangle in the bottom right remains for charcoal, because as a battery is For the productive use case highlighted in part 2 of Case not required to support the cooking load, upfront cost are Study 4 (boiling heavy foods in an EPC), Figure 3.35 shows low. For less reliable or weaker grids (panels b and c), where that the threshold drops to just $0.1/kg, meaning that in a battery is required to support part or all of the cooking virtually all contexts, highly efficient eCooking outcompetes ­ load, the upfront costs increase considerably, leaving just charcoal. If reliable grid electricity is available, directly a small window for grid/battery-powered eCooking in the plugging in an AC EPC is the most cost-effective option for middle/top left. tariffs up to $0.55/kWh, above which solar battery–powered eCooking becomes most cost-effective. Optimal-system diagram for productive-use case (precooking beans/cereals with an FIGURE 3.35  electric pressure cooker) on reliable grid, 2025 0.6 0.5 0.4 Charcoal Price ($/kg) 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 Electricity Tari ($/kwh) Note: Areas show which system architecture is most cost-effective at each electricity/charcoal price combination. Blackouts/voltage instability are assumed to have a negligible effect on grid-connected eCooking, so no battery is required for grid-connected architectures. Batteries for solar battery–powered eCooking are sized at 0.17kWh, to allow the system to meet 100 percent of daily cooking load, assuming that 20 percent of the load can be met directly by PV. Grid-eCooking and solar battery–powered eCooking system architectures were modelled with a five-year financing horizon. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 77 Exploring opportunities for cost-effective eCooking in grid/battery-powered eCooking (where the entire day’s diverse economic contexts cooking demand has to be supported by the battery) is no longer cost-effective. In rural charcoal-­ producing In contexts where charcoal is cheap and electricity tariffs are areas, charcoal prices are typically low ($0.10–0.20/kg), high, only the most efficient eCooking solutions are cost-­ so 100 percent grid-powered eCooking and 50 percent effective. As charcoal prices increase and tariffs decrease, grid/battery-powered eCooking (EPC only) become more opportunities for cost-effective eCooking open up. more expensive than charcoal. With medium tariffs ($0.25–0.55/kWh), typical of optimized solar hybrid mini Table 3.5 aggregates the findings from the case studies grids in 2025 (ESMAP 2019) and high charcoal prices, and summarizes the types of grid-connected eCooking all eCooking solutions except 100 percent grid/battery-­ that are likely to be cost-effective at different combi- powered eCooking are cost-effective. With high tariffs nations of tariffs and charcoal prices. With low tariffs (above $0.55/kWh), typical of solar hybrid mini grids today (below $0.25/kWh), typical of national grid (AFREA and (ESMAP 2019), 100 percent grid-powered eCooking and ESMAP 2016) or micro-/mini-hydro (Skat 2019) and high 50 percent grid/battery-powered eCooking (EPC only) charcoal prices typical of urban deforestation hotspots also become more expensive than charcoal. In contexts ($0.40–0.60/kg), all eCooking solutions are cost-­ effective. with high tariffs and low charcoal prices, only the most In peri-­ urban areas, where charcoal prices are typi- efficient forms of eCooking (boiling heavy foods on an ­ cally at medium levels ($0.20–$0.40/kg), 100 percent EPC) can compete with charcoal on cost. TABLE 3.5 Range of opportunities for cost-effective eCooking that open up at different tariff levels  TYPICAL TARIFF RANGE ($/KWH) NATIONAL GRID AND OPTIMIZED SOLAR MINI-/MICRO-HYDRO HYBRID MINI GRIDS IN TYPICAL SOLAR HYBRID TODAY 2025 MINI GRIDS TODAY TYPICAL CHARCOAL COST RANGE LOW MEDIUM HIGH AREA ($/KG) (LESS THAN 0.25) (0.25–0.55) (MORE THAN 0.55) Urban High All eCooking All AC and fuel stacking Fuel stacking AC deforestation (0.40–0.60) battery-supported eCooking hotspot eCooking Peri-urban area Medium All AC and fuel stacking Fuel stacking AC Most efficient eCooking (0.20–0.40) battery-supported eCooking eCooking Rural charcoal- Low Fuel stacking AC Most efficient eCooking Most efficient eCooking producing (0.10–0.20) eCooking region Note: Figures are based on modelling outcomes using projected component costs for 2025. “All eCooking” includes 100 percent grid/battery-powered eCooking, 100 percent grid-powered eCooking, 50 percent grid/battery-powered eCooking (EPC only), 50 percent grid-powered eCooking (EPC only) and boiling heavy foods in an EPC. “All AC and fuel stacking battery-supported eCooking” includes 100 percent grid-powered eCooking, 50 percent grid/battery-powered eCooking (EPC only), 50 percent grid-powered eCooking (EPC only) and boiling heavy foods in an EPC. “Fuel stacking AC eCooking” includes 50 percent grid-powered eCooking (EPC only) and boiling heavy foods in an EPC. “Most efficient eCooking” includes only boiling heavy foods in an EPC. 78 ESMAP  |  Cooking with Electricity: A Cost Perspective The global perspective understanding of the range of opportunities for eCooking by ­including grid-connected and mini-/micro-hydropower Figure 3.36 shows the outlook for eCooking at a global system architectures, as well as fuel-stacking scenarios. level by comparing the range of costs of the eCooking Input data were drawn from across the four case study coun- technologies explored in this paper with those of the most tries (Kenya, Zambia, Tanzania, and Myanmar) and the three widely used cooking fuels. Leach and Oduro (2015) and both system architectures (grid, mini grid, solar home system). Beyond Fire papers (2016, 2019) directly compare the cost Appendix F describes the input data and assumptions. of cooking with a range of different electric cooking system architectures. However, all of these models were based on The results show that AC eCooking on national grids or secondary data or laboratory data for energy demand. mini-/micro-hydropower is already cost-effective for many people today and that battery-supported DC eCooking In contrast, Figure 3.36 is based on empirical data for and solar-hybrid mini grids become cost-effective in 2025, energy demand from the cooking diaries. It extends the although clean fuel stacks with LPG can make all of these FIGURE 3.36 Comparison of system architectures using aggregated data from all case studies Year Electrical System Architecture Fuel stacking Cooking fuel 2020 AC without household battery AC eCooking / Battery-supported DC eCooking Charcoal 2025 Battery-supported DC LPG / AC eCooking LPG (household battery) Battery-supported DC eCooking / LPG Firewood 100% Electric Clean Fuel Stack (Electricity and LPG) Cooking Fuels 40 35 Cost of Cooking Service ($/month) 30 25 20 15 10 5 0 Note: The cost of cooking service is calculated over a five-year financing period for all system architectures. The range on each bar represents sensitivities to energy demand, to the grid tariff or solar resource and to key system performance and cost parameters. The ranges for energy demand are derived from the range of median values from the four country cooking diary studies for 100 percent eCooking (0.87–2.06kWh/household/day). The ratios of energy demand for cooking fuels: electricity calculated from the cooking diaries were used to model demand for LPG (2: 1), charcoal (10: 1) and firewood (10: 1). Grid-connected system architectures use a tariff range encompassing 90 percent of Sub-Saharan African utilities from AFREA and ESMAP (2016): $0.04–$0.25/kWh. National grids and mini-/micro-hydropower are grouped together, as tariff ranges are almost identical ($0.05–$0.25/kWh for mini-/micro-hydropower) (Skat 2019). Solar hybrid mini grid system architectures use a current tariff range of $0.55–$0.85/kWh and a range of $0.25–$0.38/kWh in 2025. The solar resource range is the range of average monthly solar irradiation in the least sunny months in each of the four case study countries (3.68–4.30kWh/kWpeak). eCook system performance and cost ranges are as reported in Table 2.3. Batteries are LiFePO4, sized to meet 100 percent and 50 percent of daily cooking loads, at 1–3kWh and 0.34–0.98kWh, respectively. PV is 300–700W for 100 percent and 100–200W for 50 percent. For full details of modelling input and output parameters, see appendix F. eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 79 technologies cost-effective today. Cooking with AC grid elec- of electricity and LPG can make battery-supported eCook- tricity can be the cheapest option for many people ($3–$17/ ing cost-competitive for some households today ($6–$29/ month), but is not always possible due to access and grid month). stability challenges. Supporting 50 percent of cooking loads with a battery increases the cost of cooking ($5–$22/month The critical role of energy-efficient appliances in 2025) but is still competitive with LPG, charcoal, and fire- wood ($6–$24/month, $5–$41/month, and $0–$23/month, Both energy-efficient appliances and fuel stacking can respectively in 2025). Supporting 100 percent of the cooking substantially reduce the costs of battery-supported electric loads increases the cost substantially (to $8–$39/month) cooking (Figure 3.37). An uninsulated four-plate cooker but may still be competitive in contexts with low tariffs and and oven may be cost-effective for households with low energy demand. By 2025, the cost of cooking with AC reliable grid electricity and low tariffs ($7/month at $0.04/ appliances connected to solar hybrid mini grids ($8–$25/ kWh). It is unlikely that anyone would consider supporting month) and with DC appliances powered by solar home it with a battery, which would need 4.56kWh capacity ($28/ systems ($11–$24/month) become competitive. LPG can play month even at $0.04/kWh). In contrast, the appliance stack an important role as a transition fuel, as a clean fuel stack of uninsulated (hotplate, induction, infra-red cooker, or Impact of energy-efficient appliances and fuel stacking on cost of AC and battery-supported FIGURE 3.37  DC eCooking Uninsulated four-plate with Appliance stack of uninsulated (hotplate, induction, 100% grid eCook integrated oven and infra-red cooker or kettle) and insulated (EPC, (AC w/o battery) rice cooker, and electric frying pan or thermo-pot) Uninsulated single-plate 100% grid-battery-eCook (hotplate and induction or Single insulated and pressurized appliance (EPC) (DC w/battery) infra-red cooker) 40 35 Cost of Cooking Service ($/month) 30 25 20 15 10 5 0 100% of household 50% of household Boiling heavy 100% of household 50% of household Boiling heavy cooking cooking foods only cooking cooking foods only Note: The cost of the cooking service is calculated over a five-year financing period for all system architectures. Component costs are from 2025. The range on each bar encompasses 90 percent of Sub-Saharan African utility tariffs from AFREA and ESMAP (2016) ($0.04–$0.25/kWh). Daily household energy demand values are from Figure 2.2 (100 percent eCooking: uninsulated plate with oven, 3kWh; uninsulated single plate, 2kWh; appliance stack, 1.5kWh; 0.5kWh. 50 percent eCooking: EPC, 0.5kWh. Boiling heavy foods only: EPC, 0.15kWh). Fuel-stacking scenarios model only the eCooking service, not the cost of the cooking fuel. 80 ESMAP  |  Cooking with Electricity: A Cost Perspective kettle) and insulated (EPC, rice cooker, electric frying pan, reduces overall costs (from $16–$37/month to $13–$29/ or thermo-pot) can offer a much more affordable solution month), as the battery capacity is reduced (from 2.85kWh that is capable of covering 100 percent of a household’s to 2.14kWh). Although it cannot cook all food types, the everyday cooking needs. It would cost $4–$13/month EPC is likely to be an attractive first step into eCooking for for AC and $13–$29/month for battery-supported DC. many, as it can deliver the cheapest cooking service by Simply cooking with a single uninsulated appliance may some considerable margin. Systems could be designed to be slightly cheaper for some AC users ($3–$15/month), as cook 50 percent of the menu (at a cost of $2–$5/month for the upfront cost of the appliance is lower (modelled at $20 AC or $5–$11/month for battery-supported DC) or simply as opposed to $70) for the appliance stack). For the DC what the EPC does most efficiently, which is boil heavy systems, the cost of the battery dominates, so spending foods (at a cost of $2–$3/month for AC and $3–$4/month more on an additional energy-efficient appliance actually for battery-supported DC). eCooking in Grid-Connected and Off-Grid Systems: Modelling Results and Discussion 81 Ch apter 4 DELIVERY APPROACHES This section explores the delivery models that may be well Table 4.1 provides an overview of the delivery models suited for promoting eCooking solutions depending on discussed in the following section, highlighting examples the system architecture and value chain players involved. of how each approach could be applied. Of course, the eCooking appliances and systems should be integrated into most effective solutions will often combine these delivery existing delivery infrastructure as much as possible and use approaches. For example, a solar-hybrid mini grid developer payment mechanisms that have already been established by wanting to stimulate demand for electricity may partner with energy service delivery players and the electric appliance local women’s groups, which can act as sales agents by industry. carrying out live cooking demonstrations at group meetings. Delivery Approaches 83 TABLE 4.1 Applicability of various delivery approaches to each system architecture SYSTEM ARCHITECTURE DC BATTERY–SUPPORTED TOOLS AND APPROACHES AC NATIONAL GRID OR NATIONAL GRID OR DC OFF-GRID SOLAR HOME ENABLING DELIVERY MINI GRID MINI GRID SYSTEM Electricity price signaling E.g., time-of-use tariff incentivizes cooking during daylight hours on solar mini grids. On-bill financing from service E.g., existing prepaid utility customers repay cost of providers/utility appliance every time they purchase tokens. Cash purchase from service E.g., existing mini grid customers buy appliance at community providers/utility cooking demonstration. PAYG E.g., solar home system company offers existing customers upgrade from lighting to cooking system. Cash purchase from commercial E.g., appliances are sold at distributor/retailer networks supermarkets. Productive use E.g., eCooking appliances are paired with irrigation pumps to allow firewood collectors to earn income to make repayments. Peer-to-peer women-led product E.g., women food bloggers produce eCooking content and sell appliances to their social distribution models media followers Consumer lending institutions E.g., women’s savings groups set up revolving funds to purchase cooking appliances. Note:   Highly applicable business model;  Potentially applicable business model can employ more typical marketing strategies and may help 4.1.  Appliance Value Chain reach more consumers, but it also may increase margins along the distribution and retail value chain. If people are to be able to use electricity for cooking, Appliances are often considered part of the retail process. a supply chain for appliances needs to be in place that Therefore, utilities that plan customer connections may not matches consumer demand and the load management of the consider the supply side and value chain of appliances. supply. Many distribution models for eCooking appliances Supplying and financing electric appliances, creating a exist. The two most basic options are service providers supporting industry of return and repair, and making consum- (such as utilities, operators, and institutions) and commercial ers aware of the benefits of such appliances requires planning distributor and retailer networks. by and coordination of different actors in the value chain. Deployment of appliances through service providers Quality assurance is key to ensuring that the most efficient, represents a more consolidated bulk approach, in which durable, and affordable appliances that match consumer service providers can bundle the appliance with existing preferences are facilitated for market entry. Quality assur- services to their customers. Using a more decentralized ance aspects were key to the growth of the off-grid solar distribution approach through distributor retailer networks, market. The Lighting Africa and Lighting Global programs such as those working with fast-moving consumer goods, ensured that only products verified for performance and 84 ESMAP  |  Cooking with Electricity: A Cost Perspective durability were supported. These products began to gain the balance in Latin America, Africa, and the Middle East. market share through brand recognition and increased The majority of EPCs are made in China on commission. investments, displacing poorly performing appliances that had created market spoilage in the sector. eCooking appliance value chains are already well estab- lished in a number of low- and middle-income countries. Bulk orders can help achieve economies of scale, driving Analysis of data from Seair Exim Solutions (2020) reveals down the unit cost of appliances. But identifying the most that in the last six months of 2019, the top five importers appropriate appliance is often difficult. CLASP’s Global LEAP in Kenya brought a total of $12 million worth of eCooking (Lighting and Energy Access Partnership) Awards program appliances into the country. Over this period, 330,000 provides incentives to the manufacturers of energy-­ efficient electric kettles (the most popular eCooking appliance) were appliances to focus on market products that have been imported, followed by ovens/cookers (74,000) and micro- identified as high-quality, easy to use, affordable, and energy wave ovens (63,000). EPCs are gaining in popularity, but efficient (Global LEAP 2020). The program was originally import volumes are still orders of magnitude lower in Kenya. established to demonstrate the viability of off-grid appliance An in-depth market assessment is recommended in order to sales (solar lanterns, DC refrigerators, or TVs) to commer- understand which cooking appliance brands are being sold cial lenders and appliance manufacturers that are not yet in developing countries and the distribution networks and engaged in the market. The program has expanded into marketing approaches that are being used. the grid-connected market. A Global LEAP competition for EPCs was launched in 2020, in collaboration with the MECS program. Such an exercise is useful for development programs wanting to facilitate the uptake of eCooking as well as for energy service providers wanting to offer the 4.2.  Peer-to-Peer best-in-class appliances their customers can afford (in terms of both upfront costs and ongoing consumption). Women-Led Product The global EPC market exceeded $580 million in 2018 Distribution Models (IMARC 2019). Approximately 70 percent of the market for these cookers is households, with the balance restau- rants and institutions. The market leader is Instant Pot The peer-to-peer delivery model works by recruiting cookers, which are sold mainly in developed economies. sales representatives who can tap into their own social In 2018, 45 percent of EPCS were sold in the United States; networks. The model relies on word of mouth and capital- 25 percent in the European Union; 20 percent in Asia; and izes on the fact that trust and familiarity between the sales Delivery Approaches 85 representatives and the consumers (family, friends and credit model (for example, scratch cards sold to top up acquaintances) can be more persuasive than conventional customer accounts). Monitoring of payments and system retail methods. use occurs through machine-to-machine technologies and Internet-of-Things integration that send information via GSM An example of this model is Solar Sister, an organiza- (Global System for Mobile communications) networks to tion that recruits, trains, and mentors sales reps who are system management centers and facilitates real-time data expected to invest their own capital to buy the products communication and remote monitoring of energy demand, and then resell them, first to family members and friends, time-of-day usage for appliances, and so forth. PAYG then to friends of friends, and finally to the community at providers are incentivized to offer quality after-sales service, large (Chepkurui, Leary, Minja, et al. 2019). Although this because ongoing payments are tied to the system continu- model could work for efficient eCooking appliances without ing to function. batteries, such as EPCs, it would need to be adapted to focus on finding new subscribers for services that involve There are two main approaches of PAYG financing. Under making ongoing payments to spread the costs of more a lease-to-own system, consumers pay a fixed fee at set expensive battery-supported eCooking products, which the intervals until the total value of the system plus financing is modelling results suggest will likely cost several hundred paid off, at which point, they become the owner of the equip- dollars. ment. Under a fee-for-service arrangement (similar to a utility model), consumers pay for the service for the duration of the The aspirational nature of modern eCooking appliances is contract (typically long term), but ownership remains with the likely to be a strong driver in attracting new users. Watching company. someone one knows cook one’s favorite dishes and interacting directly with her could help overcome some of The current repayment horizon for energy supply systems the initial reservations about this new technology. Another designed to power lighting, TV, and other low-power advantage of the peer-to-peer business model is that sales appliances is typically one to three years. Systems sized agents can offer after-sales services, supplying parts, such for cooking will need to be an order of magnitude larger, as sealing rings for pressure cookers, and offering friendly which would push up the size of each repayment signifi- advice on how to make the tastiest meals with this new cantly. Increasing the recovery period to three to five years equipment. would reduce the required daily/weekly/monthly customer outlay, but it would also increase the need for longer Leveraging existing social media communities (through warranties from manufacturers. However, where a house- both physical and digital channels) could greatly expand the hold is currently paying for cooking fuels, payments toward scalability of the peer-to-peer business model as a market- new eCooking appliances would be offset by reductions in ing strategy for eCooking. Some cooking-themed Facebook expenditure on cooking fuel. groups in East Africa have over 1 million users, and local food bloggers regularly receive hundreds of thousands of hits Under the fee-for-service model, payments are typically on their video recipes on YouTube (Chepkurui, Leary, Numi, made when the consumer needs and can afford power. et al. 2019). This model is more compatible with the way many biomass or kerosene users pay for their fuel and with the longer repayment horizons that will be needed to make larger cooking systems affordable. Companies such as BBOXX use the fee-for-service model, in which consumers never 4.3.  Pay-as-You-Go own the system but instead pay for the ability to use it. Under the lease-to-own model, the customer eventually Models becomes responsible for maintaining the system. Doing so can be particularly challenging when expensive components with short life expectancies (such as batteries) A number of enterprises providing energy for lighting solu- inevitably fail. Under the fee-for-service model, the tions are leveraging digital consumer financing to enhance company (or utility) retains responsibility for maintaining the affordability of their products and services. M-KOPA, a the system over the contract period. With either approach, solar home system company based in Nairobi, has deployed service can be interrupted when the user runs out of credit PAYG to reach 3 million people with 750,000 units across or the financing payment is not made. Under the ownership Kenya, Tanzania, and Uganda. PAYG customers typically model, the system is lockable until the full amount of the make payments via mobile money or an agent-based energy loan is paid. 86 ESMAP  |  Cooking with Electricity: A Cost Perspective A growing number of specialized companies are now offer- ing value chain services for PAYG. This model reduces entry 4.4.  Productive costs for new companies, which can focus on their business model and relationship with customers instead of building Applications technology and systems that can now be handled by third parties that focus on building and maintaining such plat- forms. For many PAYG companies, the challenge is manag- As an alternative to designing customized systems to ing an ongoing financing relationship with lower-income power an eCooking appliance, it is possible to simply plug customers. Once established, new products and services an eCooking appliance into a larger solar home system can be offered to existing customers. Upon completion of a designed for other purposes, such as solar irrigation (see financed energy purchase, customers build a credit history Section 3.4). Although smaller systems designed for lighting and can hence become eligible for additional products using and other low-power applications would be overloaded by the stream of expenditures that helped them pay for other an eCooking appliance, many larger systems designed for appliances. As in the case with utility value-added-services, productive applications could support them with existing or some distributed energy service companies offer new additional storage capacity. Such systems could open up products and services as a way of moving customers up a new markets, as the productive applications enable a direct services or product ladder that caters to customers’ specific repurposing of time spent on fuel collection and tending preferences. Cooking could be a highly desirable service fires into income-generating activities. This additional that could both encourage existing customers to upgrade income could substitute for the lack of existing expenditure and attract new customers. on ­cooking fuels in rural communities where firewood is collected for free. The PAYG model may yield higher gross margins than direct cash sales, but it also has higher operating costs and risks Cooking can also be a productive use of energy. Many associated with default. PAYG businesses also require regu- restaurants in Sub-Saharan Africa already use task-specific lar fundraising for covering working capital costs to cover eCooking appliances, such as rice cookers and kettles. An their receivables and can be complex in their organization, early opportunity for cost-effective eCooking that has the especially if they cover services such as financial services. potential to increase revenue generation for street vendors Whether the PAYG model will be suitable for delivering and other microenterprises is precooking heavy foods, such eCooking solutions will depend on many factors, including as beans, in an EPC (see Case Study 4). the cost of appliances and systems to be financed (which can be high if dedicated batteries are included), customers’ ability to pay, the financing plan, and other features. Delivery Approaches 87 Utilities with flexible metering and detailed data on load 4.5.  Utility Model: profiles could offer discounted tariffs during off-peak times, when there is surplus power in the system, to encourage Cooking as a Service usage at these times in order to smooth the load profile. Although collecting additional revenue thanks to increased demand, such as eCooking, could improve a utility’s finan- Many utilities are starting to move toward an integrated cial position, peak loading could lead to a return of load service delivery model. Some energy service companies shedding and brown-outs on systems with limited genera- have started to shift their business model toward service tion capacity. Energy storage and smart-charge controllers packaging and delivery, going beyond selling electrons are likely to play a key role in this business model. Load and deriving value from establishing strong relationships management for cooking needs to be deeply embedded in with customers based on understanding their needs and all electrification planning. aspirations. Through such relationships, it will be possible to stimulate demand more organically and include a range of productive use and consumer appliances according to customer demand. New distributed and digital technologies 4.6.  Distribution through will become important tools through which innovative utilities distinguish themselves by developing a proactive and value- Consumer Lending driven approach to customer relationship management. Institutions Such approaches constitute value-added services to enhance customer experience but also maximize revenue. For example, utilities may offer electric appliances as part of Where microfinance institutions and savings and credit a special promotion, bundling them with existing services, cooperatives (SACCOs) are strong, they can sometimes offering on-bill financing, and amortizing the cost through double as both distribution/retail actors and financiers of utility bills (in a manner similar to that of PAYG companies energy-efficient appliances. The availability of consumer that include the price of appliances in the service fee financing from microfinance institutions has been one of the charged to the customer). Cooking as a service could consti- biggest drivers of pico PV lanterns and, to an extent, solar tute such a value-added service. Integrating eCooking thus home system sales globally. calls for utilities to become more efficient and agile, which means using different business models and offers to their Microfinance institutions can also establish agreements customers. with well-known and high-quality brands and manufactur- ers with reliable warranties, in order to mitigate the risk In urban centers where grid connections are strong enough to of nonpayment by their members. Working with eCook- supply additional demand from eCooking, the willingness of ing appliances brands (such as brands that have been distribution network operators to facilitate eCooking needs to screened and recommended by the Global LEAP Awards), be considered. Many distribution utilities struggle to maintain microfinance institutions and SACCOs can help deliver quality of service to existing customers; expanding the grid and finance high-quality products to their members. In to new customers typically requires government subsidies. countries where microfinance institutions do not have a Utilities thus need to think of ways to increase revenue from strong presence, users will be left without this option or existing connections as a way of planning for improvements in subject to high premiums from a limited range of microfi- other areas of their business, including grid expansion. nance institutions. 88 ESMAP  |  Cooking with Electricity: A Cost Perspective Ch apter 5 FINANCING THE TRANSITION TO ECOOKING 5.1.  Consolidating Investment Strategies Technological advances are helping make electric solutions to afford the upfront investments in appliances or systems an affordable new path for increasing the pace of progress covering eCooking. Distributors and retailers will require toward both the electricity access and clean cooking goals additional access to working capital to be able to finance the of SDG7. For some of the 2 billion people who have access systems and roll out supporting services related to eCooking to reliable electricity but nevertheless cook with biomass, over the potentially lengthy repayment periods, depending AC eCooking and battery-supported cooking are already on the terms offered. The sector as a whole will require affordable and less expensive than the high and increasing financing as part of a consolidated investment strategy that costs of traditional fuels. considers eCooking as one of the areas where incremental financing can make a big difference in closing the gaps in Large-scale financing mechanisms are largely unavailable electricity generation, supply infrastructure, and demand for clean cooking as a stand-alone sector. By introducing a stimulation. “single investment strategy,” incorporating clean cooking into the growth of the electrification sector and renew- able energy technology for grid and off-grid development, the various financial instruments currently in play in these sectors could encourage both growth of energy access through renewable energy and utilization of this energy for clean cooking. As renewable energy investments grow in the coming years, clean cooking has an opportunity to leverage instruments available in the renewable energy space, such as long-term loans, guarantees, and project bonds to bridge the shortfall in meeting the SDG7 clean cooking targets. Simply mobilizing further financing is not sufficient, however. This financing needs to be directed to the key aspects of the value chain to fill the investment void, in particular the innovative delivery models discussed in the previous section and innovative financing Financing will be needed across the spectrum of the value chain, as much as possible building on the mechanisms being used to mobilize finance for electrification and renew- able energy projects. End-users will require credit to be able Financing the Transition to eCooking 89 FIGURE 5.1 Market financing of electric cooking appliances Service Provider • On-bill financing mechanisms established by utility or mini-grid developer • Asset financing provided by specialist organization or financial institutions • PAYG via o -grid service provider Manufacturers Distributors Retailers Consumers • Capital investment • Trade finance to enable • Loans for bulk • PAYG or other for R&D procurement activity purchases of Consumer financing Uses of • Working capital to • Working capital for o -grid appliances mechanisms to Financing sustain production marketing, after-sales • Loans or equity enable product • Brand building and support, and investment to support purchases marketing maintenance consumer financing Source: Adapted from Global LEAP (2018). reframing of the eCooking concept as a repurposing of 5.2.  Financing the Cost of household expenditure to support the roll-out of electrical infrastructure (whether national grid, mini grid, or off-grid PV). eCooking for Households This proposition could therefore attract private and govern- ment investment in a way that improved cookstoves have not. Biomass stoves can be purchased for as little as $2–$10 The competitiveness of eCooking is highly dependent across Sub-Saharan Africa, but they tend to be lower-tier on financing options, which are needed to mitigate the stoves, with limited efficiency improvements and durability upfront cost of devices to consumers. The case studies profiles. Higher-tier stoves cost $30–$50 but reduce ongo- show that in some contexts, the discounted cost of a ing fuel expenditure. The high upfront cost of these stoves range eCooking solutions over 5 or 20-year financing has limited uptake. Electric hot plates are relatively inexpen- scenarios is already lower than expenditures on biomass sive (typically $10–$30), but the ongoing expenditures on over the same period. However, cash flow that aligns with electricity are relatively high because of their low efficiency. consumer spending remains a key constraint. An EPC Efficient electric appliances are typically more expensive, that can cook a meal for 1/10th the cost of charcoal but with basic EPCs typically retailing for $50–$100. involves an upfront cost that is orders of magnitude higher is not an attractive proposition. Delivery models must As is the case for many renewable energy and energy-­ enable consumers to pay in a way that is compatible with efficient technologies, the cost of efficient eCooking how they currently pay for biomass. solutions is heavily weighted toward CapEx, with savings possible for poorer consumers only if the initial payment is Innovative business models could enable direct substitu- not prohibitively high. Integrating a battery into the appli- tion of daily/weekly/monthly charcoal expenditure and a ance will likely increase CapEx (although proportionally 90 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE 5.2 Range of appliance financing options for utilities and mini grid developers Low Involvement High Involvement Creates open platform Partners with nationwide Finances the purchase of for customers to share retailer to market appliances certain appliances for select repayment data with to prequalifies customers. Retailer customers on utility’s balance potential lenders o ers financing and shares sheet, raising funds from bond interest revenue with the utility. markets or local banks. Source: Waldron and Hacker (2020). much less for efficient appliances than for inefficient ones). customer segments, for example), demand-side subsidies Consequently, creative consumer financing will be essen- tied to the incremental cost of eCooking that target the more tial to enable poorer consumers to access this potentially vulnerable segments of society could be implemented, with transformative opportunity. For example, utilities or mini help from the government or development partner programs. grid developers with excess generating capacity wanting to encourage more demand could offer the initial cost of an EPC on a PAYG lease basis through on-bill-financing. Subsidy may not be required; instead, the upfront costs could be 5.3.  Financing Developers’ spread over many months on the regular electric utility bill. Capital Expenses and Lifeline tariffs may be another useful instrument for financing some of the ongoing cost of cooking with electricity (see Working Capital Section 3.2). They subsidize the rate up to a certain number of kWhs, which is often enough to cover basic needs. Cooking with electricity may fall partially under the lifeline Like many renewable energy technologies, eCooking tariff and partially above it, making the ongoing cost of solutions, in particular battery-supported models designed cooking more expensive. Targeted subsidies tied to extend- for weak grid and off-grid environments, tend to have high ing the lifeline tariffs to enable cooking for households in upfront costs. Uptake of eCooking will depend substantially need could be designed. Restricting the price subsidy to the on the willingness of the private sector—solar companies, initial block of consumption offers a less costly alternative mini grid operators, and utilities—to adopt the technology as to across-the-board price subsidies while preserving their part of the suite of services offered to their customers. politically attractive universal protection feature. In the case of financing for mini grids, high upfront costs and The case study modelling explored the costs of cooking for long-term payback are particular challenges for developers. both lifeline and regular tariffs. It shows that by using highly However, recent technology innovations on metering and efficient appliances and/or cooking only part of the house- control processes by firms such as Powerhive, SteamaCo, hold’s food with electricity, households can eCook within SparkMeter, and Inensus are enabling innovations through existing lifeline tariff thresholds. Extended lifeline tariffs prepaid smart metering, mobile payments, load limits, and would allow more cooking to fall under the first (subsidized) remote monitoring/control to improve mini grid operations block of consumption. Doing so could be one targeted and offer proactive customer care. The upfront investment way of enabling the bottom of the pyramid segment of the for new consumer appliances could be made by the supply population to cook with electricity. Where such schemes may company and recovered through sales of electricity. For this not be viable, because of the inability of the utility to finance to be possible, however, supply companies will often need these lower tariffs directly (via cross-subsidies from other upfront financing. Financing the Transition to eCooking 91 Debt finance may be an effective instrument in cases where consumers without such subsidies. Financing could be larger companies with established track records (such as scaled up proportionately to cover the incremental costs of PAYG solar companies and private utilities) are looking to cooking with electricity while maintaining the tariff at similar expand to eCooking. However, as experience from the solar levels. Eligible appliances could also receive a subsidy at sector shows, many companies do not qualify for these the manufacturer or distributor level, with the remainder loans, for a host of reasons, including inadequate collateral paid by the customer in cash or installments. Preinvestment and the perception of risk by lenders. The establishment of support is also crucial in carrying out assessments related concessional credit facilities (for example, Lighting Global- to the integration of cooking loads and system optimization, supported programs) and guarantees to capitalize and understanding the market for eCooking options. de-risk commercial loans are providing some assurance to lending institutions, which as a result are able to lend to Many strategic investors have recently entered the off-grid developers at market interest rates but at longer loan tenors and mini grid sectors, including ENGIE, EDF, Total, Shell, than are typically available to them. Mitsubishi, Caterpillar, Schneider, and General Electric. The strength of these multinational companies lies in their CapEx subsidies, often provided by government programs ability to work across energy systems, bringing techno- or development agencies, have played a major role in many logical innovation, research and development resources, energy infrastructure projects, including the development capital, and financial and operational discipline to their of mini grids. Electricity tariffs are often not affordable to subsidiaries. 92 ESMAP  |  Cooking with Electricity: A Cost Perspective A results-based financing mechanism could be tied to clear 5.4.  Results-Based measurement and monetization of impact units for verified climate, health, and gender impact results. Work is under- Financing and Impact- way at the World Bank through ESMAP’s Clean Cooking Fund (ESMAP 2019c) to establish a dedicated source of Linked Financing financing to pilot this approach using established method- ologies. Critical features underpinning the monetization of these benefits include (a) the development of widely agreed Results-based financing offers a particularly attractive means methodologies for measuring and monetizing the impacts; of achieving development goals. It can be tied to outputs or (b) credible, independent, third-party verification of results; outcomes of an intervention, such as the number of inde- and (c) clear demand for financing of the verified results by pendently verified appliances sold to customers or the number donors and impact investors. of new customers with cooking service plans. Such finance would allow for partial compensation of the service provider Impact-linked finance is another important emerging area, for results (outputs), which they could reinvest in expansion of which lies at the intersection of blended finance, impact company operations. Results-based financing schemes have investing, and results-based financing. It refers to linking been applied to improved cookstoves and solar home systems financial rewards for market-based organizations to the in various programs supported by development agencies, achievements of positive social outcomes. It goes a step including EnDev, the World Bank, SNV, and others. beyond verification of connections or sales, proposing to mainstream outcomes (such as end-user welfare) as part Other forms of results-based financing also present promis- of payment. Impact performance metrics can be linked to ing transformational opportunities for promoting clean cook- different financing instruments. It has the potential to attract ing solutions. As cooking with electricity produces virtually further investment and de-risk lending. This approach can zero kitchen emissions, it can have a significant impact on attract donors and impact investors because of the stronger offsetting traditional fuels, which emit harmful particulate link of the funding to the desired outcomes. The financing matter at the household level, causing a range of respiratory mobilized could help companies obtain working capital and complications and diseases. An early form of such support open new markets by bridging the gap in contexts where was carbon finance. Additional quantifiable co-benefits with eCooking is not quite cost-effective yet. monetizing potential have started to emerge. Financing the Transition to eCooking 93 Ch apter 6 DISCUSSION, RECOMMENDATIONS, AND AREAS FOR FURTHER RESEARCH Enabling eCooking as a mainstream solution in developing This section lays out the changes needed, the actors countries requires a change in the current narrative that concerned, and the concrete actions that the development portrays electrification and clean cooking as two separate community can take to support these processes. Each problems. The mindset of actors in both spheres has to recommendation is accompanied by a short discussion and a evolve so that eCooking is seen as a viable alternative to table identifying whom to target and how. cooking with biomass and a valuable anchor load for electri- fication programs at all scales. 6.1.  Support Policy Makers’ Efforts to Create an Enabling Environment that Bridges the Division between the Electrification and Clean Cooking Sectors National governments need to develop a supportive regu- latory environment and incentives for the private sector that bring together the clean cooking and electrification sectors, including both grid and off-grid service providers. CREATE INTERMINISTERIAL SPACES TO DEVELOP SINGLE INVESTMENT STRATEGIES THAT ALIGN WITH POLITICAL OBJECTIVES The evidence in this report suggests that a single investment in modern energy inclusive of cooking can offer a more cost-effective route to meeting the twin goals of access to electricity and clean cooking than investing separately in each. Currently, most developing countries have strat- egies for electrification, and some have strategies for Discussion, Recommendations, and Areas for Further Research 95 TABLE 6.1 Targeted recommendations for creating interministerial spaces TARGET ACTIONS NEEDED Countries with ambitious but separate • Create or strengthen interministerial spaces (committees, working groups, etc.) health (household air pollution); to show how working together to make eCooking a mainstream solution can gender; environmental (climate change, achieve all these objectives. deforestation); and/or energy access • Develop national planning frameworks that establish clear goals for the adoption (electrification and clean cooking) goals, in of eCooking solutions by developing modern energy strategies that (a) include which policy support is already in place cooking as a valuable anchor load in electrification planning and (b) incorporate eCooking as an option for enabling clean cooking. clean cooking. The two are almost always totally separate, CREATE A SPACE FOR DIALOGUE BETWEEN however, with grid extension programs typically planned STAKEHOLDERS IN THE CLEAN COOKING AND without considering possible eCooking loads and clean ELECTRIFICATION SECTORS cooking programs planned without electric appliances. Interweaving electrification and clean cooking strategies can Much of the knowledge and experience required to develop, enable the combined efforts of both sectors to work toward pilot, and market innovative eCooking solutions already achieving the goals of both sectors. exists, but it is divided in two. On one side is the clean cook- ing sector, which has an in-depth understanding of people’s Separation of the two sectors is often enshrined in ministe- cooking needs and aspirations and the market for baseline rial responsibilities. Bridging the gap therefore requires the fuels today. On the other side are the solar, mini grid, and highest-level political commitment. Biomass cooking affects utility sectors, which have detailed knowledge of electrical health (the Ministry of Health) and contributes to forest products and systems, the user experience of transitioning to degradation (the Ministry of Environment). The potential solu- modern energy, and attracting investment orders of magni- tion requires access to electricity (the Ministry of Energy), the tude higher. It is likely that rapid progress could be made by manufacture of importation of devices (the Ministry of Trade creating strategic partnerships. Both nationally and interna- and Industry), and a realistic budget (the Ministry of Finance). tionally, private sector, government, and nongovernmental Engaging with parliamentary and cabinet-level ministers organizations need to come together at events that jointly and setting up interministerial spaces (committees, working work on grid, off-grid, and cooking transition planning. groups, etc.) can help overcome the siloed work of ministries and integrate planning.21 TABLE 6.2 Targeted recommendations for encouraging intersectoral dialogue TARGET ACTIONS NEEDED Industry associations in the clean cooking • Sponsor events that bring together actors from both sides. or electrification sectors (such as the Global • Fund projects that require collaboration by organizations from both sides. Off-grid Lighting Association [GOGLA], the • Advocate for high-level support for modern energy access inclusive of clean Clean Cooking Alliance [CCA], and African cooking. Mini grid Developer’s Association [AMDA]) • Support targeted exchange programs for key personnel from the two sectors. and organizations whose mandate already spans both domains (for example, Ministries of Energy, SEforALL, and multilateral organizations). 96 ESMAP  |  Cooking with Electricity: A Cost Perspective STRENGTHEN THE CASE FOR THE POOR THROUGH REDUCE THE RELATIVE COST OF COOKING STRATEGIC USE OF LIFELINE TARIFFS FINANCED WITH ELECTRICITY BY DIVERTING FOSSIL FUEL BY CROSS-SUBSIDIES OR TARGETED SUBSIDY SUBSIDIES TO ELECTRICITY ACCESS PROGRAMS PROGRAMS Countries that have already seen significant uptake of fossil The provision of a lifeline tariff is a subsidy to the poor. The fuels for cooking (particularly kerosene) could alter their rela- evidence from Case Studies 1 and 2 shows that such tariffs tive price points versus cooking with electricity by diverting can be a well-targeted tool for achieving the social benefits subsidies (or adding taxes) that would redirect funds into the of clean cooking. Section 3.2, on Affordability, highlights the development of the supply chain for eCooking. Kerosene is challenges associated with increasing the use of existing a polluting fuel, with health impacts comparable to biomass lifeline tariffs in Sub-Saharan Africa, where most utilities do cooking. LPG has an important role to play as a transi- not yet recover their costs. However, lifeline tariffs are widely tion fuel, but since it cannot be produced from renewable seen as fair and a necessary instrument of social policy to sources and is typically backed by large ongoing subsidies, increase the purchasing power of the poor and provide basic it may not offer a truly sustainable pathway to achieving and services. As the electrification and clean cooking sectors are sustaining universal access to clean cooking. Where the brought together, there will be a need to balance subsidies national generation mix is substantially renewable, or in the for lifeline tariffs with other priorities. case of solar eCooking, electricity can provide such a path- way, which needs to be reflected in clean cooking strategies Optimizing energy demand by using the most efficient by positioning renewably generated electricity as a desirable appliances to maximize the benefits of existing lifeline tariff end goal. provisions will also be critical. The comparison of modelling results in 3.34 and the demand calculations in section 3.2 High-level political support can be achieved by aligning show that a lifeline tariff allowance of 100kWh/month at eCooking with existing political objectives. In Ecuador, for $0.10/kWh would make it cost-­ effective for the majority of example, induction cookers were introduced to reduce LPG households cooking with charcoal to switch to eCooking consumption through a national program that was in line with even if grid reliability were low and a battery sized for a full the country’s objective of increasing the share of renewable day’s cooking were required. energy in its energy mix, as the national grid is predomi- nantly hydro-powered (Gould et al. 2018; Parikh et al. 2020). TABLE 6.3 Targeted recommendations for using lifeline tariffs TARGET ACTIONS NEEDED Utilities and regulators that set lifeline tariffs • Optimize energy demand by encouraging the use of the most energy-efficient appliances to make the most of existing lifeline tariff allowances. • Ensure that utilities are still financially viable with increased consumption at the lifeline rate by (a) ensuring cost-reflective regular retail tariffs; (b) securing official development assistance or investing national budgets for gender, health, or environmental goals; and (c) diverting fossil fuel subsidies. Utilities with tariffs above $0.10/kWh • Enable poorer households to cook with electricity by implementing lifeline tariffs of at least 100kWh/month at $0.10/kWh or below (threshold at which, according to author analysis, would enable poorer households cook with electricity cost-effectively). Utilities with a large proportion of customers • Enable poorer households to access lifeline tariffs by (a) developing on-bill on shared meters financing for connection fees, (b) reducing connection fees for additional meters in the same building, and (c) redesigning meters to enable multiple connections. Mini grid developers • Create comparable lifeline tariffs by accessing the same subsidies as utilities. Discussion, Recommendations, and Areas for Further Research 97 TABLE 6.4 Targeted recommendations for diverting fossil fuel subsidies TARGET ACTIONS NEEDED Countries with fossil fuel subsidies for • Conduct evidence-based research on the benefits and drawbacks of subsidy cooking fuels diversion into (a) financing the higher upfront cost of energy-efficient and battery-supported eCooking solutions that can ultimately reduce both the cost of cooking for consumers and national expenditures, (b) extending lifeline tariffs to make eCooking more attractive to poorer households, and/or (c) providing tax exemptions for key components (energy-efficient eCooking appliances, as well as higher-capacity lithium-ion batteries). • Conduct evidenced-based research on “feebates” or cross-subsidies for highly efficient eCooking solutions from levies on commercialized polluting fuels and technologies or inefficient appliances (for example, kerosene/charcoal or electric four-plate cookers with ovens). STREAMLINE SUPPLY CHAINS, IN ORDER TO The Global LEAP Awards competition for EPCs is a good DECREASE THE LIFETIME COST OF ECOOKING example of how such a quality assurance program could be BY REDUCING THE UPFRONT COST OF QUALITY- implemented (Global LEAP 2020). Manufacturers are invited ASSURED ENERGY-EFFICIENT APPLIANCES to submit their products for testing in both lab and field-test- ing categories. The key output is a buyer’s guide that can To avoid market spoilage associated with poor-quality prod- inform future bulk purchasing of high-quality appliances that ucts, there is a need to reduce the upfront cost and increase are well matched to the needs and aspirations of consumers the availability of high-quality eCooking appliances. To in underserved markets. optimize the cost of cooking with electricity, these products should also be the most energy efficient, especially in the case of battery-supported systems, as the most expensive component (the battery) can then be considerably smaller. TABLE 6.5 Targeted recommendations for enabling quality-assured energy-efficient appliances TARGET ACTIONS NEEDED Appliance manufacturers and distributors • Support quality assurance programs, such as the Global LEAP. All countries • Develop national quality standards with energy-efficiency criteria. Countries with strong supply chains for • Identify the most appropriate energy-efficient appliances for local cooking imported products already in place practices. • Provide tax exemptions for quality-assured, energy-efficient, and culturally appropriate appliances. Countries with strong local manufacturing • Incentivize local manufacture of culturally appropriate appliances and/or industries assembly of eCooking systems, through local market development programs. 98 ESMAP  |  Cooking with Electricity: A Cost Perspective 6.2.  Conduct Strategic Evidence-Based Research to Inform Decision Makers, Private Sector Players, and Consumers of Emerging Opportunities More evidence on cooking with electricity is needed in IDENTIFY CULTURALLY APPROPRIATE ENERGY- developing country contexts. Cooking is a deeply cultural EFFICIENT ECOOKING APPLIANCES AND EXPLORE experience; only by fully understanding the compatibility of FUEL STACKING the broad range of solutions on offer with each local context can the right decisions be made. The affordability of eCooking is directly linked to the amount of energy required to cook and how much fuel stacking is Key inputs for the analysis in this report are the prices people likely to occur. Given the diverse range of foodstuffs avail- are paying for traditional fuels and how much of their current able and cooking practices used across the world, there is a cooking practice can be readily transferred to eCooking. need to match eCooking appliances with cuisines. Both of these factors vary widely, as the development of traditional fuel markets and cultural cooking practices are An array of modern eCooking appliances are now available different in each setting. Key questions include the following: on the market. However, many of them are highly specialized (toasters, kettles) or too expensive for poorer households ● Who is already paying for cooking fuel (and how much to afford. There is a need to determine which appliances are they paying)? are most desirable and offer the greatest energy and time savings on popular local foods in each local context. ● Which eCooking appliance and system architecture are Predicting where eCooking will slot into potential customers’ best matched to people’s needs and aspirations? fuel-stacking behavior is critically important: If a new eCook- ● How often will consumers actually use each device? ing device is used for only half the cooking, then only half the baseline expenditure is available to repay the cost of the device. This metric is also important for understanding how much progress is being made toward the achievement of SDG 7 and in unlocking results-based funding, in particular, climate finance. TABLE 6.6 Targeted recommendations for identifying culturally appropriate appliances TARGET ACTIONS NEEDED Local research institutions specializing in • Conduct market assessments and value chain analyses on appliances currently action research in (a) cultures with major available, and facilitate market entry for ones that are not. staples that are easy to cook in energy- • Map out local menus to establish which dishes are cooked and how often. efficient eCooking appliances (such as rice, • Test appliances in kitchen laboratory settings to establish which are most couscous, and maize) and (b) cultures in compatible with local cooking practices. which “heavy foods” represent a significant • Conduct cooking diary studies to understand how people cook and the amount portion of the menu, as EPCs can provide of energy required, and test the most promising appliances as part of real kitchen significant energy, cost, and time savings. routines. Discussion, Recommendations, and Areas for Further Research 99 TABLE 6.7 Targeted recommendations for understanding target market segments TARGET ACTIONS NEEDED Household survey designers • Include the more holistic questioning framed in the Multi-Tier Framework for energy access. Data analysts for household surveys • Analyze the clean cooking and electrification responses together, to characterize the key target market segments for eCooking. International finance experts • Develop reliable price indexes for biomass fuels. Research institutes • Conduct detailed market assessments identifying, quantifying, and characterizing the target market segments in each context. GAIN A DEEPER UNDERSTANDING OF KEY TARGET data often indicate how many people use charcoal as their MARKET SEGMENTS, IN PARTICULAR OF EXISTING primary cooking fuel and how many people live off the grid. EXPENDITURES ON COOKING FUELS However, without going back to the raw data, it is not possi- ble to know how many of these charcoal users are off-grid or Gaining a deeper understanding of who is paying for what other fuels they may be using. The SEforALL Multi-Tier cooking fuel and how much, as well as the level of electric- Framework for energy access, developed by ESMAP and ity access, is more important for eCooking solutions than its partners, includes a much broader range of questions for improved cook stoves, where people simply use less (including questions on the quality of electricity access, of the same locally produced fuel. In the coming years, PV expenditures on cooking fuels, and fuel stacking) that could panels, lithium-ion batteries, energy-efficient appliances, enable much more detailed market segmentation if the data and other system components will need to be imported in can be analyzed by cutting across the clean cooking and most contexts, so the relative value of fuel expenditures in electrification responses. local currency must be compared with the international price points of the components. Data on fossil fuel prices and utility tariffs are available; reliable data on current prices of ENHANCE TECHNO-ECONOMIC MODELS biomass fuels are often much harder to find. BY INCLUDING THE EXPECTED COSTS OF MARKETING, SELLING, AND SUPPORTING SOLAR The size of the market for specific eCooking solutions can BATTERY–POWERED ECOOKING DEVICES IN be estimated by matching it with specific customer groups RURAL AREAS based on expenditures and access to electricity. For exam- ple, a key target market segment for solar eCooking is likely The costs of marketing, selling, and supporting AC appli- to be rural charcoal users, who are likely to be off-grid yet ances in urban areas is already included in the retail price are almost certainly paying for their fuel. National survey of the appliances in grid-connected scenarios. These costs TABLE 6.8 Targeted recommendations for enhancing the modelling of solar battery–powered eCooking TARGET ACTIONS NEEDED Industry associations (such as GOGLA) • Conduct anonymized surveys of member organizations on cost estimates based on actual cost breakdowns of most similar products (larger solar home systems). Solar home system companies • Conduct feasibility studies and pilot projects to develop viable business models that incorporate these costs. 100 ESMAP  |  Cooking with Electricity: A Cost Perspective are likely to be much higher in rural areas. What is more, if it occurs when there is spare capacity. However, the although the estimated costs of shipping and import taxes increased load must be balanced against the cost of any (plus the cost of financing in the discounted monthly costs) is upgrades to the infrastructure (larger transformers, additional added to the factory gate prices of the other system compo- generation) that may be needed (Lombardi et al. 2019). nents, retail costs are not included. The cost of establishing marketing, distribution, and after-sales support network in There is a need to model future scenarios in which uptake sparsely populated rural areas with limited infrastructure may happen at scale, in order to enable planners to design is likely to be substantially higher than the relatively slim appropriate generating capacity, transmission, and distri- margins retailers add to the products they sell in urban bution infrastructure and delivery models. Multiple options areas. It is unclear exactly how much higher it is, however, as are often available. For example, a mini grid that is likely to no commercial solar eCooking products are currently avail- exceed its peak generating capacity if all customers adopt able and many companies are reluctant to share this type of eCooking, may want to compare the cost of additional commercially sensitive data. centralized energy storage or generation with the cost of decentralized household storage. Such analysis could also explore techniques to shape the load profile by influencing consumer behavior, such as flexible tariff structures (for MODEL THE IMPLICATIONS OF ENCOURAGING example, off-peak tariffs) or smart appliances/storage that ECOOKING FOR LOAD MANAGEMENT ON can be controlled by the grid operator. In many countries of NATIONAL GRIDS AND MINI GRIDS, IN ORDER interest, the regulatory regimes are complex and prescrip- TO ESTABLISH THE LIKELY IMPACT ON OVERALL tive, stifle innovation, and act as a barrier to new entrants. COSTS AND THE INTEGRITY OF THE SYSTEMS If clean cooking is to be integrated with planning for elec- trification, then detailed analysis of the implications for the Cooking with electricity puts additional load on grid infra- power system will be vital to enable policy makers to make structure, which can increase revenue for the grid operator evidence-based decisions. TABLE 6.9 Targeted recommendations for modelling load management on grid systems TARGET ACTIONS NEEDED Electrification planners, regulatory • Conduct studies on load management, to evaluate the relative costs and benefits agencies, utilities, and mini grids developers of future scenarios for managing scaled uptake of eCooking. considering eCooking Discussion, Recommendations, and Areas for Further Research 101 6.3.  Support Private Sector Efforts to Develop Products and Services Tailored to the Needs and Aspirations of the Poor ENABLE UTILITIES AND MINI GRID DEVELOPERS (see, e.g., Case Study 2). Others are keen to increase the TO DEVELOP, PILOT, AND SCALE UP ECOOKING electricity consumption per connection to increase their SERVICES THAT ARE COMPATIBLE WITH THEIR profitability in challenging markets (see Case Studies 1, EXISTING BUSINESS MODELS 3, and 4). Many utilities and mini grid developers are now taking a more holistic approach, going beyond simply selling Knowledge of and attitudes toward eCooking vary widely units to trying to understand the needs and aspirations of in the utility and mini grid sectors. Understanding them is their customers. Raising awareness of the opportunities for crucial to ensuring that support is correctly targeted. Some delivering eCooking services that align with the priorities of providers already have many customers cooking with elec- each service provider will be key. tricity and are looking to manage demand more sustainably TABLE 6.10 Targeted recommendations for developing utility and mini grid business models TARGET ACTIONS NEEDED Utilities or mini grid developers with • Conduct feasibility studies to determine (a) which grids are already able to demand stimulation programs (typically support AC eCooking and which need strengthening and (b) which energy- energy-limited grids [such as solar] or efficient appliances are most attractive to their customers and how much they are power-limited grids [such as micro-hydro] paying for their fuel. with spare capacity at peak times) • Support knowledge exchange and partnerships with the clean cooking sector to understand cooking needs and aspirations. • Establish on-bill financing mechanisms for eCooking appliances and/or subsidized appliance costs (recovering costs through sales of electricity units). • Experiment with different tariffs, including extended lifeline and off-peak tariffs, to stimulate demand. Utilities with a large share of renewable • Establish time-of-use tariffs to encourage users to cook with electricity when energy or renewable/hybrid mini grid renewable power is available. developers Utilities or mini grid developers with • If customers are not already cooking with electricity, conduct feasibility studies demand-side management programs to determine whether (a) battery-supported eCooking can also enable 24-hour (typically power-limited grids [such as micro- electricity access without overloading the grid and (b) off-peak tariffs can hydro] without spare capacity at peak times encourage users to cook with electricity without significantly increasing peak grids or grids with frequent load shedding, loading. blackouts, or voltage instability or grids that • If many customers are already cooking with electricity, explore the viability of provide power only during set hours) decreasing peak loading with demand management techniques, such as time- shifting demand with battery-supported appliances or encouraging users to adopt more energy-efficient eCooking appliances. 102 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 6.11 Targeted recommendations for producing and selling appliances that appeal to customers at the bottom of the pyramid TARGET ACTIONS NEEDED Manufacturers of energy-efficient eCooking • Conduct research and development on the most promising appliances to enable appliances with a presence in developing customers at the bottom of the pyramid to (a) cook a wider range of foods, (b) countries cook even more efficiently, (c) integrate energy metering into the device to indicate to the user exactly how much has been spent on each meal, (d) facilitate the adoption of energy-saving practices, (e) withstand blackouts and voltage/ frequency fluctuations, and (f) develop DC and battery-integrated models. • Develop longer warranties, in line with longer repayment horizons. • Establish service networks in rural and poor urban areas to make spare parts and expertise available locally. • Partner with financing institutions to enable households to use innovative financing mechanisms (such as on-bill financing, microcredit, and PAYG) to repay the high upfront cost of appliances. • Develop social marketing campaigns based on (a) cost (cheaper than other common local fuels); (b) convenience (faster cooking, cleaner kitchen, multitasking); and (c) ways to save even more time and money by cooking efficiently. • Package (or repackage) international models with advice on (a) how to cook local foods (putting stickers on EPCs indicating cooking times, for example ) and (b) energy-efficient cooking practices (recipe books, community cooking demonstrations). INCENTIVIZE ECOOKING APPLIANCE ENABLE SOLAR HOME SYSTEM COMPANIES TO MANUFACTURERS TO DEVELOP PRODUCTS DEVELOP, PILOT, AND SCALE UP INNOVATIVE NEW TARGETED AT THE BOTTOM OF THE PYRAMID, ECOOKING PRODUCTS AND SERVICES PARTICULARLY DC- AND BATTERY-SUPPORTED ECOOKING PRODUCTS Developers/distributors of solar home systems can lever- age existing customer relationships and credit histories to The needs and aspirations of poor people are often differ- offer cooking as a new energy service. They may also be ent from those of better-off people. To achieve widespread able to attract new customers, who can repurpose their uptake, eCooking appliances must be seen as accessible existing expenditures on cooking fuels to sign up for a solar but highly desirable products by the poor. eCooking appli- eCooking service that also offers lighting, phone-charging, ances are usually designed for urban elites, who will likely TV, and radio. Solar home systems are often packaged with already own an array of kitchen gadgets and be familiar with customized appliances that have been carefully selected to digital technologies. eCooking appliances targeted at poorer match the system’s power generation and storage capabili- consumers can be developed by simplifying control mech- ties. For most companies, eCooking appliances will be a big anisms, tailoring them to local foods, and ensuring that they step up, requiring a significant product redesign. Cooking is offer a strong value proposition to consumers over charcoal, a highly culturally embedded process, so understanding the kerosene, coal, and LPG. The poor are much more likely to cooking needs/aspirations of the customer base is likely to be living in off-grid and weak-grid areas. Designing DC and require much more detailed market research than lighting, battery-integrated versions of existing eCooking appliances phone-charging, TV, or radio, use of which is much more is therefore a key pro-poor action that major manufacturers homogeneous. should be encouraged to take. Discussion, Recommendations, and Areas for Further Research 103 TABLE 6.12 Targeted recommendations for developing business models for solar home systems TARGET ACTIONS NEEDED Solar home systems companies that • Support knowledge exchange and partnerships with actors in the clean cooking • already offer fee-for-service (utility) sector to understand their cooking needs and aspirations. business models (which are likely to be • Conduct feasibility studies and pilot projects with grant funding. more compatible with longer repayment horizons) • have customers paying high prices for polluting fuels and technologies • have strong relationships with their customers (to facilitate the gathering of in-depth information on their cooking needs and aspirations) • have a history of innovative product/ service design. PAYG solar companies with one- to two-year • Develop and pilot longer recovery periods (three to five years) or fee-for-service recovery periods (utility) business models. Solar home systems designed for • Conduct feasibility studies and piloting to assess the viability of adapting existing productive applications products to power eCooking appliances. ENABLE PLAYERS IN THE EXISTING CLEAN financing. Most improved cookstoves do not contain elec- COOKING VALUE CHAIN TO EXPAND THEIR trical components and are sold without consumer financing, PRODUCT RANGE TO INCLUDE ECOOKING because the upfront cost is much lower than for electric APPLIANCES appliances. Capacity building of actors in the clean cook- ing value chain will be needed to develop the specific skill Improved cookstove manufacturers/distributors are likely sets needed to design, manufacture, and support electrical to lack expertise in electrical system design and consumer products. TABLE 6.13 Targeted recommendations for enhancing the role of players in the clean cooking value chain TARGET ACTIONS NEEDED Improved cookstove manufacturers/ • Expand from using simple thermal efficiency of heat transfer from the fuel into distributors that already use innovative the cooking pot toward a more holistic understanding of how much energy is financing mechanisms to sell their products required to cook popular local foods with energy-efficient eCooking appliances. • Develop/extend innovative consumer financing (by connecting with specialist PAYG providers, for example). • Support knowledge exchange and partnerships with actors in the electrification sector to understand electrical system design and the user experience of transitioning to electricity. 104 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 6.14 Targeted recommendations for empowering women to promote eCooking TARGET ACTIONS NEEDED Women-led businesses or women in key • Empower women to develop innovative eCooking solutions within their roles in the electrification and clean cooking organizations (through targeted exchange programs between the solar and clean sectors cooking industries, for example). Electrification or clean cooking initiatives • Expand the range of products/services promoted by women into eCooking. designed to empower women as entrepreneurs as well as end-users Organizations already using women-led peer-to-peer business models for other products/services EMPOWER WOMEN ENTREPRENEURS TO LEAD IDENTIFY VIABLE BUSINESS MODELS THAT WILL THE DEVELOPMENT AND DISSEMINATION OF BOTH UNLOCK CONSUMER RESPONSES AND INNOVATIVE NEW ECOOKING SOLUTIONS MEET PRIVATE SECTOR FINANCING NEEDS No one understands the needs and aspirations of women as The flow of finance needs to be better understood, in order well as women themselves. As the primary beneficiaries of to stimulate the development of new service-orientated eCooking solutions, women should be at the center of any business models and help ensure the longer-term recovery eCooking initiative. They cannot simply be passive beneficia- of investment. For example, utilities looking to become more ries of products/services developed and marketed primar- agile and user-focused could proactively stimulate energy ily by men. Women must be empowered to co-create the demand with eCooking (see Case Study 1). Doing so may eCooking solutions they aspire to use and to leverage their require the use of on-bill financing for appliances, as well as social networks to enable successful solutions to rapidly price signaling (for example, time-of-use tariffs) to smooth out reach scale. the daily load profile. However, it is important for the market capitalization of the company to consider who owns the TABLE 6.15 Targeted recommendations for balancing consumer and private sector financing needs TARGET ACTIONS NEEDED Private sector organizations seeking to • Develop delivery models for investment scenarios (for example, discounted attract investment in eCooking, such returns and cash flow projections for PAYG models). as service providers (utilities, mini grid developers) wanting to stimulate demand via eCooking Large-scale private sector investors in • Establish partnerships with specialist asset financing companies willing to take on renewable technology, including donors the financial risk of appliance ownership. who contribute to special purpose vehicles Discussion, Recommendations, and Areas for Further Research 105 assets “on the books.” Where the operating company retains developmental outcomes, making it a promising tool for ownership, the investment is depreciated over time and the leveraging the transformational potential of eCooking. Such loss is a part of the company’s operational expenses. Where financing can support the development of supply chains for the company hands over the equipment to the consumer eCooking through bulk procurement of culturally appropri- at the end of the loan or lease period, the asset comes off ate, energy-efficient, high-quality appliances and eCooking the balance sheet. A more attractive proposition may be a systems. Social investment finance—through patient equity partnership between a service provider and a separate asset capital, soft debt finance, or crowd-sourced funding—could financing organization that could take on this financial risk. help facilitate the emerging convergence of electric modern energy provision and clean cooking. Enabling new players to explore new approaches and business models on each side of the clean cooking and electrification divide would acceler- BRIDGE INITIAL COST–VIABILITY GAPS IN ate the convergence of these sectors. NEW MARKETS BY COMBINING FINANCING INSTRUMENTS SUCH AS GRANTS, SOCIAL IMPACT INVESTMENT, AND RESULTS-BASED FINANCING TIED TO ENVIRONMENTAL, GENDER EQUITY, AND HEALTH OUTCOMES Grant funding has an important role to play in facilitating early stage experimentation among private sector actors that are curious to explore emerging opportunities. Results- based financing and impact-linked finance tie finance to TABLE 6.16 Targeted recommendations for bridging initial cost–viability gaps TARGET ACTIONS NEEDED Private sector organizations interested in • Conduct feasibility studies and piloting with grant funding. investing in or implementing eCooking solutions Established companies in the clean cooking • Forge partnerships with companies on the other side of the “great divide.” or electrification sectors Governments and donors developing • Link eCooking impacts to addressing local and global development challenges, results-based financing programs or impact- leveraging climate finance and financing linked to other impacts to mobilize funds linked financing focused on health, gender, for de-risking impact investors and service providers. and environmental impacts Private sector actors with proven eCooking • Create a toolkit to assess how attractive the various forms of results-based solutions financing may be for their products/services, focusing on (a) the displaced fuels (for example, whether the biomass fuel is sustainably sourced); (b) the utilization rate of the eCooking solution (how much of the traditional fuel will be replaced); and (c) calculation of environmental, health, and gender equity key performance indicators (such as carbon equivalent emissions reduction, disability-adjusted life years averted, women’s time saved). Social investment funds and private sector • Broker between social investors and businesses. actors with proven eCooking solutions • Develop clean cooking funds that include a focus on eCooking within their mandate. 106 ESMAP  |  Cooking with Electricity: A Cost Perspective 6.4.  Help Consumers Understand the Benefits of Adopting Modern eCooking Solutions and Reduce Barriers to Behavioral Change DEVELOP “PAY-AS-YOU-COOK” FINANCING because it can be purchased in small quantities, a regular (FLEXIBLE REPAYMENT SCHEMES BASED ON HOW monthly repayment on a battery-supported eCooking CONSUMERS CURRENTLY PAY FOR BIOMASS) device is not likely to be attractive. Prepaid electricity meters allow consumers to buy just enough units to cook If consumers are going to switch to eCooking, it must a single meal, but doing so does not allow them to see be cheaper than alternatives on a levelized cost basis, how much they paid to cook the meal, as it is unclear and consumers must be able to pay for it in the same whether the units are consumed by cooking devices or way they currently pay for biomass. For consumers who other appliances. In addition, they still face the upfront currently buy kerosene or charcoal each time they cook, cost of the appliance. TABLE 6.17 Targeted recommendations for developing “pay-as-you-cook” financing TARGET ACTIONS NEEDED Improved cookstove manufacturers/ • Develop PAYG and utility business models to reframe clean cooking as a service distributors rather than a product. PAYG companies • Offer more flexible repayment plans, including by (a) extending PAYG contracts for battery replacement; (b) aligning payment with income-generating events such as harvests, providing flexibility in repayments beforehand, and planning marketing activities to recruit new customers shortly after; and (c) leveraging progress with mobile money to enable smaller and more irregular repayments. System developers • Pair eCooking devices with productive appliances such as water pumps to enable users to repurpose the time they currently spend on fuel collection with income- generating activities. Microfinance organizations • Support interventions that can enable micro–savings and loan groups (such as self-help groups, Savings and Credit Cooperative Organisations [SACCOs], and chamas) to understand how repayment of a loan for an eCooking device can be achieved by savings on cooking fuels. Banks, agricultural finance companies, and • Conduct case studies and illustrations that show that the risk is low if the loan credit companies to consumers is for quality-assured equipment and is based on realistic data on existing expenditures. Discussion, Recommendations, and Areas for Further Research 107 TABLE 6.18 Targeted recommendations for helping consumers understand the cost of eCooking TARGET ACTIONS NEEDED Food bloggers, cooking shows, and retail • Demonstrate cooking local foods with energy-efficient appliances while outlets monitoring energy consumption to compare cost with traditional fuels. Appliance distributors and utilities • Offer appliance-level submetering via plug-in electricity meters displaying local electricity tariffs. Utilities • Tackle the issue of shared meters (see table 6.3). Appliance manufacturers • Integrate energy meters into cooking appliances. HELP CONSUMERS UNDERSTAND HOW MUCH CONDUCT PARTICIPATORY ECOOKING IT WOULD REALLY COST THEM TO COOK WITH DEMONSTRATIONS AND OFFER TRIAL PERIODS ELECTRICITY WITH LIMITED FINANCIAL RISK TO THE CONSUMER TO ENABLE THEM TO EXPLORE ECOOKING Many consumers do not even consider cooking with elec- tricity, because even people who have access to reliable People who have not cooked with electricity often worry electricity often assume it is too expensive for cooking. Even that the appliances are too complicated and the food will with prepaid meters, most consumers are unaware of how not taste as good. They need to be assured that they can much electricity each appliance is consuming. Charcoal and produce the same delicious food they are used to cooking kerosene can be bought in small quantities, and it is very and that eCooking can make the process quicker and easier. clear how much is used to cook each meal. In contrast, elec- tricity is invisible, and a meter is needed to show how much has been consumed at each point in the network. TABLE 6.19 Targeted recommendations for conducting eCooking demonstrations and offering trial periods for consumers TARGET ACTIONS NEEDED Consumers without experience cooking • Support peer-to-peer women-led initiatives that enable entrepreneurs to with electricity demonstrate eCooking to people in their social network. • Offer trial periods that allow consumers to take home eCooking devices and cook for their family without having to commit to a full service contract. • Conduct cooking demonstrations in public places, where people can taste the finished product. 108 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE 6.20 Targeted recommendations for translating evidence into easy-to-understand content TARGET ACTIONS NEEDED Consumers that: • Launch media campaigns focusing on cost (cheaper than charcoal) and • have already adopted improved cooking convenience (faster cooking and multitasking). solutions (as they have shown a • Have appliance retailers and service providers carry out cooking demonstrations willingness to change) showing how easy it is to cook popular local foods with electricity and how • have access to electricity (grid or off-grid) delicious the food is. but have not yet adopted eCooking • Improve consumer-facing communication of quality assurance and safety, with • already use a range of modern electrical regard to cooking with an EPC, electric shocks, and lithium-ion batteries. goods (as they are likely to be familiar • Have social media groups share recipes and cooking techniques. with electric devices and value modern • Use a variety of delivery mechanisms, including mainstream media outlets (social solutions) media, TV, radio, billboards, live cooking demonstrations). • use mobile money or other mechanisms that can facilitate smaller transactions TRANSLATE EVIDENCE-BASED RESEARCH INTO ENCOURAGE CONSUMERS TO COOK AS MUCH EASY-TO-UNDERSTAND CONTENT THAT CAN BE OF THEIR TYPICAL MENU ON ENERGY-EFFICIENT SHARED ON POPULAR MEDIA APPLIANCES AS POSSIBLE As new opportunities emerge and early adopters take up The cost savings from increased appliance efficiency are the new approaches, it will be important to communicate directly proportional to the amount of cooking that is done early successes to the wider population in accessible and using energy-efficient appliances. Many energy-efficient engaging formats, such as cookbooks, social media toolkits, cooking appliances are highly specialized, performing just and live cooking demonstrations. Social media, TV, and radio one task very well (an example is a toaster). Others, such provide opportunities to communicate the benefits enjoyed as the EPC or rice cooker, can be used to cook many foods by early adopters. but tend to be used for specific foods (heavy foods and rice, respectively). Encouraging users to cook as much as possi- ble using energy-efficient appliances can reduce the use of inefficient appliances and/or fuel stacking with biomass. TABLE 6.21 Targeted recommendations for encouraging wider use of energy-efficient appliances TARGET ACTIONS REQUIRED Consumers who have purchased energy- • Use local cookbooks, YouTube video recipes, and cooking demonstrations to efficient appliances showcase the range of dishes that can be cooked using the new appliance. • Train sales agents in women-led peer-to-peer business models and social media groups to share tips for making the most of their new appliances. Discussion, Recommendations, and Areas for Further Research 109 Ch apter 7 CONCLUSION This report highlights the transformative potential of eCook- 1. Support policy makers’ efforts to create an enabling ing to achieve a broad range of development goals span- environment that bridges the gap between the ning the gender, environmental, health, and energy access electrification and clean cooking sectors. domains by simultaneously enabling access to clean cooking 2. Conduct strategic, evidence-based research to inform and reliable electricity. The five case studies illustrate early decision makers, private sector players, and consumers potential markets—contexts where eCooking is not only of emerging opportunities. cost-effective but also offers additional value to consumers and/or service providers. 3. Support private sector efforts to develop appropriate technical and financial products and services tailored to New energy-efficient appliances, such as the EPC, can the needs and aspirations of the poor. already make eCooking affordable for the ever-increasing 4. Help consumers understand the benefits of adopting number of grid-connected households. There is also an modern eCooking solutions, and reduce barriers to emerging opportunity with battery-supported eCooking. Cost behavioral change. trends suggest that the price of components will continue to fall while the cost of biomass fuels continues to rise. Battery support would also extend the opportunity to cook with elec- The MECS program is supporting strategic interventions in tricity to off-grid households and households with unreliable each of the five case study contexts featured in this report grid access. (as well as many more). Over the next decade, the relative price points of key technologies will continue to evolve, likely Realizing this potential will take concerted global effort. The opening the door to an even broader range of cost-effective report therefore concludes with a call for action, highlighting eCooking solutions. The program intends to keep close track how support for eCooking should be delivered to achieve of these developments, create a range of market-ready inno- the greatest development impact for the nearly 3 billion vations, and shape enabling environments in order to make people currently cooking with biomass: a valuable contribution toward achievement of SDG7. Conclusion 111 ENDNOTES 1. The MTF redefines the way energy access is measured, going beyond the traditional binary measure of “connected or not connected” and allowing for a more nuanced tracking of SDG7 targets. For more information, see ESMAP (2020b). 2. Access for households is defined as meeting Tier 4 standards or above (following ISO/TR 19867- 3:2018 Voluntary Performance Targets) across all six measurement attributes of the MTF: convenience, (fuel) availability (a proxy for reliability), safety, affordability, efficiency, and exposure (a proxy for health related to exposure to pollutants from cooking activities). 3. The figure is 0 in five regions, 10 percent in East Asia including China, 26 percent in South Asia including India, and 37 percent in southern Sub-Saharan Africa (Smith et al. 2014). 4. The scale and severity of the environmental impacts of wood-based biomass fuel use vary greatly over space and time. There is general agreement that collecting fuelwood has not led to major deforestation, although it can lead to local landscape degradation and alterations, sometimes causing local fuelwood shortages. Analysis is much more complex and divergent for charcoal, which is nearly exclusively consumed for cooking and heating in urban settlements, including for industrial and commercial uses. 5. Levels of electricity access are as defined by the MTF for the case study countries (except Tanzania, data for which are not yet available)]. 6. The Africa Renewable Energy Initiative (AREI) is an Africa-led initiative that aims to accelerate and scale up the harnessing of the continent’s renewable energy potential. 7. These devices were previously referred to as PV-eCook and Grid-eCook/Battery-eCook, respectively, although both contain a battery. 8. Levelized cost is a measure of net present cost averaged over some period or some output quantity. The levelized monthly cost of cooking is the net present value of initial capital investment, any required replacement capital investments, and recurrent electricity and fuel purchases throughout a specified financing period, averaged as cost per month. It is directly comparable with the costs of traditional cooking fuel for a household. 9. “Heavy foods” refer to foods such as beans that require significant energy, cost and time to cook, and which are often particularly amenable to cooking in an EPC. 10. A 9.6 percent real discount rate is used throughout this report, following Lombardi et al. (2019). Reported interest rates are frequently nominal rates, taking no account of inflation. For a country with average inflation of 10 percent (typical in parts of East Africa, for example), a 9.6 percent real rate equates to a 19.6 percent nominal rate. 11. Techniques include centralized or decentralized battery storage, smart metering, distributed-load control, and collaborative agreements. 12. Ethanol gel was found to be the most expensive fuel. LPG was approximately 15 percent, paraffin 35 percent, and electricity 70 percent cheaper. 13. Of course, many customers exceed the lifeline threshold and are paying the regular retail tariff of $0.23/kWh, suggesting that average consumption is actually lower, as such customers make up a larger share of average spending. 14. Figures from the SDG7 Global Tracking Framework differ from official government statistics, but they also report a comparably sized increase (from 29 percent to 73 percent) over a similar period (five years) (Kenya Power 2018). 15. The Last Mile Connectivity Project (LMCP) was launched in 2015 to scale up connectivity in rural and peri-urban areas by providing subsidies for grid extension to enable customers purchase electricity at an affordable cost. 16. LPG prices have a wider uncertainty than grid tariffs, so the green bar in Figure 3.8 is wider, but the midpoints are aligned. 17. The proposed restructured ZESCO tariffs are as follows: K 0.56Z (($0.04)/kWh for less than 100kWh/month, K 1.01 ($0.08)/kWh for 100–300kWh/month, and K 2.31 ($0.18)/kWh for more than 300kWh/month. 18. LCoE is the discounted cost of producing each unit of electricity (the minimum tariff that will enable the mini grid developer to break even). 19. Tests for charcoal and LPG were originally carried out with highly efficient cooking practices. Fuel consumption was scaled up by one-third to model everyday cooking conditions. 112 ESMAP  |  Cooking with Electricity: A Cost Perspective 20. The battery lifetimes chosen for the lower- and higher-cost system assumptions are stated in Section 2.1. For a discussion of battery lifetime modelling, see appendix E. 21. For example, the Health and Energy Platform for Action (HEPA) is a recent initiative co-led by the World Health Organization (WHO), the United Nations Development Programme (UNDP), the United Nations Department of Economic and Social Affairs (UNDESA), and the World Bank that links energy (including clean cooking) to health provision. It calls for political and technical cooperation between the health and energy sectors at both the global and country levels, in order to recognize the health burden of cooking with polluting fuels and technologies and use energy investment to make progress toward SDG3 (ensuring healthy lives and promoting well-being for all at all ages). Endnotes 113 REFERENCES Adam Smith International. 2016. “Black Gold—the Real Cost of Charcoal in Africa.” https://medium.com/@ adamsmithinternational92/black-gold-the-real-cost-of-charcoal-in-africa-7d241a2f3084. AFREA (Africa Renewable Energy and Access Program) and ESMAP ( Energy Sector Management Assistance Programme). 2016. Power Affordable for Africa and Viable for Its Utilities. World Bank (AFREA and ESMAP). Washington, DC. Akena, A.K., and E. Wanless. 2020. Achievements and Challenges of Uganda’s Power Sector. Rocky Mountain Institute. Basalt, CO. https://rmi.org/achievements-and-challenges-of-ugandas-power-sector/. Banerjee, M., Prasad, R., Rehman, I.H. and Gill, B. 2016. “Induction Stoves as an Option for Clean Cooking in Rural India.” Energy Policy. https://doi.org/10.1016/j.enpol.2015.10.021. Batchelor, S. 2013. “Is It Time for Solar Electric Cooking for Africa?” Gamos Concept Note, May, Reading, United Kingdom. Batchelor, S., Brown, E., Scott, N. and Leary, J. 2019. “Two Birds, One Stone: Reframing Cooking Energy Policies in Africa and Asia.” Energies 12 (9): 1591. https://doi.org/10.3390/en12091591. ———. 2015. Solar Electric Cooking in Africa in 2020: A Synthesis of the Possibilities. UK Department for International Development, London. https://doi.org/10.12774/eod_cr.december2015.batchelors. Batchelor, S., Leary, J., Sago, S., Minja, A., Chepkurui, K., Sawe, E., Shuma, J., and Scott,N. 2019. “Opportunities & Challenges for eCook Tanzania.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Batchelor, S., M. Talukder, M. Uddin, S. Mondal, S. Islam, R. Redoy, R. Hanlin, et al. 2018. “Solar E-Cooking: A Proposition for Solar Home System Integrated Clean Cooking.” Energies 11 (11): 2933. https://doi. org/10.3390/en11112933 . Battery University. 2019. Lead-Acid Batteries. https://batteryuniversity.com/learn/article/ lead_based_batteries. Bekker, Bernard, Anton Eberhard, Trevor Gaunt, and Andrew Marquard. 2008. “South Africa’s Rapid Electrification Programme: Policy, Institutional, Planning, Financing and Technical Innovations.” Energy Policy 36 (8): 3125–37. https://doi.org/10.1016/j.enpol.2008.04.014. Beute, Nico. 2012. “Electrification in South Africa over the Last 20 Years.” In Proceedings of the 20th Domestic Use of Energy (DUE) Conference, 55–63. Cape Town. Buskirk, Robert Van. 2019. Can Efficiency Be Sufficient for African Cooking? European Council for Energy Efficient Economy Summer Study. Chepkemoi, M., J. Leary, V. Chengo, and J.F. Todd. 2019. East Africa Launch Workshop Report. Modern Energy Cooking Services (MECS). www.MECS.org.uk. Chepkurui, K., J. Leary, A. Minja, S. Sago, S. Batchelor, E. Sawe, E. Brown, M. Leach, N. Scott, and J. Shuma. 2019. ECook and Gender in Tanzania. MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Chepkurui, K., J. Leary, A. Numi, J. Fodio-Todd, R. Hanlin, S. Batchelor, E. Brown, and N. Scott. 2019. Nairobi eCookBook Focus Group Discussion. MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Couture, T., and Jacobs, D. 2019. Beyond Fire: How to Achieve Electric Cooking. Hivos & World Future Council. https://www.worldfuturecouncil.org/beyond-fire/. Cowan, B. 2008. Identification and Demonstration of Selected Energy Best Practices for Low-Income Urban Communities in South Africa. Alleviation of Poverty through the Provision of Local Energy Services (APPLES), Energy Research Centre, University of Cape Town. CSIS (Center for Strategic and International Studies). 2018. Urbanization in Sub Saharan Africa. Washington, DC. https://www.csis.org/analysis/urbanization-sub-saharan-africa. Das, Ipsita, Joseph Pedit, Sudhanshu Handa, and Pamela Jagger. 2018. “Household Air Pollution (HAP), Microenvironment and Child Health: Strategies for Mitigating HAP Exposure in Urban Rwanda.” Environmental Research Letters 13 (045011). Dlamini, C., Moombe, K.B., Syampungani, S., Samboko, P.C., 2016. Load Shedding and Charcoal Use in Zambia: What Are the Implications on Forest Resources? Indaba Agricultural Policy Research Institute (IAPRI), Lusaka, Zambia. Drigo, R., Bailis, R., Ghilardi, A., Masera, O., and Suber, M. 2014. Tier I Final Report: Pan-tropical Analysis of Woodfuel Supply, Demand and Sustainability. Yale-UNAM Project, Global Alliance for Clean Cookstoves (GACC). 114 ESMAP  |  Cooking with Electricity: A Cost Perspective Eberhard, A., Gratwick, K. and Kariuki, L. 2018. “A Review of Private Investment in Tanzania’s Power Generation Sector.” Journal of Energy in Southern Africa 29 (2): 1–11. https://doi. org/10.17159/2413-3051/2018/v29i2a4389. EcoDirect. Solar Panels for Your Home and Energy Efficient Products 2020. https://www.ecodirect.com/ Default.asp. EEG (Energy and Economic Growth). 2016. “Tanzania Energy Sector Overview.” Energy for Economic Growth Research Programme. https://energyeconomicgrowth.org/index.php/publication/ tanzania-energy-sector-overview. ESMAP (Energy Sector Management Assistance Program). 2015. Beyond Connections: Energy Access Redefined. Washington, DC. https://openknowledge.worldbank.org/bitstream/handle/10986/24368/ Beyond0connect0d000technical0report.pdf?sequence=1&isAllowed=y. ———. 2019a. Mini Grids for Half a Billion People: Market Outlook & Handbook for Decision Makers. Executive Summary. Washington, DC. https://openknowledge.worldbank.org/bitstream/ handle/10986/31926/Mini Grids-for-Half-a-Billion-People-Market-Outlook-and-Handbook-for-Decision- Makers-Executive-Summary.pdf?sequence=1&isAllowed=y. ——— 2019b. “UK Commits $250 Million; France, Netherlands Pledge $120 Million to World Bank Initiatives on Energy Storage, Solar Power and Clean Cooking.” ESMAP News. Washington, DC. https://www.esmap.org/node/187647. ——— 2019c. “ESMAP Leads New Clean Cooking Program.” ESMAP News. Washington, DC. https://www.esmap.org/node/181461. ———. 2020a. Tracking SDG7|Progress Towards Sustainable Energy. Washington, DC. https://trackingsdg7.esmap.org//. ———. 2020b. Multi-Tier Framework|Tracking Progress Toward Sustainable Energy Goals. https://mtfenergyaccess.esmap.org/. ESMAP (Energy Sector Management Assistance Program), and MECS (Modern Energy Cooking Services). Forthcoming. The State of Access to Modern Energy Cooking Services. Washington, DC: World Bank. EWURA (Energy and Water Utilities Regulatory Authority). 2017. Annual Report for the Year Ended 30th June 2016, vol. 255. Dar es Salaam. Export.gov. 2019. “Tanzania: Energy.” 2019. https://www.export.gov/article?id=Tanzania-Energy. Falcão, M. P. 2008. “Charcoal Production and Use in Mozambique, Malawi, Tanzania, and Zambia: Historical Overview, Present Situation and Outlook.” Paper presented at the conference on Charcoal and Communities in Africa, June,16–18. Maputo, Mozambique. https://energypedia.info/images/b/ bb/EN-”Charcoal_production_and_use_in_Mozambique,_MalawEN_i,_Tanzania,_and_Zambia_-_ historical_overview,_present_situation_and_outlENook”-Falcão,_M._P..pdf. Frith, J. 2017. The 2017 Lithium-Ion Battery Price Survey. Bloomberg New Energy Finance. Gamos. 2017. Beyond Fire. https://elstove.com/2017/07/28/beyond-fire/. Gammon, R., Boait, P. and Advani, V. 2016. “Management of Demand Profiles on Mini Grids in Developing Countries Using Timeslot Allocation.” Paper presented at the Institute of Electrical and Electronics Engineers (IEEE) PowerAfrica Conference, June 28–July 2, Livingstone, Zambia. Gius, G., Walker, M., Li, A., Adams, N.J., Van Buskirk, R. and Schwartz, P. 2019. “Hot Diodes! Dirt Cheap Cooking and Electricity for the Global Poor?” Development Engineering 4 (October): 100044. https:// doi.org/10.1016/j.deveng.2019.100044. Global LEAP. 2016. The State of the Off-Grid Appliance Market. https://dalberg.com/wp-content/ uploads/2016/06/Global_LEAP_The_State_of_the_Global_Off-Grid_Appliance_Market_1.pdf. Global LEAP (Lighting and Energy Access Partnership). 2020. “2020 Global LEAP Awards Electric Pressure Cooker Competition.” https://globalleapawards.org/electric-pressure-cookers. GOGLA (Global Off-Grid Lighting Association). 2019. Global Off-Grid Solar Market Report Semi-Annual Sales and Impact Data, January–June 2019, Public Report. Utrecht, the Netherlands. Goldie-Scot, L. 2019. “A Behind the Scenes Take on Lithium-Ion Battery Prices.” Bloomberg New Energy Finance, March 5. https://about.bnef.com/blog/behind-scenes-take-lithium-ion-battery-prices/. Gould, C.F.., Schlesinger, S., Toasa, A.O., Thurber, M., Waters, W.F., Graham, J.P. and Jack, D.W. 2018. “Government Policy, Clean Fuel Access, and Persistent Fuel Stacking in Ecuador.” Energy for Sustainable Development 46: 111–22. https://doi.org/10.1016/j.esd.2018.05.009. Howells, M., Victor, D.G., Gaunt, T., Elias, R.J. and Alfstad, T. 2006. “Beyond Free Electricity: The Costs of Electric Cooking in Poor Households and a Market-Friendly Alternative.” Energy Policy 34 (17): 3351–58. https://doi.org/10.1016/j.enpol.2005.07.006. IEA (International Energy Agency). 2018. SDG7: Data and Projections. Access to Electricity. https://www.iea.org/reports/sdg7-data-and-projections/access-to-electricity. IEA (International Energy Agency), IRENA (International Renewable Energy Association), UNSD (United Nations Statistics Division), WB (World Bank), and WHO (World Health Organization). 2019. Tracking SDG 7: The Energy Progress Report 2019. Washington DC. https://trackingsdg7.esmap.org/. IMARC (Information Management Research Centre). 2019. Multi Cooker Market: Global Industry Trends, Share, Size, Growth, Opportunity and Forecast 2019–2024. IRENA (International Renewable Energy Agency). 2012. “Solar Photovoltaics. International Renewable Energy Agency.” Renewable Energy Technologies: Cost Analysis Series 1 (4/5). Abu Dhabi. References 115 ———. 2016. Remap: Roadmap for a Renewable Energy Future. Abu Dhabi. ———. 2018. Renewable Power Generation Costs in 2017. Abu Dhabi. ———. 2019. Renewable Capacity Statistics 2019. Abu Dhabi. ———. 2020. “Scaling Up Renewable Energy Deployment in Africa.” Abu Dhabi. https://www.irena.org/-/ media/Files/IRENA/Agency/Publication/2020/Feb/IRENA_Africa_Impact_Report_2020.pdf. ITT (Institute for Transformative Technology). 2016. Achieving Universal Electrification in India. A Roadmap for Rural Solar Mini Grids. Rockefeller Foundation. https://www.rockefellerfoundation.org/wp-content/ uploads/Achieving-Universal-Electrification-in-India.ITTReport.April2016.pdf/. Jacobs, D., Schurig, D., Matser, E., Leidreiter, A. and García, I. 2016. Beyond Fire: How to Achieve Sustainable Cooking. World Future Council, Hamburg, Germany. https://www.worldfuturecouncil.org/ beyond-fire/. Kenya Power. 2018. Annual Report and Financial Statements for the Year Ended 30th June 2018. Nairobi. KPLC (Kenya Power and Lighting Company). 2019a. Implementation of New Tariffs. https://kplc.co.ke/img/ full/rZr6LLCjVqVw_1st November 2018 - Tarrifs.jpg. ———. 2019b. “Pika Na Power.” 2019. https://www.facebook.com/PikaNaPower/. Lall, S.V., Henderson, J.V. and Venables, A.J. 2017. “Africa’s Cities: Opening Doors to the World.” Nouvelles Dermatologiques, vol. 15. World Bank and UK Aid. Leach, M., J. Leary, N. Scott, S. Batchelor, X. Chen, K.-S. Ng, R. Oduro, and E. Brown. 2019. eCook Modelling. MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leach, M. and Oduro, R. 2015. “Preliminary Design and Analysis of a Proposed Solar and Battery Electric Cooking Concept : Costs and Pricing.” Evidence on Demand (prepared at the request of the UK Department for International Development). https://doi.org/10.12774/eod_cr.november2015.leachm. ———. 2019b. “Why Understanding Real Cooks Is Fundamental to Going beyond Fire.” MECS blog. https:// www.mecs.org.uk/why-understanding-real-cooks-is-fundamental-to-going-beyond-fire/. Leary, J., S. Batchelor, M. Leach, E. Brown, and A. Alsop. 2018. “eCook Global Market Assessment Where Will the Transition Take Place First ?” MECS Working Paper, Modern Energy Cooking Services. www. MECS.org.uk. Leary, J., S. Batchelor, and N. Scott. 2019. “Cooking Diaries 3.0 Protocols.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., S. Batchelor, N. Scott, E. Brown, and M. Leach. 2018. “Electric Pressure Cooker Selection Guide.” MECS Working Paper, Modern Energy Cooking Solutions. www.mecs.org.uk/wp-content/ uploads/2019/10/Pressure-cooker-selection-guide-v1.2-8-3-18.pdf. Leary, J., J. Fodio Todd, S. Batchelor, K. Chepkurui, M. Chepkemoi, A. Numi, R. Hanlin, N. Scott, and E. Brown. 2019. The Kenya eCookBook: Beans & Cereals Edition. MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., F. Mwila, N. Serenje, F. Yamba, S. Batchelor, E. Brown, M. Leach, and N. Scott. 2019. “eCook Zambia Design Challenge Workshop Report.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk Leary, J., N. Scott, W. W. Hlaing, A. Myint, S. Sane, P. P. Win, T. M. Phyu, et al. 2019. “eCook Myanmar Cooking Diaries.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., N. Scott, A. Numi, K. Chepkurui, R. Hanlin, M. Chepkemoi, S. Batchelor, et al.. 2019. “eCook Kenya Cooking Diaries.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., N. Scott, S. Sago, A. Minja, S. Batchelor, K. Chepkurui, E. Sawe, M. Leach, and E. Brown. 2019. “eCook Tanzania Cooking Diaries.” MECS Working Paper, Modern Energy Cooking Services. www. MECS.org.uk. Leary, J., N. Scott, N. Serenje, F. Mwila, S. Batchelor, M. Leach, E. Brown, and F. Yamba. 2019a. “eCook Zambia Cooking Diaries.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. ———. 2019b. “Opportunities & Challenges for ECook in Zambia.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., N. Serenje, F. Mwila, S. Batchelor, M. Leach, E. Brown, N. Scott, and F. Yamba. 2019. “eCook Zambia Focus Group Discussions Summary Report.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Leary, J., P. P. Win, A. Myint, S. Sane, P. P. Win, T. M. Phyu, E. T. Moe, et al. 2019. “eCook Myanmar Focus Group Discussions.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Lemaire, X. 2011. “Off-Grid Electrification with Solar Home Systems: The Experience of a Fee-for- Service Concession in South Africa.” Energy for Sustainable Development 15 (3): 277–83. https://doi.org/10.1016/j.esd.2011.07.005. Lighting Global. 2018. Off-Grid Solar Market Trends Report 2018. Washington, DC: World Bank Group. https://www.lightingglobal.org/resource/2018-global-off-grid-solar-market-trends-report/. Lombardi, F., Fabio, R., Marco, S. and Emanuela, C. 2019. “Enabling Combined Access to Electricity and Clean Cooking with PV-Microgrids: New Evidences from a High-Resolution Model of Cooking Loads.” Energy for Sustainable Development 49: 78–88. https://doi.org/10.1016/j.esd.2019.01.005. Musoke, R. 2017. “Charcoal Users Feel Price Heat.” Independent Uganda, October 16. https://www.independent.co.ug/charcoal-users-feel-price-heat/. 116 ESMAP  |  Cooking with Electricity: A Cost Perspective Odarno, L., Sawe, E., Swai, M., Katyega, M.J.J. and Lee, A. 2017. Accelerating Mini Grid Deployment in Sub-Saharan Africa: Lessons from Tanzania. World Resources Institute (WRI) and Tanzania Traditional Energy Organisation (TaTEDO), Dar es Salaam. Parikh, J., Cloke, J., Puzzolo, E., Pope, D., and Singh, C. 2020. Electric Cooking. MECS Briefing Paper, Modern Energy Cooking Services. www.MECS.org.uk. Power Africa. 2018. Power Africa in Uganda. Nairobi. https://www.usaid.gov/powerafrica/uganda. REAM (Renewable Energy Association of Myanmar), DRI (Department for Research and Innovation), Gamos, Loughborough University, and University of Surrey. 2018. eCook Myanmar Stakeholder Workshop Report. www.MECS.org.uk. Republic of South Africa. 2012. A Survey of Energy-Related Behaviour and Perceptions in South Africa. Department of Energy, Pretoria. Scott, N., J. Leary, W. W. Hlaing, A. Myint, S. Sane, P. P. Win, T. M. Phyu, E. T. Moe, S. Batchelor, et al.. 2019. “Opportunities & Challenges for eCook in Myanmar.” Working Paper.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Scott, N., J. Leary, W.W. Hlaing, A. Myint, S. Sane, P.P. Win, T.M. Phyu, E.T. Moe, T. Htay, et al. 2019. “eCook Myanmar Discrete Choice Modelling.” MECS Working Paper, Modern Energy Cooking Services. www. MECS.org.uk. Scott, N., J. Leary, S. Sago, A. Minja, S. Batchelor, K. Chepkurui, E. Sawe, M. Leach, and E. Brown. 2019. “eCook Tanzania Discrete Choice Modelling.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Scott, N., J. Leary, N. Serenje, F. Mwila, S. Batchelor, M. Leach, E. Brown, and F. Yamba. 2019. “eCook Zambia Discrete Choice Modelling.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Scott, N., S. Batchelor, and T. Jones. 2019. “Discrete Choice Modelling Survey: Kenya.” MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Seair Exim Solutions. 2020. Market Intelligence Data, Kenya. https://www.seair.co.in/. Sebitosi, A. B., and Pillay, P. 2005. “Energy Services in Sub-Saharan Africa: How Conducive Is the Environment?” Energy Policy 33 (16): 2044–51. https://doi.org/10.1016/j.enpol.2004.03.025. Skat. 2019. Mini Grid Technologies. https://energypedia.info/images/a/ad/Mini-grid_Technology_ Comparison.pdf. Smith, K.R., Bruce, N., Balakrishnan, K., Adair-Rohani, H., Balmes, J., Chafe, Z., Dherani, M., Hosgood, H.D., Mehta, S., Pope, D. and Rehfuess, E.. 2014. “Millions Dead: How Do We Know and What Does It Mean? Methods Used in the Comparative Risk Assessment of Household Air Pollution.” American Review of Public Health 35: 185–206. Šúri, M., Huld, T. A. and Dunlop, E. D.. 2005. “PVGIS: A Web-Based Solar Radiation Database for the Calculation of PV Potential in Europe.” International Journal of Sustainable Energy 24 (2): 55-67. TANESCO (Tanzania Electricity Supply Company Ltd.). 2019. Bei Za Umeme Zilizoidhinishwa. http://www. tanesco.co.tz/index.php/customer-service/tariffs/7-bei-za-umeme-zilizoidhinishwa/file. Tesga Power. 2019. Dc 12v 24v Stainless Steel Electric Pressure Cooker 5l. https://www.alibaba.com/ product-detail/DC-12V-24V-stainless-steel-electric_60760544437.html?spm=a2700.7724857. normalList.26.24e5228dC893aj. UBOS (Uganda Bureau of Statistics) 2017. Uganda National Household Survey 2016/17. Kampala. https:// www.ubos.org/wp-content/uploads/publications/03_20182016_UNHS_FINAL_REPORT.pdf. Umeme. 2019. Power Tariffs. 2019. https://www.umeme.co.ug/tariffs. UN (United Nations). 2017. “World Population Prospectus: The 2017 Revision, Key Findings and Advance Tables.” Working Paper ESA/P/WP/248. New York. https://esa.un.org/unpd/wpp/publications/Files/ WPP2017_KeyFindings.pdf. Villema, N., S. Sago, E. Sawe, A. Minja, J. Leary, K. Chepkurui, J. Shuma, E. Brown, S. Batchelor, and M. Leach. 2018. The National Stakeholders’ Solar Electric Cooking Workshop Report. MECS Working Paper, Modern Energy Cooking Services. www.MECS.org.uk. Waldron, D., and Hacker, S. 2020. Electric Bankers: Utility-Enabled Finance in Sub-Saharan Africa. CGAP (Consultative Group to Assist the Poor) and CLASP (Collaborative Labelling and Appliance Standards Programme). https://www.cgap.org/sites/default/files/publications/2020_05_Working_Paper_Electric_ Bankers.pdf. Wang J., P. Liua, J. Hicks-Garner, E. Sherman, S. Soukiazian, M. Verbrugge, H. Tataria, J. Musser, and T. Finamore. 2011. “Cycle-Life Model for Graphite-LiFePO4 Cells.” Journal of Power Sources 196. doi:10.1016/j.jpowsour.2010.11.134. Watkins, T., P. Arroyo, R. Perry, R. Wang, O. Arriaga, M. Fleming, C. O’Day, et al. 2017. “Insulated Solar Electric Cooking: Tomorrow’s Healthy Affordable Stoves?” Development Engineering 2 (January): 47–52. https://doi.org/10.1016/j.deveng.2017.01.001. WHO (World Health Organization). 2016. Burning Opportunity: Clean Household Energy for Health, Sustainable Development, and Wellbeing of Women and Children. Geneva. http://apps.who.int/iris/bitst ream/10665/204717/1/9789241565233_eng.pdf?ua=1. References 117 ———. 2017. Household Energy Database. Geneva. https://www.who.int/airpollution/data/household- energy-database/en/#:~:text=The%20WHO%20Household%20energy%20database,cooking%2C%20 heating%20and%20lighting%20fuels. ———. 2018. “Household Air Pollution and Health.” Geneva. https://www.who.int/en/news-room/fact-sheets/ detail/household-air-pollution-and-health. World Bank. 2013. On Thin Ice. Washington, DC. https://www.worldbank.org/content/dam/Worldbank/ document/SDN/Full_Report_On_Thin_Ice_How_Cutting_Pollution_Can_Slow_Warming_and_Save_ Lives.pdf. ———. 2014. Clean and Improved Cooking in Sub-Saharan Africa: A Landscape Report. Africa Clean Cooking Energy Solutions Initiative, Washington, DC. ———. 2018. Off-Grid Solar Market Trends Report 2018. Washington, DC. ———. 2019a. Getting Electricity: Doing Business. Washington, DC. https://www.doingbusiness.org/en/data/ exploretopics/getting-electricity. ———. 2019b. World Development Indicators. Washington, DC. https://datacatalog.worldbank.org/. ———. 2019c. World Development Indicators (WDI) Data Catalog. Washington, DC. https://datacatalog. worldbank.org/dataset/world-development-indicators. World Bank. 2020. Data. https://data.worldbank.org/indicator/SP.URB.GROW?locations=ZG World Bank, and IHME (Institute for Health Metrics and Evaluation). 2016. The Cost of Air Pollution. Washington, DC. http://documents.worldbank.org/curated/en/781521473177013155/pdf/108141- REVISED-Cost-of-PollutionWebCORRECTEDfile.pdf. Wu, X., R. Nethery, R., M.B. Sabath, D. Braun, F. Dominici. 2017. “Exposure to Air Pollution and COVID-19 Mortality in the United States: A Nationwide Cross-Sectional Study.” Journal of Chemical Information and Modeling. DOI: 10.1017/CBO9781107415324.004. ZESCO (Zambia Electricity Supply Corporation Ltd.). 2019. Current ZESCO Tariffs. https://www.zesco.co.zm/ home. Zubi, G., F. Spertino, M. Carvalho, R. S. Adhikari, and T. Khatib. 2017. “Development and Assessment of a Solar Home System to Cover Cooking and Lighting Needs in Developing Regions as a Better Alternative for Existing Practices.” Solar Energy 155 (October): 7–17. https://doi.org/10.1016/j. solener.2017.05.077. 118 ESMAP  |  Cooking with Electricity: A Cost Perspective APPENDIX A THE MODERN ENERGY COOKING SOLUTIONS PROGRAM The Modern Energy Cooking Services (MECS) program was socio-technical innovations that will drive the transition set up to pave the way for the development and dissemination forward. MECS will also identify and generate evidence on of innovative eCooking solutions. It is a five-year program that other drivers for transition, including (a) understanding and combines creating a stronger evidence base for transitions to optimizing multi-fuel use (fuel stacking), cooking demand, modern energy cooking services in Foreign, Commonwealth and behavior change and (b) establishing the evidence base and Development Office (FCDO) priority countries with to support enabling policy environments that can underpin FIGURE A.1 Overview of the Modern Energy Cooking Services (MECS) program Deliver and communicate high Actors implement Enhance quality research to key high quality evidence integration of clean stakeholders based interventions cooking in SDG7 Transitional pathways Assumption national goverment development plans • Conceptual framework for understanding capture the benefits of MECS. current price trends in polluting demand and supply for clean cooking fuels and technologies continue & MECS technologies reach a ordability tipping points. Value chains for MECS supported by private sector. Technology and business innovation • Traials of tech. and business model prototypes Assumption • Life cycle analysis Goverments, • Cultural studies and consumer feedback guide design donors and lending institutions continue to invest in programmes and Scaled up experimentation Participating countries initiatives that • Transitional theory of changes incorporate promote access • Standards and specifications for technologies MECS in energy to clean energy • Develop and test operational models for scalling policy, planning private sector and programmes responds at scale to MECS Market ready opportunities. range of MECS innovations SDG 7 (technology and • Incorporate results into access to MECSSDG indicator 7 business models) Impact: Accelerated Changing the narrative uptake of • Pushing results into practice MECS in • Bringingtogether the clean cooking and electrification sectors Africa and Asia Modern energy cooking services (MECS) is a 5 year programme funded by UK aid. it is managed by loughborough university UK in partnership with ESMAP (World ban) Appendix A: The Modern Energy Cooking Solutions Program 119 a pathway to scale and support well-understood markets to the MECS program. The countries of interest will be and enterprises. The program is managed as an integrated reassessed every six months throughout the duration of the whole but split into two complementary workstreams, one MECS program, based on the following criteria: lead by Loughborough University and the other led by the World Bank’s Energy Sector Management Assistance Program ● The main source of fuel for cooking is biomass, and the (ESMAP). Figure A.1 outlines the program. government is seeking to do something different. ● Access to modern energy is poor, as a result of weak The intended outcome of MECS is a market-ready range of supply chains and key infrastructure, but the govern- innovations (both technology and business models) that lead ment wants to improve the situation. to improved choice of sustainable, affordable, and reliable modern energy cooking services for consumers. MECS ● The country is a FCDO priority country, and a large includes a series of challenge funds, designed to facilitate share of the population is poor. feasibility studies, prototyping, piloting, and scaling up of ● A substantial resource of renewable energy exists but is innovative eCooking solutions, as exemplified by many of the barely tapped. case studies in this report. MECS principles will be integrated into the SDG 7.1 global tracking framework, with the aim of Tier 1 countries fulfill the criteria and have existing connec- encouraging participating countries to incorporate modern tions and activities with MECS. They include Bangladesh, energy cooking services in energy policies and planning. Ethiopia, Ghana, Kenya, Malawi, Nepal, Rwanda, Tanzania, Uganda, and Zambia. Tier 2 countries also seem to fulfill the The MECS program focuses on 15 countries in the Global criteria, but MECS has so far had more limited connections South. They are divided into Tier 1 and Tier 2 categories, with them. They include Cambodia, Cameroon, The Gambia, depending on the strength of their connection and relevance Myanmar, and Nigeria. 120 ESMAP  |  Cooking with Electricity: A Cost Perspective APPENDIX B TYPOLOGY OF ECOOKING SYSTEM ARCHITECTURES FIGURE B.1 Typology of eCooking devices for strong, weak, and off-grid regions Direct or Battery-less eCooking Battery-Supported eCooking Direct AC Direct DC AC eCooking eCooking battery-supported eCooking + + DC Battery-Supported eCooking + Grid-Connected eCooking Grid-Connected Battery-Supported eCooking National Grid Direct AC eCooking National-Grid Battery-Supported eCooking (DC) or AC + + + + + Mini-Grid Direct AC eCooking Mini-Grid Battery-Supported eCooking (DC) or AC + + + + + Solar eCooking Solar Powered Battery Supported eCooking Direct Solar eCooking (DC) Solar Power Battery-Supported eCooking (DC) or AC + + + + Note: eCooking = Electric Cooking The system architectures modelled in this report are listed in table 2.4. Appendix B: Typology of eCooking System Architectures 12 1 APPENDIX C ASSESSING ELECTRICITY DEMAND FOR COOKING Cooking diaries are a novel methodology for addressing the Cooking with Electricity lack of data about how people currently cook with biomass and how they might cook with electricity. Cooking is a deeply culturally embedded practice. Understanding the nuances of A barrier to cooking with electricity is the high level of power how the intended beneficiaries of a clean cooking interven- required, which can be an issue in terms of both the quality tion actually cook is therefore critical. of the connection to an individual house and the aggregate loads the additional power may impose on a distribution Data on cooking practices, fuel/electricity use, and the network. Boiling and simmering can easily be done on lower- user experience were collected in each of the four case power insulated devices. Higher power is needed for other study countries. Focus groups offered deeper qualitative processes, however, such as frying. insights into how people currently cook, how they aspire to cook in the future, and the compatibility of their cooking Many people think that eCooking appliances consume their practices with the strengths and weaknesses of cooking rated power constantly. In fact, most appliances are automati- on battery-supported electrical appliances. The results cally controlled to oscillate between no power and full power show that unlike many other clean cooking ­ technologies, (figure C.1), depending on user input or the temperature in which have struggled to achieve acceptance, many the pot. As a result, they rarely consume their maximum rated ­ energy-efficient eCooking appliances are highly desirable capacity, even when cooking on high, so even a 1kW hot plate to everyday cooks. is unlikely to use a full 1kWh if left on for an hour. FIGURE C.1 Electricity demand profile of a 600W electric pressure cooker cooking for one hour 1,000 800 600 Watt 400 200 0 1 201 401 601 801 1,001 1,201 1,401 1,601 1,801 2,000 2,201 2,401 2,601 2,801 3,001 3,201 3,401 3,601 Seconds Source: Couture and Jacobs (2019). 122 ESMAP  |  Cooking with Electricity: A Cost Perspective The amount of electricity required for cooking depends on the savings and comparisons are particularly sensitive to the following factors: what is cooked. 1. the efficiency of heat transfer into the pot (for example, Cowan (2008) studied energy use by 80 households in induction) or (better) directly into the food (as in a South Africa cooking a wide range of foods and meal types microwave) using a number of fuels, including electricity (figure C.2). The study subdivided meals into quick (for example, rice); 2. control of the cooking process (through, for example, a medium (for example, chicken stew); and long (for e ­ xample, timer on a microwave or a temperature sensor on a rice offal) categories. A typical meal of rice and chicken stew for cooker) four people used 0.71kWh. Three meals a day with a mix of 3. the efficiency of heat transfer out of the pot (which is meal types could thus be delivered for perhaps 2kWh/day. reduced by lids and insulation) However, Cowan used a cheap and inefficient ­ commercial hot plate. Recent research has shown that an EPC can yield 4. the temperature in the pot significant savings over such a device and ­ significantly 5. energy-efficient cooking practices (such as soaking reduce the energy required (Leary, Fodio Todd, et al. 2019). beans as chopping ingredients finely). Building on these insights into meal-based energy consump- The focus of the clean cooking industry has been on the tion, the ground-breaking Beyond Fire series of reports first factor, often using the efficiency of heat transfer from investigated the viability of four potential pathways to the fuel into the pot as the key performance indicator for achieve truly sustainable cooking: solar eCooking, eCook- improved cookstoves. Many people claim that induction ing on mini grids, biogas, and renewably generated “power stoves increase the “efficiency of cooking” by 10–20 percent to gas” (Jacobs et al. 2016; Couture and Jacobs 2019). The over hot plates. This claim is based on the first factor only. researchers concluded that of all the pathways, the two Induction stoves can be used in tandem with other equip- eCooking configurations offered the greatest co-benefits, ment that address the third and fourth factor (insulation and although they were also the most expensive. pressurization) through the use of insulated and/or pressur- ized stove-top pots. However, in rice cookers and electric Key to understanding the viability of each pathway was pressure cookers (EPCs), insulation and pressurization (for establishing the amount of energy needed to cook. In their EPCs) are integrated into the appliance itself. Rice cookers first report, the authors used a figure of 1GJ (278kWh)/ and EPCs may not use induction to heat the pot, but their person/year in the pot, which is equivalent to 3.2 kWh/ strategic use of insulation means that there is minimal wast- household/day for an average household of 4.2 people.1 age in the heat transfer process; in many cases they mimic They focused solely on hot plates and induction stoves, the efficiencies of the induction hob and exceed it by also ignoring insulation, pressurization, automatic control and retaining heat with insultation. The EPC also offers significant energy saving practices, which offer significant energy advantages over the combination of induction and stove- saving potential (Gamos 2017). top pressure pans in relation to the second factor, through the level of automatic control. The integrated appliance is In their second report (Couture and Jacobs 2019), which completely controlled to avoid excessive pressurization, focused exclusively on the two eCooking pathways, they yielding further energy savings, increasing safety, and expanded their analysis to include two more energy-efficient ­ reducing the need for monitoring of the cooking process eCooking appliances: the EPC and the slow cooker. The by the cook. energy demand modelled in their original report seemed to understate the opportunity for eCooking solutions. The Much of the research on the performance of improved findings in the second report are more positive. This report cooking appliances has used standardized water boiling also concludes that induction stoves provide limited savings tests, which are effective at measuring heat transfer and thus over hot plates and highlights the substantial energy savings losses and efficiency in a laboratory setting. However, the for the most efficient appliances (slow cookers and EPCs). amount of energy actually saved depends on the meal being Their electricity consumption figures are based on the cooked. The greater control offered by electricity means that 1 The authors state that the exact number is not necessary for their analysis, because they focus on monthly cooking costs rather than the initial capital cost. They therefore need only a baseline to compare the four technological pathways. Appendix C: Assessing Electricity Demand for Cooking 12 3 FIGURE C.2 Electricity required to prepare typical meals for four people in South Africa 12 3,0 10 2,5 Energy Consumption per Meal (in kWh) Energy Consumption per Meal (in MJ) 8 2,0 6 1,5 4 1,0 2 0,5 0 0 Staple Starch Boiled Rice Pasta Soft Meats Medium-Length Longer-Cooking Longer-Cooking Very Slow Meals (e.g., Liver) Stews Meat Stews Dried Vegetable Cooking Meat (e.g., Potatoes, (e.g., Chicken (e.g., Beef/ Dishes Cassava) and Vegetable Mutton Stew) (e.g., Beef/Maize Stew) and Beans) Electricity Parafin LPG Ethanol Gel Source: Adapted from Cowan (2008) by Couture and Jacobs (2019). extrapolation of laboratory-based measurements, however, 80 households in four countries as input data for a compa- with the lid left closed for the full duration of the cooking.2 rable techno-economic model. These data were collected using a range of multidisciplinary techniques, including cook- Fully understanding how much electricity is needed to ing diaries, focus groups, and kitchen laboratories, building cook underpins all eCooking cost comparisons. Couture on Cowan’s practical controlled cooking tests. The studies and Jacobs (2019, 8) therefore end their report with a call aimed to understand how households in Sub-Saharan Africa to “governments and donors around the world… to fund a and Southeast Asia currently cook and how they aspire greater range of R&D projects, including… providing further to cook. analysis of cooking with different electric appliances, such as slow cookers, pressure cookers and even infrared These techniques were applied in Kenya, Myanmar, cookers [and] analysis of [their] behavioral and cultural Tanzania, and Zambia.3 Under the MECS program, they acceptance.” will be applied in all 15 focus countries. As of June 2020, data collection was already underway in Ethiopia, Ghana, This report extends Beyond Fire’s analysis to take forward Nepal, and Uganda. The analysis described below focuses the collective understanding of how much it really costs to on the data from Kenya, the most detailed dataset currently cook with electricity by using empirical data recorded by available. 2 For discussion of the implications of the test methods, see Leary and Batchelor (2019b). 3 The project reports for each country are available at https://www.mecs.org.uk/working-papers/. 124 ESMAP  |  Cooking with Electricity: A Cost Perspective plug-in kWh meters [figure C.3]); registration surveys (simple The Cooking Diaries demographic data); and exit surveys (qualitative user expe- rience feedback and observational eCooking challenges). Studies In each country studied, 20 households recorded data in two stages. In the first stage, participants collected baseline data for two weeks. During this time, they cooked the way By gathering data on how people cook in their own homes, they always do. In the second “transition” stage, participants cooking diary studies provide insights into the unique cooked only with electric appliances for two weeks. cooking practices of individual households and quantita- tive measurements of the energy used in the home (Leary, Details on the methods used and the findings in each coun- Batchelor, and Scott 2019). It is usually easier to control heat try can be found in the synthesis and country reports for levels with modern fuels such as gas and electricity than it is Kenya, Myanmar, Tanzania, and Zambia (see table 2.1). The with biomass, as they can be turned up/down and on/off in following sections present some key results and learning an instant. There is also a wide range of eCooking appli- from the Kenyan study. ances, each designed for specific processes (for example, kettles for heating water). Therefore, it is important to know how often people need to fry, boil, reheat, or use other cook- ing methods. BENEFITS OF USING ELECTRIC PRESSURE COOKERS TO PREPARE “HEAVY FOODS” This mixed-methods approach gathered data from cook- ing diary forms (foods cooked, cooking processes/times, Almost all 19 households that participated in the Githeri appliances used); energy measurements (weighing fuels and eCooking Challenge at the end of the Kenya cooking diaries FIGURE C.3 Enumerator training study participant to record cooking diary data in Nairobi Note: eCooking appliances are plugged into an energy meter in the top right of the photo. Appendix C: Assessing Electricity Demand for Cooking 12 5 Energy consumption during the Githeri eCooking Challenge, by participant, appliance, FIGURE C.4  and process 0.9 0.8 0.7 0.6 Energy (kWh) 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Household ID EPC First pressurization EPC Second pressurization EPC fry Hotpate fry study achieved dramatic energy savings cooking “heavy the biggest difference to the time and money spent in the foods” in a kitchen laboratory setting. On a hot plate, cook- kitchen. They show that EPCs can save up to 85 percent of ing 500 grams of githeri (a traditional Kenyan meal of maize the cost of cooking heavy foods using charcoal, LPG, or an and beans) usually consumes more than 2kWh and can electric hot plate (figure C.5).4 consume as much as 4kWh if no efficiency measures are in place (using the slowest-cooking beans, leaving the lid off, In the Kenya cooking diaries, households were provided and so forth). Using an EPC, 16 of 19 households prepared with three key devices: a hot plate, a rice cooker, and the meal using less than 0.4kWh—an 80 percent saving over an EPC and were thus able to cook all their food with hot plates (figure C.4). One participant managed to beat the electricity. The menu did not vary significantly from the figure the Beyond Fire report cites as low consumption for baseline data obtained during the preceding weeks with one hour of cooking on an EPC (0.164kWh); many others participants’ stoves and fuels. The analysis below shows consumed around the average level (0.221kWh). that it is possible to cook over 90 percent of this typical Kenyan menu in an EPC. After limited training, with three Four of the participants in the Kenya cooking diaries study appliances to choose from, participants chose to cook were featured in the Kenya eCookBook (Leary, Fodio approximately half their menu using efficient appliances Todd, et al. 2019). Kitchen laboratory experiments involve (EPC or rice cooker). When the did, they used about half controlled experiments to explore which factors make the energy of a hot plate. 4 Leary et al. did not test the use of stove-top pressure cookers on charcoal, kerosene, LPG, hot plates, or stove-top pressure cookers in combination with a fireless cooker. The difference between nonelectric fuels and an EPC would presumably be smaller if the nonelectric fuels used a conventional stove-top pressure cooker. 126 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE C.5 Cost of preparing 500 grams of dried yellow beans using most popular fuels in urban Kenya 40 30 KSh 20 10 0 Note: Costs are based on prices in Nairobi in July 2018. Source: Leary, Fodio Todd, et al. (2019). GOING BEYOND HEAVY FOODS ● Heavy foods: Heavy foods usually require boiling the main ingredient (for example, beans) for over an hour on Energy savings on heavy foods are substantial in controlled a conventional stove. Their preparation may also involve and semi-controlled conditions. It is important to understand a separate frying stage with extra ingredients to add how they fit into the kitchen routines of cooks at home. flavor (for example, a tomato and onion sauce). ● Staples: Staples are normally boiled for about half an Evidence from the cooking diaries shows that heavy foods hour. Some (for example, ugali, porridge) require stirring; comprise about a third of all dishes on a typical urban others (for example, rice) are simply left to boil. Kenyan household’s menu (table C.1). Many other dishes can also be cooked on an EPC. Some (such as rice) are quickly ● Quick fryers: Food is usually fried for 5–15 minutes. A grasped and require little in the way of behavior change. shallow pan and high heat are often preferred but are Others (such as using a heatproof material to hold the pot not essential. Access to the pan is usually required to still while stirring ugali, a maize flour porridge), are less intu- stir the food and prevent burning. itive and require some behavior change. A few dishes (such ● Deep fryers: Food is completely submerged in oil at as chapati) cannot be prepared using the EPCs available on 175°C–190°C. the market today. ● Flatbreads: Medium heat, evenly distributed across a shallow pan is required to cook flatbread at the same A typical East/Southern African menu consists of various rate. Access to the pan is required to turn the bread categories of dishes. Leary, Scott, Numi, et al. (2019) propose frequently. the following categories: Appendix C: Assessing Electricity Demand for Cooking 127 Categorization of typical Kenyan foods by compatibility with electric pressure cooker and TABLE C.1  associated energy savings FREQUENCY OF HOME COOKING IN ENERGY URBAN KENYAN MENU COMPATIBILITY SAVINGS VERSUS REQUIREMENTS FOOD CATEGORY (PERCENT OF TOTAL) TYPICAL DISHES WITH EPC HOTPLATE AND ENABLERS Heavy foods 32 Beans, matumbo Users High (50–90 Provide cooks (tripe), meat instinctively use percent) with cooking stews EPCs times and water quantities for popular local foods Staples 39 Ugali (maize Users use EPCs if Moderate (20–50 Demonstrations, meal), rice encouraged percent) extra EPC Quick fry 20 Sukuma wiki Users use EPCs if Low (5–20 Demonstrations, (kale), eggs encouraged percent) manual heat control, extra EPC, shallow pan Deep fry 2 Mandazi (donuts), Users cannot Low (5–20 Manual heat fried chicken, currently use percent) control or deep chips EPCs fry settings (160°C–190°C) Flatbreads 4 Chapati Users cannot Low (5–20 Manual heat (flatbread) currently use percent) control and EPCs shallow pan Other 3 Unknown The Kenya cooking diaries data suggest that EPCs use appliance available. A total of 645 dishes were cooked on roughly half the energy of electric hot plates across the EPCs and rice cookers (see figure C.6). Ignoring all other full range of dishes they are able to cook. On average, rice appliances (which were used for only 150 dishes and were cookers used 39 percent (median of 0.09 kWh/person/event, mainly microwaves) and comparing directly to the 739 dishes n = 46) and EPCs used 76 percent (0.18 MJ/person/event, cooked on a hot plate, roughly half (47 percent) of a total n = 49) of the energy of a hot plate (0.23 MJ/person/event, of 1,387 dishes were cooked by choice on an EPC or rice n = 119). However, EPCs were chosen to cook heavier(and cooker. Without additional training or design modifications, therefore more energy-intensive) dishes (figure C.6). They households with an EPC as their efficient appliance are thus can also be used for lighter staples (such as rice). As all likely to choose to cook roughly half their menu with it. participants in the Kenya cooking diaries had an electric hot plate, a rice cooker, and an EPC, it can be assumed that all Broadening to results from all four of the country studies, the dishes that were cooked in a rice cooker could also have table 2.5 in the main report shows the median daily energy been cooked in the EPC with the same energy consumption. consumption figures from the 100 percent eCooking stage of The average per capita, per heating event energy consump- the cooking diaries in each country. Table 2.6 shows compa- tion figure for rice cookers and EPCs comes to just under rable figures for the traditional fuels studied in each country. half (45 percent) that of the electric hot plate. Inspection of the daily cooking demands across the cooking diary samples shows that the distribution is not normal but has Further analysis of the Kenya cooking diaries data suggests instead a substantial tail toward higher loads. It represents that with minimal training, households would choose to use a mixture of days on which special meals are cooked and an EPC to cook half their menu if it were the only electric the presence of some cooks who routinely use more energy 128 ESMAP  |  Cooking with Electricity: A Cost Perspective Number of each category of dish cooked on inefficient (hot plate) and efficient (rice cooker FIGURE C.6  or electric pressure cooker) appliances reported in the Kenya cooking diaries 300 250 200 Number of Dishes 150 100 50 0 Heavy Foods Staples Quick Fryers Deep Fryers Long Fryers Other Electric Hotplate EPC Rice Cooker Note: Omitted from these figures are 127 records for dishes cooked on microwaves and kettles already owned by some participants. than others when cooking comparable dishes (and may thus This hypothesis is supported by the focus groups and be described as engaging in energy-inefficient practices). kitchen laboratory experimentation that followed the cooking The median is lower than the mean. Using the median in this diary study to produce the eCookBook (Chepkurui, Leary, analysis represents the sort of eCooking device that would be Numi, et al. 2019; Leary, Fodio Todd, et al. 2019). There was needed to cook food on the majority of days. also very limited choice of EPC models in Nairobi at the time, so some participants had to use models with known If indeed urban Kenyan households could cook over issues, such as intermittent shallow frying, complicated user 90 percent of their menu on an EPC with greater user interfaces, and inability to deep fry. What is more, as many training and experience combined with design improve- participants were used to cooking on a four-plate gas stove, ments, total energy consumptions would likely drop below the hot plate may well have been chosen simply to allow the figures shown in table 2.5. Heavy foods, staples, and more dishes to be cooked simultaneously. Households with quick-frying foods can all be cooked on an EPC, which only a single burner cooking device are forced to cook each together make up 91 percent of the urban Kenyan menu dish sequentially. (table C.1). With the exception of sausages, participants attempted to cook every dish in these three categories at least once. For instance, there were 102 meal events for ENERGY USED FOR COOKING ugali with a hot plate, but there were also 105 events with a rice cooker and 11 with an EPC. These data could be Table C.2 shows the measured energy demand values for interpreted as meaning that the EPC was not the preferred cooking fuels recorded during the cooking diary study (elec- device for these foods. However, the cooking diary study tricity demand values can be found in table 2.6). Tables 2.5 looked only at the first month in which participants used and 2.6 present the normalized values for electricity and fuel these appliances and only minimal training was given. It is consumption (for a 4.2-person household) that were used as likely that experimentation with cooking a broader range inputs for the case study modelling in this report. Table E.7, of dishes in the EPC did not occur until the end of that in appendix E, lists the calorific values of each fuel that were period. used to convert between energy content and weight. Appendix C: Assessing Electricity Demand for Cooking 12 9 Measured energy consumption for cooking all food on individual traditional fuels, by case TABLE C.2  study country FIREWOOD CHARCOAL KEROSENE LPG ENERGY PER HHe (MJ) ENERGY PER HHe (MJ) ENERGY PER HHe (MJ) ENERGY PER HHe (MJ) CAPITA ENERGY (MJ) CAPITA ENERGY (MJ) CAPITA ENERGY (MJ) CAPITA ENERGY (MJ) MEDIAN DAILY PER MEDIAN DAILY PER MEDIAN DAILY PER MEDIAN DAILY PER MEDIAN DAILY MEDIAN DAILY MEDIAN DAILY MEDIAN DAILY COUNTRY HH SIZEf HH SIZEf HH SIZEf HH SIZEf nd nd nd nd Kenyaa,b 17 10 4 2.50 129 8.1 3.2 2.53 Myanmar 62 23.9 4.2 5.69 26 32.1 5.9 5.44 26 7.2 3.3 2.18 Tanzaniaa 31 80.4 6.1 13.18 109 14.8 4.1 3.61 Zambiac 71 49.3 6.3 7.83 Note: Data are from cooking diary periods for traditional fuel use only. HH = household. a. Firewood data were not available in the Kenya or Tanzania cooking diary datasets, so consumption data were estimated using the ratio of firewood: charcoal energy consumption (approximately 1:1) from Myanmar and the ratio of firewood: charcoal energy density (approximately 1:2). b. Insufficient records were available for charcoal cooking from the cooking diaries study in Kenya to make a reliable measurement of charcoal consumption. As cooking practices are similar to those in Tanzania and electricity consumption was measured to be similar in the two counties, the values for charcoal cooking in Tanzania were also used as model inputs for Kenya. c. LPG data were not available in the Zambia cooking diary datasets, so consumption data were estimated using the average of the Tanzania: Zambia ratios for charcoal and eCooking with energy-efficient appliances (approximately 2:1). d. n = number of days of data using only that fuel. e. Median daily energy per HH = median daily energy consumption (kWh or MJ/household/day). f. Household size = household members cooked for (mean of means). ELECTRICITY LOAD PROFILE eCooking systems or off-peak electricity on mini grid or national grid systems. In Myanmar, most cooking occurs in It is important to understand not only the total amount of the morning, which means that battery storage is required for energy required for cooking but also when that energy is stand-alone systems. There is an evening peak in all coun- needed. Figure C.7 aggregates the data from all households tries, but it is clearly greatest in Kenya, Zambia, and Tanzania. over all the days on which they cooked solely with elec- For mini grid or national grid systems where demand peaks tricity to produce a set of load profiles. In Kenya, Zambia, in the evening and generating capacity is already at its limit, and Tanzania, the lunchtime peak occurs just after midday, battery storage or additional generation will be required to enabling solar electricity to be used directly on off-grid meet this additional load. 1 30 ESMAP  |  Cooking with Electricity: A Cost Perspective Normalized 24-hour load profile aggregated from all households in Myanmar, Kenya, FIGURE C.7  Tanzania, and Zambia 100% 80% 60% 40% 20% 0% 0 00 0 0 0 0 0 0 0 00 00 00 00 00 00 00 00 00 00 0 00 00 00 00 00 :0 :0 :0 :0 :0 :0 :0 :0 :0 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 00 00 00 00 00 00 00 00 00 1:0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 :0 0: 2: 3: 4: 5: 7: 8: 9: 6: 10 11 12 13 14 16 17 18 22 23 15 19 20 21 24 Kenya Myanmar Tanzania Zambia Appendix C: Assessing Electricity Demand for Cooking 131 APPENDIX D COMPARISON OF ECOOKING APPLIANCES Table D.1 compares a broad range of eCooking appliances, categorizing them into inefficient conventional, more effi- cient, and most efficient modern appliances. Section 1.4. discusses some of these appliances. TABLE D.1 Energy efficiency and versatility of eCooking appliances featured in this report HEAT TRANSFER HEAT TRANSFER TYPICAL POWER APPLIANCE INTO POT OUT OF POT REQUIREMENTS SPEED VERSATILITY Inefficient conventional appliances Hot plate Conduction when Convection and 1–2kW per hob Average Any pot (round bottom pot in contact with radiation from (DC: 300–700W) difficult); frying and element uninsulated pot; boiling evaporation without lid Electric oven Convection Cooking chamber 1–5kW Slow Baking, roasting, insulated, but not grilling only sealed; whole oven space around pot/dish heated More efficient modern appliances Kettle Conduction via Convection and 1.5–2.5kW Fast Single vessel; water immersed element radiation from boiling only uninsulated pot; fixed lid, but not completely sealed Slow cooker Conduction via Insulation and fixed 100–200W Very slow Single deep pot; insulated element lid, but not completely simmering only sealed Electric frying Conduction via Convection and 1–2kW Average Single shallow pot pan element stuck to pan radiation from only; frying and boiling uninsulated pot; evaporation without lid 1 32 ESMAP  |  Cooking with Electricity: A Cost Perspective HEAT TRANSFER HEAT TRANSFER TYPICAL POWER APPLIANCE INTO POT OUT OF POT REQUIREMENTS SPEED VERSATILITY Induction stove Induction Convection and 1–2kW per hob Fast Any flat-bottomed radiation from frying and ferrous pot; frying and uninsulated pot; bringing to boiling evaporation without lid boil Infra-red stove Radiation Convection and 1–2kW per hob Fast Any flat-bottomed pot; radiation from frying and frying and boiling uninsulated pot; bringing to evaporation without lid boil Halogen oven Radiation Convection and 700W–1.5kW Average Baking, roasting, radiation from grilling only uninsulated chamber; lid, but not completely sealed Most efficient modern appliances Rice cooker Conduction via Insulation and fixed 300W–1kW Average Single deep pot only; insulated element lid, but not completely (DC: 200–400W) boiling and some sealed frying, Microwave Microwave Cooking chamber 700W–1.5kW Fast Any nonmetallic dish; insulated, but not boiling sealed Insulated Conduction via Insulation; evaporation 700W–1.5kW Fast Single shallow pot electric frying insulated element without lid frying and only; frying and boiling pan stuck to pan bringing to boil Thermo-pot Conduction via Insulation and fixed 500W–1.5kW Slow Single vessel; water immersed element lid, but not completely (DC: 200–400W) boiling only sealed Electric Conduction via Insulation and fixed 700W–1.2kW Very fast Single deep pot only; pressure insulated element lid; completely sealed (DC: 200–400W) (pressurized) boiling and some cooker boiling frying Note: Results are based on findings of focus group discussions in Kenya, Tanzania, Zambia, and Myanmar, as summarized in table 2.1.  Advantage over other appliances   No particular advantage over other appliances   Disadvantage compared with other appliances Appendix D: Comparison of eCooking Appliances 133 APPENDIX E OUTLINE OF THE ECOOKING MODEL This appendix overviews the model structure and describes The eCooking model uses numerical simulations of the cook- and explains the assumptions made in the modelling, includ- ing undertaken by a household and the energy required, ing for the values used for key model parameters (Leach, linked to a system design model for an eCooking device, Leary, Scott, and Batchelor 2019) provide more detail). The either stand-alone (powered by a solar panel and battery) or full set of model inputs are summarized for each case study grid connected. The aim is to be able to compare the costs in appendix F. of eCooking with a baseline alternative (such as cooking with charcoal or LPG). The model includes a detailed treatment of cooking prac- Model Structure tices based on primary data, characterization of the costs of the major components based on learning rates, and an empirically based model for battery degradation that The model is designed to explore alternative ways to deliver captures the high current drain and harsh operating condi- the cooking service currently delivered by traditional stoves tions for this application. and fuels. The important metric is not cost per unit of elec- tricity delivered from an eCooking system but cost per meal. FIGURE E.1 eCooking model schematic showing key parameters Charcoal, firewood, Solar insolation Cooking kerosene or LPG stove • Daily peak Sun • Calorific value of fuel • Foods cooked Hours • E ciency of stove • Electricity required • Unit price of fuel • Fuel stacking choices Solar PV • Lifetime of PV Inverter Electric Cooking • Peak power rating Charge controller Battery pack (for AC Cooking) Appliances (AC or DC) • PV unit price • Power rating • Voltage, Storage factor • Power rating • Peak load • Lifetime of controller • Minimum depth of charge • Inverter e ciency • E ciency • Purchase cost • Round trip e ciency • Other cabling losses • Lifetime Grid • Degradation • Lifetime • Purchase cost (National or Mini-) • Lifetime • Purchase cost • Tari s • Battery capacity • Battery unit price Replacements required over 20 year system lifetime Financing Total discounted investment assumptions and operating cost • Discount rate • Financing period Cost of Cooking per Day 1 34 ESMAP  |  Cooking with Electricity: A Cost Perspective The energy needs for cooking define the requirements of irradiance, as additional battery capacity cannot help smooth one or more eCooking appliances and a matching inverter (if out month by month changes in PV output. Irradiance appliance uses AC electricity). The eCooking system is sized declines in the winter or monsoon months and rises in the to meet a user-defined fraction of this total cooking demand, summer or dry seasons. The PV-battery cooking system with the balance met from a specified traditional fuel. can be sized in two ways: (a) with a larger PV to operate year round as the principal means of cooking or (b) with a The required battery storage capacity can then be deter- smaller PV, capable of producing sufficient energy each day mined, along with a suitable charge controller. The solar PV in sunnier periods. The latter might work perfectly well for can be sized based on the daily need for battery charging some households happy to fuel stack. However, although and the solar insolation available. Alternatively, for an on-grid a smaller system would be cheaper, the capital cost of the or on-mini grid application, the load on the grid is calculated. system is spread out over fewer days, affecting its afford- The battery, solar PV (if used), and balance of plant are sized ability. In some locations, the variation in irradiance can be for daily load balancing. A user-defined factor is included, large—easily a factor of two—making the choice between a oversizing storage to allow for both unusually high cooking small or large system important. However, in many places in demands and/or reduction in grid or PV supply input so Africa the variation is much smaller. The model defaults to that the system can “ride through” an unusually cloudy day sizing the system so that it operates year-round and explores without the battery running flat, for example. Power losses alternatives in sensitivity analysis. The model also calculates are modelled in the wiring (typically 5 percent), the charge/ the likely “surplus” electricity stored in the battery each day discharge cycle (typically 10 percent), and the inverter (if during sunnier periods, which will be available for other elec- used). tricity end-uses as a co-benefit. Appropriate information is thus needed on each of the The online calculator of the European Union’s PVGIS project elements in this system, first for the system design and (Šúri, Huld, and Dunlop 2005) can be used to estimate PV sizing and then for costing. Financing assumptions follow electricity generation (per 1kW peak or rated output), with the business model to be represented, with the final result user selection of the location of the system. The main result a comparison of the daily costs of the eCooking system and is the average daily electricity output per kWpeak. The aver- cooking with traditional fuel purchases. age value for the month with the lowest output is used to size the system to operate year-round). The solar PV is sized as follows: Solar CPV: Capacity of PV (kWp) Edischarge: Daily battery discharge required (kWh/day) SOLAR INSOLATION AND PV OUTPUT Ed,min: Average daily electricity production in the least sunny month (kWh/kWp/day) The job of the PV system in this application is to deliver suffi- cient electricity each day to recharge the batteries so that ηbattery: Battery roundtrip efficiency ( percent) they are able to deliver the required electricity for cooking. A FPV decay: PV performance decay factor ( percent over lifetime) simple deterministic approach is taken to balancing electric- ity inputs and outputs within each day, estimating the aver- FPV oversize: Uprating factor to match battery oversize ( percent). age daily electricity output per kW peak and sizing the PV panels so that this average output is sufficient to recharge the batteries that day. However, a factor is added to explore PV COSTS the cost of increasing battery capacity so that it can “ride through” one or more days of low PV output, delivering A solar PV system can be described as a set of PV modules cooking service without running out before it is recharged; comprising individual solar cells held in some form of casing PV capacity is increased to match any such increase in and the balance of system, made up of wiring, installation battery capacity. equipment, and any inverter needed. For most PV systems, such as residential power or utility-scale solar farms, there is Although a fully dynamic model is beyond the scope of also a significant installation cost. For the current application, this study, it is essential to consider seasonal variation in the installation costs should be low. Appendix E: Outline of the eCooking Model 135 TABLE E.1 PV prices by model year ($/Wp) MODULE COST (AT OTHER BALANCE OF YEAR TOTAL PRICE FACTORY GATE) SYSTEM COSTS SALES COSTS 2018 0.72 0.46 0.07 0.19 2019 0.68 0.44 0.07 0.18 2020 0.65 0.42 0.06 0.17 2021 0.62 0.40 0.06 0.16 2022 0.59 0.38 0.06 0.15 2023 0.57 0.37 0.06 0.15 2024 0.56 0.36 0.05 0.14 2025 0.54 0.35 0.05 0.14 2026 0.53 0.34 0.05 0.14 2027 0.52 0.33 0.05 0.13 2028 0.50 0.33 0.05 0.13 2029 0.49 0.32 0.05 0.13 2030 0.48 0.31 0.05 0.12 PV cost projections are derived from historic data on module Batteries prices (IRENA 2018), demonstrating current prices of about $0.4/kWh. Price projections are based on expectations of This study focuses on the specific end-use of residen- growth in PV installed capacity leading of 1,760GW by 2030 tial-scale off-grid battery storage coupled with generation (IRENA, 2016), up from some 500GW today (IRENA 2019) from solar PV, with relatively rapid discharge on a daily cycle, and the learning rate (the percent cost reduction for each in what may well be hot and dusty conditions. The set of doubling of installed capacity). Historically, the PV module technical performance characteristics and specifications learning rate has been 18–22 percent, but IRENA (2019) for batteries is complex and interwoven. For example, the suggest 35 percent between 2010 and 2020. A continued number of cycles possible depends directly on the typical learning rate of 20 percent for modules out to 2030 is depth of discharge, and the relationships between these assumed here. parameters is highly dependent on the specific battery type To the resulting factory gate prices for modules alone, costs and chemistry as well as the management systems applied. are added for balance of plant (estimated as 15 percent, prin- This modelling seeks to identify key characteristics and cipally for wiring costs, following IRENA [2012]) and on-costs realistic ranges of values for each parameter, through which for transport and retailing in the study country (estimated as sensitivity analysis of the performance and costs of the 40 percent). Table E.1 shows the default PV price assump- system can be performed. tions used in the model. It is not easy to transfer data on battery performance in the literature to this specific eCooking application, because battery lifetime is complex, influenced by battery chemistry and construction, the conditions in the operating environ- ment, and the loads drawn. Leach and Oduro (2015) discuss these issues. 1 36 ESMAP  |  Cooking with Electricity: A Cost Perspective BATTERY SIZING temperature conditions is thin. The model of Wang et al. (2011) is used, which presents a generalized model for graph- The model focuses on lithium-Ion batteries, using ite-LiFePO4 cells (table E.3 sets out the parameters and the iron-phosphate chemistry. Table E.2 shows the parameters illustrative values). used in the model for the required battery capacity, with illustrative values representing an example system. −31700 + 370.3 × Crate Fcycledecay = A × exp × I 0.55 R×T 1 Cbatt = Edischarge × Fstorage × × 1 + Fdecay 1 − DoDmin So rearranging: Ebatt,avg Edischarge = × 1 + ωcable ηinverter Fcycledecay I = 0.55 The evidence base on battery performance and decay for −31700 + 370.3 × Crate A × exp cooking applications (high power draw) in high ambient R×T TABLE E.2 Lithium-Ion battery capacity model parameters PARAMETER NOTATION ILLUSTRATIVE VALUE Daily battery discharge required Edischarge 0.60 kWh/day Average electricity input to cooking appliance from battery Ebatt,avg 0.51 kWh/day Inverter efficiency ηinverter 0.9 (90 percent) Cable losses ωcable 0.05 (5 percent) Required battery capacity Cbatt 0.83 kWh Daily battery discharge required Edischarge 0.6 kWh/day Storage oversize factor (days [1 = full charge/discharge each day]) Fstorage 1 Minimum remaining charge level DoDmin 0.2 (20 percent) Additional design capacity added to account for decay loss in capacity Fdecay 0.1 (10 percent) of the battery (default is to add half the capacity lost by end of life: 20 percent/2 = 10 percent) TABLE E.3 Battery decay model parameters PARAMETER NOTATION ILLUSTRATIVE VALUE Ah-throughput (amount of charge delivered by battery during its lifetime) I 3,600Ah Loss of capacity as a result of charge and discharge over the operating Fcycledecay 0.2 (20 percent) life; normally chosen to be 20 percent Pre-exponential factor, empirically dependent on Crate, calculated below A 31,630 (see figure E.2) C rate (discharge current divided by the theoretical current draw under Crate 0.5 (full discharge in 2 which battery would deliver its nominal rated capacity in one hour) hours) Universal gas constant R 8.314 J/mol K Battery temperature T 40°C Appendix E: Outline of the eCooking Model 137 TABLE E.4 Battery cycle life model parameters PARAMETER NOTATION ILLUSTRATIVE VALUE Charge/discharge cycles before battery is replaced Cycleslife 2,300 Ah-throughput (amount of charge delivered by battery during its lifetime) I 3,600Ah Average depth of discharge in cycling DoDavg 0.8 (80 percent) Full cell capacity (standard cell size used to derive empirical relationship) 2Ah Now, BATTERY PRICES I = Cycleslife × DoDavg × Full cell capacity There is little evidence reported in the literature on the costs of modern batteries as implemented in developing coun- So tries at household or mini grid scales. Projection of battery prices is undertaken here based on expectations for electric Cycleslife = I vehicles, with assumptions for the transfer of learning to the DoDavg × Full cell capacity eCooking market. Leach and Oduro (2015) provide a detailed discussion of the historical evolution of battery prices. For the value of A, Wang et al. (2011) present an empirical relationship between A and discrete values of Crate (from 0.5 The annual Bloomberg New Energy Finance (BNEF) battery to 10). Figure E.2 shows the discrete values and the esti- price survey shows that prices have continued to fall—at mated continuous C rate used in the model. a faster rate than anticipated. In its publicly available data, BNEF shows only a single set of average historic prices and Thus the cycle life of the battery can be estimated based forecasts, acknowledging that there will be variation around on the operating temperature and the current drawn for the mean (Goldie-Scot 2019). Figure E.3 shows their most cooking. It would typically be 2,300 cycles, or six years of recent results. daily use. FIGURE E.2 Empirical relationship for battery life model A values 35000 30000 y = 448.96x2 – 6301.1x + 33840 Pre-exponential A value 25000 20000 15000 10000 5000 0 0 2 4 6 8 10 12 C Rate 1 38 ESMAP  |  Cooking with Electricity: A Cost Perspective FIGURE E.3 Actual and forecasted prices of battery packs, 2010–25 1200 1100 1000 900 Battery Pack Price ($/kWh) 800 700 600 500 400 300 200 100 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 EV battery packs EV battery packs (forecast) eCook battery pack (forecast) Note: EV = electric vehicle. Source: Goldie-Scot (2019). Factors were used to transfer these projections for electric BATTERY CHARGE CONTROLLER vehicle battery pack prices into estimates for eCooking battery packs. As suggested by Frith (2017), 51 percent was A controller is needed to manage the interaction between added to account for the typical cost premium for stationary the source of electricity (grid or PV panel) and the batteries, battery pack prices. Another 20 percent was added to reflect in order to protect the battery from being overcharged or the costs for transport and import into Africa, leading to esti- overdischarged, to protect against battery overheating, and mates of $270/kWh in 2020 and $161/kWh by 2025. to maximize the efficiency of the use of the solar power. The model allows user choice between cheaper pulse width The use of static factors of this sort represents a major modulation (PWM) and more expensive and more efficient simplification. However, there is little evidence on the real maximum power point tracking (MPPT) controllers: for the costs of household-scale battery pack prices in Africa; further current report, PWM has been chosen. analysis is required in this area. New battery chemistries may also enter the market, offering improved performance, All PV controllers need to be sized to cope with the system’s longer lifetimes, and lower costs. Limited data or market voltage and the maximum amount of current that might flow intelligence on these developments are available, however. through them. To size the PWM controller, the required rated Sensitivity analysis will need to take account of the uncer- current in amps is calculated from the PV output wattage tainty about battery performance and prices . divided by the PV’s peak power output voltage (for exam- ple, 17v), which is taken from the solar PV panel or array specifications. It is recommended to oversize the controller to allow for peak outputs and to provide a safety margin against overheating in continuous use. The modelling here Balance of System allows separate user-defined safety factors for each of these margins, defaulting to +25 percent each, following typical industry practice. In addition to the PV/grid supply, battery, and cooking appli- ances, other components are required to ensure an efficient, The cost of the controller depends strongly on its rated safe, and long-lasting eCooking device. They include the capacity. However, at any capacity level, there is a wide price battery charge controller, the battery management system, range for battery charge controllers, reflecting the features the inverter, and additional wiring. (for example, the degree of battery temperature protection Appendix E: Outline of the eCooking Model 139 TABLE E.5 Rated load and cost of selected charge controller models MODEL RATED LOAD CURRENT (AMPS) COST ($) Morningstar SHS-6 more than 6 Amp 12 Volt 6 24 Morningstar SHS-10 more than 10 Amp 12 Volt 10 32 Morningstar SK-12 more than SunKeeper 12 Amp 12 Volt 12 71 Morningstar PS-15 (12/24V) 15 96 Morningstar PS-30 (12/24V) 30 128 Morningstar TS-45 (12/24/48V) 45 167 Morningstar TS-60 (12/24/48V) 60 222 Note: The device is chosen to have a rated load current that exceeds the maximum cooking current expected multiplied by the peak load safety factor and the continuous use safety factor. Source: www.ecodirect.com/. and efficiency) and overall quality and hence expected life. A dedicated BMS can perform all of these functions. It would also be expected that significant savings could However, high-power batteries on the market are usually be made between one-off purchase of a standard charger sold in packs of sets of cells chosen to match each other retail and the cost of a bespoke controller designed into an well and assembled in parallel or series to deliver the eCooking system. required voltage and current discharge capability. These packs contain built-in protection circuits of one sort or The modelling includes a database of a sample of stan- another. As such, the BMS functions can be split between dard retail models. An appropriately sized controller is then the battery pack and the charge controller, obviating the selected to match the characteristics (notably the voltage need for a stand-alone BMS. This study therefore does not and maximum current expected) of the eCooking system model the need for an additional BMS, instead specifying (table E.5). The Morningstar SHS-6 or SHS-10 will typically charge controllers that perform the necessary functions. For provide the required capacity for the eCooking systems a bespoke eCooking design, batteries, a BMS, and charge modelled to date. controller could be integrated in different configurations. Research will be needed to determine the optimal design, balancing performance, lifetime, and cost within necessary safety limits. FOR GRID-CONNECTED ECOOKING SYSTEMS, A LESS SOPHISTICATED BATTERY CHARGER IS ASSUMED, AS MANAGEMENT OF THE PV OUTPUT IS NOT REQUIRED.
BATTERY MANAGEMENT INVERTER SYSTEM It is possible to cook by connecting a DC hot plate or a DC A battery management system (BMS) of some sort is essen- EPC directly to a battery pack. However, achieving the required tial for any rechargeable battery system. At its simplest, a power for cooking from the hot plate or the initial heating BMS prevents the battery from operating outside its safe period for an EPC (for example, 500W–1,000W) implies high limits—for example, by protecting against discharge outside current flow in the cables to the hot plate, with commensurate current limits. In practice batteries should be managed more losses. It is difficult to buy high-power DC hot plates, and only a actively, with monitoring of the state of charge (of the pack few DC EPCs are available, although there is growing commer- or ideally of each cell) and measurement of temperature and cial interest in them. This study assumes that DC appliances voltage. The quality of control achieved will influence the become widely available. It also assumes that hybrid appliances performance of the battery pack as well as its degradation are available, to allow operation from a mix of grid AC and and lifetime. battery DC, as and when AC supply is available. 140 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE E.6 Capacity and cost of selected modified sine wave (12v) inverter models MODEL INVERTER OUTPUT CONTINUOUS LOAD (KW) COST ($) Samlex SAM-1000-12 1.0 96.29 Samlex SAM-1500-12 1.5 174.93 Samlex SAM-2000-12 2.0 251.75 Samlex SAM-3000-12 3.0 367.50 Source: EcoDirect (2020). The model also allows for an alternative approach: use of a of impact on likely costs of a real system, this approach has DC to AC inverter, allowing the use of readily available and both positive and negative effects. A bespoke eCooking low-cost AC electric hot plates and other AC cooking appli- design should be able to achieve some cost savings by ances. The other advantage of integrating an inverter is that integrating functionality in, for example, battery control and the household could potentially use the resulting AC power sizing components more precisely. However, it would add for other purposes (lighting, mobile phone charging, radio, costs (for system wiring, for example) that are not captured TV). The use of an inverter is not modelled in the current by assuming a collection of stand-alone components. study, but the model is able to do so. Modelling of systems with an inverter and the use of different combinations of AC The potential benefits of tighter system integration in and DC appliances will be necessary in the future. mass-market eCooking design are ignored at this stage, in order to avoid making overly optimistic assumptions. Different inverter technologies offer varying quality in output However, the additional balance of system, such as wiring, is power. There are three main types: sine wave, modified sine reflected in a user-defined parameter, defaulting by adding wave, and square wave. Most devices will operate accept- 5 percent to the total system investment cost. ably well with a modified sine wave. There is a wide range of specifications and prices for inverters on the market. This model includes a small data- base of popular types (table E.6). Sizing is based on both Cooking Appliances the continuous power rating required, to meet anticipated cooking loads, and the “surge” rating (the maximum power Data from the cooking diary studies were used to estimate the inverter can supply for a short period), to cope with the the energy required for cooking with different fuels and high start-up load some devices draw. Modern inverters can electricity. The current version of the model was set up to typically cope with a surge of up to 300 percent for 3–15 represent hot plates and EPCs. Devices were assumed to seconds, which is sufficient to cope with the load profile of be either low-cost two-ring hot plates purchased for $20 or almost all appliances. Inverters are not 100 percent efficient: EPCs purchased for $50 (Leary et al. 2018 provide a review The model includes a user-defined value for efficiency, of EPCs). defaulting to 90 percent. The efficiency affects the required battery sizing and hence the PV sizing (or power drawn from the grid). Component Lifetimes and ADDITIONAL BALANCE OF SYSTEM Replacements The rest of the model looks at the major components as Each component is assigned a technical lifetime within individual items, sizing and choosing them from databases an overall system modelling horizon of 20 years, chosen of typical options on the market rather than attempting an to reflect the notional lifetime of the longest-lived major engineering design of an integrated whole system. In terms component, the PV. The battery lifetime is derived from the Appendix E: Outline of the eCooking Model 141 TABLE E.7 Lifetime and price assumptions about eCooking components COMPONENT LIFETIME (YEARS) PROJECTED PRICE TRAJECTORY Overall eCooking system 20 Based on components PV 20 Decline, based on learning rate Battery Calculated in model Decline, based on learning rate Inverter 10 Decline by 2 percent a year Charge controller 6 Decline by 2 percent a year Cooking appliances 5 Constant decay model presented above. Component replacement is cost of cooking per month, which can be directly compared modelled throughout the system life, with additional capital with the cost of traditional fuel purchases to undertake the cost added each time a replacement component is needed. same cooking. The model allows for changes in the cost of components over time, such that replacements are made with the costs The core business model envisages a supplier of the eCook- expected at the time of replacement. Cost changes are most ing service who pays the initial and replacement capital costs significant for batteries, for which significant cost reductions over the 20-year period in exchange for a daily or monthly are projected and which require one or more replacements user fee, as it is unrealistic to imagine that a low-income user during the 20-year system lifetime. Table E.7 shows the would make any form of agreement for 20 years. A 20-year assumptions made for lifetime and replacement cost param- financing model could reflect some form of utility-based eters. All parameter values are user-definable and therefore business model, where the electricity supplier bears the risk open to sensitivity analysis. and recovers the investment costs through an additional fee alongside the regular bill. Some other risk-bearing arrange- ment with the same effect is conceivable if installation of eCooking devices is made part of national energy access Business Models and infrastructure or via development aid or carbon finance. Investment Financing The more traditional model would be for cost recovery over a shorter period, as for solar home systems. The model thus incorporates a shorter time period for this form of commer- The model is structured to calculate the costs required to cial business model. The time period is user-definable. This deliver the eCooking service for 20 years, including replace- study uses five years—still longer than for many commer- ment costs for the other components during that period. The cial services, including solar home systems (for example, basis of the service fee calculation is the levelized cost of the two years). Further innovation in business models may be cooking service, expressed as the monthly cost of cooking. needed to make this high-capital cost type of appliance and service accessible. System cost is the sum of operating costs (grid electricity purchase, traditional fuel purchase); the initial capital cost; The normal practice in similar markets is that after the end of installation costs; and component replacement costs. Costs the financing period, ownership of the equipment transfers are discounted back to present-day values using a user-­ to the user. This means that responsibility for further compo- defined discount rate. The cost basis throughout the model nent replacements also transfers to the user, with the risk uses real costs, ignoring inflation; a real discount rate is that the system falls into disrepair and thus disuse when a therefore applied.5 The key output metric is the net present major component fails. 5 A 9.6 percent real discount rate is used throughout, following (Lombardi et al. 2019) the techno-economic feasibility of e-cooking has never been evaluated through (i. Reported interest rates are frequently nominal rates, taking no account of the effect of inflation. For a country with average inflation of 10 percent (typical in parts of East Africa), a 9.6 percent real equates to a 19.6 percent nominal rate. 142 ESMAP  |  Cooking with Electricity: A Cost Perspective TABLE E.8 Calorific value and greenhouse gas emissions of selected fuels CALORIFIC VALUE (KWH/KG, LOWER GREENHOUSE GAS EMISSIONS FACTOR FUEL HEATING VALUE) (KG CO2-EQ EMISSIONS PER KWH) Charcoal 7.9 0.32168 Firewood 4.1 0.015a LPG 12.6 0.2303 Kerosene 11.9 0.2574 Note: a. The greenhouse gas emission factor for firewood depends on assumptions about the sustainability of wood harvesting. It is assumed here that wood is harvested sustainably, with replanting and regrowth. The low emission factor reflects non-CO2 emissions. All parameter values are user-definable; ideally, they come Fuel/Appliance Stacking from the area being studied. Fuel and electricity prices in particular vary widely by location and change over time, and electricity emission factors vary by country. The assump- The model seeks to represent the energy used to meet a tions used in this study are described in the main body of household’s daily cooking requirements, which can be met the report. Table E.8 shows the default calorific values and by an eCooking system sized to meet the full cooking load greenhouse gas emission factors for traditional fuels. If itself or by a combination of a smaller eCooking system and These figures can be tailored to reflect local conditions. fuel stacking with a traditional fuel and/or grid electricity directly. The proportions of each source are user defined. The characteristics of each energy source are defined by parameters for traditional fuels (calorific value, CO2-eq Treatment of Uncertainty emissions per kWh, price per kWh) and for grid electricity (marginal CO2-eq emissions per kWh, price per kWh for a There is a wide range of uncertainty in the design, sizing, series of tariff bands [free lifeline plus up to three tariff bands and costing of a system to deliver cooking services. The can be user-defined]). modelling distinguishes between parameters for which the values are uncertain as a result of four factors: TABLE E.9 Parameter values used for the high- and low-cost scenarios for eCooking systems 2020 2025 PARAMETER LOW-COST VALUE HIGH-COST VALUE LOW-COST VALUE HIGH-COST VALUE Battery price (lithium-ion, $/kWh) 280 350 180 220 Battery minimum depth of charge 10 20 10 20 (percent) Battery life (cycles) 3,000 2,000 3,000 2,000 PV-battery roundtrip efficiency 90 85 90 85 (percent) Fuel prices 2/3 of 2018 mean 4/3 of 2018 mean 2018 low value + 2018 high value + valuea value 3 percent/year 3 percent/year Note: All financial values are in 2018 dollars. a. Some values are from late 2017 or early 2019. Appendix E: Outline of the eCooking Model 143 ● different or varying household cooking needs and practices Model Implementation ● uncertainty in appropriate values for parameters ● different financing assumptions The model is implemented in Microsoft Excel, using a set of Visual Basic macros to automate certain processes. The ● changes in parameter values over time. structure is modular, with separate tabs for each major component and modelling process. Application of the model in this study is intended to be illustrative and not comprehensive. It focuses on appropriate Implementation is intended to be user-friendly, but the values for parameters, captured through sets of assumptions model remains a research tool, lacking a comprehen- that lead to lower or higher cost systems, and changes in sive graphical user interface. Drop-down boxes are values over time, through a comparison between eCooking used for discrete choices, and parameter values can be systems implemented in 2020 or 2025. For changes over entered directly throughout the spreadsheet, with cells time, two main differences are represented: (a) declining intended for user-input indicated by red outlines. The costs for eCooking as a result of technical and organizational screenshot in figure E.4 is of the front tab, where key learning and (b) an assumption of increasing charcoal, LPG, user choices are brought together, along with reporting firewood, and kerosene prices. of key results. Table E.9 shows the assumptions, most of which are This is a simulation model and is computationally not inten- discussed in the report. Fuel prices reflect results obtained sive. The input data and output results can be saved at any from household surveys carried out alongside the cooking moment to a tab that accumulates results, so that a series of diary studies in 2017–19 and an assumption of 3 percent scenario or sensitivity runs (for example, incrementing the annual price increases thereafter. A high/low range is then cost of charcoal) can be run very easily. applied around these values by adding/subtracting one-third. FIGURE E.4 Screenshot of eCook model user interface 144 ESMAP  |  Cooking with Electricity: A Cost Perspective APPENDIX F MODEL INPUT DATA Table F.1 presents the parameter values used as inputs to every case study, and not all components are used in every the model and shows key intermediate and final calculated case study (for example, PV is used only in case 5). For this outputs. Not all variants shown in the table are relevant to reason, some cells are empty. Appendix F: Model Input Data 145 TABLE F.1 Model parameters and values used 146 Installation Case 1 Nairobi, Kenya 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 year Urban households switch from charcoal to Cooking Description eCook & LPG scenario Input, INT mediate, Output 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Component Parameter Units Cooking eCook electricity use for cooking I kWh/day 1.92 1.92 1.92 0.64 0.64 1.92 1.92 1.92 0.64 0.64 Fuel stacked with eCook: charcoal INT kg/day 0.87 0.87 0.87 0.87 Fuel stacked with eCook: LPG INT kg/day 0.11 0.11 0.11 0.11 Fuel stacked with eCook: fuelwood INT kg/day Fuel stacked with eCook: kerosene INT kg/day 0.12 0.12 0.12 0.12 Baseline: fuel used: charcoal INT kg/day 1.75 1.75 Baseline: fuel used: LPG INT kg/day 0.23 0.23 Baseline: fuel used: fuelwood INT kg/day Baseline: fuel used: kerosene INT kg/day 0.25 0.25 Cooking Appliance(s) used I EPC & EPC & EPC & EPC EPC EPC & EPC & EPC & EPC EPC appliances hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 70.0 70.0 70.0 50.0 50.0 63.3 63.3 63.3 45.2 45.2 Electricity inputs Source of electricity I Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid AC or DC cooking I AC DC Hybrid AC DC AC DC Hybrid AC DC Fuel prices Electricity tariffs I $/kWh 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 Charcoal I $/kg (Low, High) 0.4, 0.6 0.4, 0.6 0.4, 0.6 0.46, 0.70 0.46, 0.70 0.46, 0.70 LPG I $/kg (Low, High) 0.8, 1.5 0.8, 1.5 0.8, 1.5 0.93, 1.74 0.93, 1.74 0.93, 1.74 Firewood I $/kg (Low, High) Kerosene I $/kg (Low, High) 0.87, 1.62 0.87, 1.62 0.87, 1.62 1.01, 1.88 1.01, 1.88 1.01, 1.88 PV sizing Lifetime I Years Energy from solar irradiance I KWh/KWpeak PV performance decay (and oversize factor) I % over life Cooking directly from PV I % Design peak power rating INT Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels O $ Charge Lifetime I Years 6 6 6 6 6 6 controller Charge controller capacity (incl. safety factors) INT Amps 12 6 6 12 6 6 Battery charging roundtrip efficiency I % (Low, High) 90, 85 90, 85 90, 85 90, 85 90, 85 90, 85 Purchase cost of controller O $ 9.22 9.22 9.22 8.34 8.34 8.34 Purchase cost of additional Balance of System O $ 39.4, 49.1 13.4, 16.7 13.4, 16.7 25.4, 31.0 8.8, 10.6 8.8, 10.6 Battery Battery voltage I V 24.0 24.0 24.0 24.0 24.0 24.0 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 20.0 20.0 Useable max capacity remaining at replacement I % (Low, High) 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 Capacity addition to account for decay I % 10.0 10.0 10.0 10.0 10.0 10.0 Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (Low, High) 3000, 3000, 3000, 3000, 3000, 3000, 2000 2000 2000 2000 2000 2000 Battery capacity INT kWh 2.8 0.93 0.93 2.8 0.93 0.93 Unit price (delivered) I $/kWh (Low, High) 280, 350 280, 350 280, 350 180, 220 180, 222 180, 226 Purchase cost of battery O $ (Low, High) 777.8, 259.0, 259.0, 500.0, 167.4, 167.4, 972.3 324.0 324.0 611.1 204.6 204.6 Total system Initial purchase cost O $ (Low, High) 70.00 896.4, 351.6, 50.0, 50.0 331.6, 63.30 597.0, 247.8, 45.2, 45.2 229.7, 1100.6 419.9 399.9 713.7 286.9 268.8 ESMAP  |  Cooking with Electricity: A Cost Perspective Case 2: efficient Lusaka, Zambia Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 appliances Urban households using electricity, Cooking Description switch to efficient appliances scenario Component Parameter Units Input, INT mediate, Output 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Cooking eCook electricity use for cooking I kWh/day 0.87 0.87 0.87 0 29 0.29 0.87 0.87 0.87 0.29 0.29 Fuel stacked with eCook: charcoal INT kg/day 0.52 0.52 0.52 0.52 Fuel stacked with eCook: LPG INT kg/day 0.03 0.08 0.08 0.08 Appendix F: Model Input Data Fuel stacked with eCook: fuelwood INT kg/day Fuel stacked with eCook: kerosene INT kg/day Baseline: fuel used: charcoal INT kg/day 1.04 1.04 Baseline: fuel used: LPG INT kg/day 0.17 0.17 Baseline: fuel used: fuelwood INT kg/day Baseline: fuel used: kerosene INT kg/day Cooking Appliance(s) used I EPC & EPC & EPC & EPC EPC EPC & EPC & EPC & EPC EPC appliances hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances $ 70.0 70 70 50 50 63.3 63 63 45 45 Electricity inputs Source of electricity I Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid AC or DC cooking I AC DC Hybrid AC DC AC DC Hybrid AC DC Fuel prices Electricity tariffs I $/kWh 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 Charcoal I $/kg (Low, High) 0.15, 0.30 0.15, 0.30 0.15, 0.30 0.17, 0.35 0.17, 0.35 0.17, 0.35 LPG I $/kg (Low, High) 1.50, 2.50 1.50, 2.50 1.50, 2.50 1.74, 2.90 1.74, 2,90 1.74, 2.90 Firewood I $/kg (Low, High) Kerosene I $/kg (Low, High) PV sizing Lifetime I Years Energy from solar irradiance I KWh/KWpeak PV performance decay (and oversize factor) I % over life Cooking directly from PV I % Design peak power rating INT Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels 0 $ Charge Lifetime I Years 6 6 6 6 6 6 controller Charge controller capacity (incl. safety factors) INT Amps 6 6 6 6 6 6 Battery charging roundtrip efficiency I % (Low, High) 90,85 90,85 90, 85 90,85 90, 85 90, 85 Purchase cost of controller 0 $ 9.22 9.22 9.22 8.34 8.34 8.34 Purchase cost of additional Balance of System 0 $ 18.1, 22.5 6.3, 7.8 6.3, 7.8 11.8, 14.3 4.2, 5.0 4.2, 5.0 Battery Battery voltage I V 24.0 24.0 24.0 24.0 24.0 24.0 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 20.0 20.0 Useable max capacity remaining at replacement I % (Low, High) 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 Capacity addition to account for decay I % 10,0 10.0 10.0 10.0 10.0 10.0 Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (Low, 3000, 3000, 3000, 3000, 3000, 3000, High) 2000 2000 2000 2000 2000 2000 Battery capacity INT kWh 1.3 0.42 0.42 1.3 0.42 0.42 Unit price (delivered) I $/kWh (Low, High) 280.00 280.00 280.00 180.00 180.00 180.00 Purchase cost of battery 0 $ (Low, High) 355.1, 117.7, 147.1 117.7, 147.1 227.0, 75.7, 92.5 75.7, 92.5 441.4 277.4 Total system Initial purchase cost 0 $ (Low, High) 70.00 450.4, 203.3, 50.0, 50.0 183.3, 63.27 310.4, 151.5, 45.2, 45.2 133.4, 543.1 234.2 214.2 363.3 169.1 151.1 147 148 Case 3 Shan State, Myanmar Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 Micro-hydro minigrid; housdeholds switch to Cooking Description efficient appliances scenario Component Parameter Units Input, INT mediate, Output 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Cooking eCook electricity use for cooking I kWh/day 1.08 1.08 1.08 0.36 0.36 1.08 1.08 1.08 0.36 0.36 Fuel stacked with eCook: charcoal INT kg/day Fuel stacked with eCook: LPG INT kg/day 0.10 0.10 0.10 0.10 Fuel stacked with eCook: fuelwood INT kg/day 0.77 0.77 0.77 0.77 Fuel stacked with eCook: kerosene INT kg/day Baseline: fuel used: charcoal INT kg/day Baseline: fuel used: LPG INT kg/day 0.20 0.20 Baseline: fuel used: fuelwood INT kg/day 1.54 1.54 Baseline: fuel used: kerosene INT kg/day Cooking Appliance(s) used EPC & EPC& EPC& EPC EPC EPC & EPC& EPC & EPC EPC appliances hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances | $ 70.0 70 70 5O 50 63.3 63 63 45 45 Electricity inputs Source of electricity | Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid AC or DC cooking | AC DC Hybrid AC DC AC DC Hybrid AC DC Fuel prices Electricity tariffs | $/kWh 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 Charcoal | $/kg (Low, High) LPG I $/kg (Low, High) 1.0, 2.0 1.0, 2.0 1.0, 2.0 1.16, 2.32 1.16, 2.32 1.16, 2.32 Firewood | $/kg (Low, High) 0.10, 0.20 0.10, 0.20 0.10, 0.20 0.12, 0.23 0.12, 0.23 0.12, 0.23 Kerosene I $/kg (Low, High) PV sizing Lifetime I Years Energy from solar irradiance | KWh/KWpeak PV performance decay (and oversize factor) | % over life Cooking directly from PV | % Design peak power rating INT Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels 0 $ Charge Lifetime | Years 6 6 6 6 6 6 controller Charge controller capacity (incl. safety factors) INT Amps 6 6 6 6 6 6 Battery charging roundtrip efficiency I % (Low, High) 90, 85 90, 85 90, 85 90, 85 90, 35 90, 85 Purchase cost of controller 0 $ 9.22 9,22 9.22 8.34 8.34 8.34 Purchase cost of additional Balance of System 0 $ 22.3, 27.8 7.8, 9.6 7.8, 9.6 15.6, 18.9 5.1, 6.1 5.1, 6.1 Battery Battery voltage | v 24.0 24.0 24.0 24.0 24.0 24.0 Battery minimum depth of charge | % 20.0 20.0 20.0 20.0 20.0 20.0 Useable max capacity remaining at replacement | % (Low, High) 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 Capacity addition to account for decay | % 10.0 10.0 10.0 10.0 10.0 10.0 Efficiency loss in wiring | % 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (Low, 3000, 3000, 3000, 3000, 3000, 3000, High) 2000 2000 2000 2000 2000 2000 Battery capacity INT kWh 1.6 0.52 0.52 1.6 0.52 0.52 Unit price (delivered) | $/kWh (Low, High) 280.00 280.00 280.00 180.00 180.00 180.00 Purchase cost of battery 0 $ (Low, High) 437.3, 145.8, 145.8, 304.0, 93.7, 114.5 93.7, 114.5 546.7 182.2 182.2 369.6 Total system Initial purchase cost 0 $ (Low, High) 70.00 538.9, 232.8, 50.0, 50.0 212.8, 63.27 391.2, 170.4, 45.2, 45.2 152.4, 653.7 271.0 251.0 460.1 192.3 174.2 ESMAP  |  Cooking with Electricity: A Cost Perspective Case 4 Kibindu district, Tanzania Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 Solar-hybrid minigrid; households Cooking Description eCook & LPG. Microbusiness scenario eCook Component Parameter Units Input, INT mediate, Output Micro business 100% AC electric 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Micro business 100% AC electric 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Cooking eCook electricity use for cooking I kWh/day 0.15 2.06 2.06 2.06 0.685 0.6S5 0.15 2.06 2.06 2.OS 0.69 0.69 Fuel stacked with eCook: charcoal INT kg/day 0.87 0.87 0.87 0.87 Fuel stacked with eCook: LPG INT kg/day 0.16 0.16 0.16 0.16 Appendix F: Model Input Data Fuel stacked with eCook: fuelwood INT kg/day 1.75 1.75 1,75 1.75 Fuel stacked with eCook: kerosene INT kg/day Baseline: fuel used: charcoal INT kg/day 1.75 1.75 Baseline: fuel used: LPG INT kg/day 0.33 0.33 Baseline: fuel used: fuelwood INT kg/day 3,50 5.50 Baseline: fuel used: kerosene INT kg/day Cooking Appliance(s) used I EPC EPC & EPC & EPC & EPC EPC EPC EPC & EPC & EPC & EPC EPC appliances hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 50.0 70.0 70 70 50 50 45 63.3 63 63 45 45 Electricity Source of electricity Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid Miningrid inputs AC or DC cooking AC AC DC Hybrid AC DC AC AC DC Hybrid AC DC Fuel prices Electricity tariffs $/kWh 0.11.4 1.35 1.35 1.35 1.35 1.35 0.1–1.4 1.35 1.35 1 35 1.35 1.35 0.10, 0.20 Charcoal $/kg (Low. High) 0.10 0.10 0.10, 0.20 0.12 0.12 1.0, 2.0 LPG $/kg (Low. High) 1.00 1.00 1.0, 2.0 1.16 1.16 0.03, 0.05 Firewood $/kg (Low. High) 0.03 0.03 0.03, 0.03 0.03 0.05 Kerosene $/kg (Low. High) PV sizing Lifetime I Years Energy from solar irradiance I KWh/KWpeak PV performance decay (and l % over life oversize factor) Cocking directly from PV I % Design peak power rating INT Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels 0 $ Charge Lifetime I Years 6 6 6 6 6 6 controller Charge controller capacity (incl. INT Amps 12 6 6 12 6 6 safety factors) Battery charging round trip efficiency I % (Low, High) 90, 85 90, 85 90, 85 90, 85 90, 85 90, 85 Purchase cost of controller 0 $ 9.22 9.22 9.22 8.34 8.34 8.34 Purchase cost of additional Balance 0 $ 42.1, 52.5 14.3, 17.8 14.3, 17.8 77.2, 33.1 9.3, 11.5 9.3, 11.3 of System Battery Battery voltage I v 24.0 24.0 24.0 24.0 24.0 24.0 Battery minimum depth of charge I % 20.0 20.0 20,0 20,0 20.0 20.0 Useable man capacity remaining at % (Low, High) 90.0, 90.0, 90.0, 90.0, 90.0, 90.0, replacement 80.0 80.0 80.0 80.0 80.0 80.0 Capacity addition to account for decay % 10.0 10.0 10.0 10.0 10.0 10.0 Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (low. High) 3000, 3000, 3000, 3000, 3000, 3000, 2000 2000 2000 2000 2000 2000 Battery capacity INT kWh 3.0 0.99 0.99 3.0 0.99 0.99 Unit price (delivered) l $/kWh (Law, High) 280.00 280.00 280.00 180.00 180.00 180.00 Purchase cost of battery 0 $ {Low. High) 833.0, 277.7, 277.7, 535.5, 178.5, 173.5, 1041.3 347.1 347.1 654.5 218.2 213.2 Total system Initial purchase cost 0 $ (Law, flight) 50.00 70.00 954.4, 371.2, 50.00 351.2, 45.20 63.27 634.3, 259.5, 45.20 241.4, 1373.0 444.1 424.1 759.3 301.1 283.0 149 150 Case 5 Echariria, Kenya Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 Cooking Description Offgrid solar-battery cooking scenario Component Parameter Units Input, INT mediate, Output 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Traditional 50% batt DC, 50% Traditional 100% charcoal 100% LPG 100% firewood 100% Kerosene Cooking eCook electricity use far cooking I kWh/day 1.92 0.64 1.92 0.64 Fuel stacked with eCook: charcoal INT kg/day 0.87 0.87 Fuel stacked with eCook: LPG INT kg/day 0.11 0.11 Fuel stacked with eCook: fuel wood INT kg/day Fuel stacked with eCook: kerosene INT kg/day 0.12 0.12 Baseline: fuel used: charcoal INT kg/day 1.75 1.75 Baseline: fuel used: LPG INT kg/day 0.23 0,23 Baseline: fuel used: fuel wood INT kg/day 3.50 3.50 Baseline: fuel used: kerosene INT kg/day 0.25 0.25 Cooking Appliance(s) used I EPC & EPC EPC & EPC appliances hotplate hotplate Lifetime of cooking appliances I Vears 5 5 5 5 Purchase cost of cooking appliances I $ 70 50 63 45 Electricity inputs Source of electricity I PV PV PV PV AC or DC cocking I DC DC DC DC Fuel prices Electricity tariffs I $/kwh Charcoal I $/kg (Low, High) 0.20, 0.40 0.20, 0.40 0.23,0.46 0.23, 0.46 LPG I $/kg (Low, High) 0.80, 1.50 0.80, 1.50 0.93, 1.74 0.93, 1.74 Firewood I $/kg (Low, High) 0.09, 0.17 0,09, 0.17 0.10, 0.20 0.10, 0,20 Kerosene I $/kg (Low, High) 0,87, 1.62 0.87, 1.62 1.01, 1.88 1.01, 1.88 PV sizing Lifetime I years 20.00 20.00 20.00 20.00 Energy from solar irradiance I KWh/KWpeak 3,85 3.85 3,85 3.85 PV performance decay (and oversee factor) I %over life 10.00 10.00 10.00 10.00 Cooking directly from PV I % 20.00 20.00 20.00 20.00 Design peak power rating INT Wpeak 0.63 0.21 0.63 0.21 Unit price (delivered) I $/Wpeak 0.42 0.42 0.35 0.35 Purchase cost of PV panels O $ 264.46 88.15 221.95 73.98 Charge Lifetime I Vears 6 6 6 6 controller Charge controller capacity (incl. safety factors) INT Amps 45 45 45 45 Battery charging roundtrip efficiency I % (Low, High) 90, 85 90, 85 90, 85 90, 85 Purchase cost of controller 0 $ 154.04 154.04 139.24 139.24 Purchase cost of additional Balance of System O $ 52.0, 60.6 22.5, 25.3 38.1, 43.2 17.3, 19.0 Battery Battery voltage I V 24.0 24.0 24.0 24.0 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 Useable max capacity remaining at replacement I % (Low, High) 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 Capacity addition to account for decay I % 10.0 10.0 10.0 10.0 Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (Low, 3000, 3000, 3000, 3000, High) 2000 2000 2000 2000 Battery capacity INT kWh 2.2 0.74 2.2 0.74 Unit price (delivered) I $/kWh (Low, High) 280.00 280.00 180.00 180,00 Purchase cost of battery 0 $ (Low, High) 622.3, 207.4, 400.0, 133.3, 777.8 259.3 488.9 163.0 Total system Initial purchase cost 0 $ (Low, High) 1162.8, 522.1, 852.5, 409.1, 1342.5 582.0 969.6 444.8 ESMAP  |  Cooking with Electricity: A Cost Perspective Case 2: inefficent Lusaka, Zambia Installation year 2025 2025 2025 2020 2020 2020 appliance owners Urban households using electricity, switch to Cooking Description mix of efficient appliances and inefficient scenario Component Parameter Units Input, INT mediate, Output 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC 100% AC electric 100% DC battery- electric 50% AC, 50% Batt DC Cooking eCook electricity use for cooking I kWh/day 3.15 3.15 3.15 3.15 3.15 3.15 Fuel stacked with eCook: charcoal INT kg/day Fuel stacked with eCook: LPG INT kg/day Appendix F: Model Input Data Fuel stacked with eCook: fuelwood INT kg/day Fuel stacked with eCook: kerosene INT kg/day Baseline: fuel used: charcoal INT kg/day Baseline: fuel used: LPG INT kg/day Baseline: fuel used: fuelwood INT kg/day Baseline: fuel used: kerosene INT kg/day Cooking appliances Appliance(s) used I EPC & EPC & EPC & EPC & EPC & EPC & hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 70.0 70 70 63.3 63 63 Electricity inputs Source of electricity I Grid Grid Grid Grid Grid Grid AC or DC cooking I AC DC Hybrid AC DC Hybrid Fuel prices Electricity tariffs I $/kWh 0.01 0.01 0.01 0.01 0.01 0.01 Charcoal I $/kg (Low, High) LPG I $/kg (Low, High) Firewood I $/kg (Low, High) Kerosene I $/kg (Low, High) PV sizing Lifetime I Years Energy from solar irradiance I KWh/KWpeak PV performance decay (and oversize factor) I % over life Cooking directly from PV I % Design peak power rating INT Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels O $ Charge controller Lifetime I Years 6 6 6 6 Charge controller capacity (incl. safety factors) INT Amps 6 6 6 6 Battery charging roundtrip efficiency I % (Low, High) 90, 85 90, 85 90, 85 90, 85 Purchase cost of controller O $ 9.22 9.22 8.34 8.34 Purchase cost of additional Balance of System O $ 67.8, 83.2 25.2, 30.6 44.6, 53.7 17.3, 20.3 Battery Battery voltage I V 24.0 24.0 24.0 24.0 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 Useable max capacity remaining at replacement I % (Low, High) 90.0, 80.0 90.0, 80.0 90.0, 80.0 90.0, 80.0 Capacity addition to account for decay I % 10.0 10.0 10.0 10.0 Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles (Low, 3000, 3000, 3000, 3000, High) 2000 2000 2000 2000 Battery capacity INT kWh 4.6 1.52 4.6 1.52 Unit price (delivered) I $/kWh (Low, 280.00 280.00 180.00 180.00 High) Purchase cost of battery O $ (Low, High) 1276.6, 425.5, 820.7, 273.6, 1585.7 531.9 1003.0 334.3 Total system Initial purchase cost O $ (Low, High) 70.00 1423.6, 530.0, 63.27 936.9, 362.5, 1748.2 641.7 1128.3 426.2 151 Generic (as used for 152 cases modelled in Everywhere Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 Figure 3.36) Cooking Description 100% electric solutions scenario DIRECT OR BATTERY 100% AC electric 100% AC electric 100% DC battery- electric 100% DC battery- electric 100% DC battery- electric 100% DC battery- electric 50% AC, 50% Batt DC 50% AC, 50% Batt DC 50% AC, 50% Batt DC 50% AC, 50% Batt DC 100% AC electric 100% AC electric 100% AC electric 100% AC electric 100% DC battery- electric 100% DC battery- electric 100% DC battery- electric 100% DC battery- electric Low or high Component Parameter Input, INTmediate output, Output Low High High Low High Low High Low High Low High Low High Low High Low High cost Cooking eCook electricity use for cooking I kWh/day 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 0.87 2.06 Fuel stacked with eCook: charcoal I kg/day Fuel stacked with eCook: LPG I kg/day Fuel stacked with eCook: fuelwood I kg/day Fuel stacked with eCook: kerosene I kg/day Baseline: fuel used: charcoal I kg/day Baseline: fuel used: LPG I kg/day Baseline: fuel used: fuelwood I kg/day Baseline: fuel used: kerosene I kg/day Cooking appliances Appliance(s) used I EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & EPC & hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate hotplate Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 70.0 70.0 70.0 70.0 63.3 63.3 70.0 70.0 63.3 63.3 70.0 70.0 63.3 63.3 70.0 70.0 63.3 63.3 Electricity inputs Source of electricity I Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid PV PV PV PV AC or DC cooking I AC AC DC DC DC DC Hybrid Hybrid Hybrid Hybrid DC DC DC DC DC DC DC DC Fuel prices Electricity tariffs I $/kWh 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.55 0.85 0.25 0.38 Charcoal I $/kg LPG I $/kg Fuelwood I $/kg Kerosene I $/kg PV sizing Lifetime I Years 20.00 20.00 20.00 20.00 Energy from solar irradiance I KWh/KWpeak 4.30 3.68 4.30 3.68 PV performance decay (and oversize factor) I % over life 10 10 10 10 Cooking directly from PV I 20 20 20 20 Design peak power rating I Wpeak 1 1 1 1 Unit price (delivered) I $/Wpeak 0.42 0.42 0.35 0.35 Purchase cost of PV panels O $ 264.46 280.01 221.95 235.01 Charge controller Lifetime I Years 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 Charge controller capacity (incl. safety I Amps 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 45.0 45.0 45.0 45.0 factors) Battery charging roundtrip efficiency I 90 85 90 85 90 85 90 85 90 85 90 85 Purchase cost of controller O $ 9.2 9.2 8.3 8.3 9.2 9.2 8.3 8.3 154.0 154.0 139.2 139.2 Battery Battery voltage I V 24 24 24 24 24 24 24 24 24 24 24 24 Battery minimum depth of charge I 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Useable max capacity remaining at I 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 replacement Capacity addition to account for decay I 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Effficiency loss in wiring I 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles 3000 2000 3000 2000 3000 2000 3000 2000 3000 2000 3000 2000 Battery capacity NT kWh Unit price (delivered) I $/kWh 280 350 180 220 280 350 180 220 280 350 180 220 Purchase cost of battery O $ 777.80 972.30 500.00 611.10 259.00 324.00 75.66 92.48 622.26 777.82 400.02 488.92 Total system Initial purchase cost O $ (Low, High) 857.0 1051.5 571.6 682.7 338.2 403.2 147.3 164.1 70.0 70.0 63.3 63.3 1110.8 1281.9 824.5 926.4 ESMAP  |  Cooking with Electricity: A Cost Perspective Generic (as used for cases Everywhere Installation year 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 modelled in Figure 3.36) Cooking Description Clean fuel stack scenario 50% AC electric 50% AC electric 50% DC battery- electric 50% DC battery- electric 50% DC battery- electric 50% DC battery- electric 50% AC electric 50% AC electric 50% AC electric 50% AC electric 50% DC battery- electric 50% DC battery- electric 50% DC battery- electric 50% DC battery- electric 100% LPG 100% LPG 100% charcoal 100% charcoal 100% firewood 100% firewood 100% LPG 100% LPG 100% charcoal 100% charcoal 100% firewood 100% firewood Low or high Component Parameter cost Appendix F: Model Input Data Input, INT mediate, Output Low High Low High Low High Low High Low High Low High Low High Low High Low High Low High Low High Low High Low High Cooking eCook electricity use for cooking I kWh/day 0.29 0.68 0.29 0.68 0.29 0.68 0.29 0.68 0.29 0.68 0.29 0.68 0.29 0.68 Fuel stacked with eCook: charcoal I kg/day Fuel stacked with eCook: LPG I kg/day 0.07 0.16 0.07 0.16 0.07 0.16 0.07 0.16 0.07 0.16 0.07 0.16 0.07 0.16 Fuel stacked with eCook: fuelwood I kg/day Fuel stacked with eCook: kerosene I kg/day Baseline: fuel used: charcoal I kg/day 1.10 2.61 1.10 2.61 Baseline: fuel used: LPG I kg/day 0.14 0.33 0.14 0.33 Baseline: fuel used: fuelwood I kg/day 2.12 5.02 2.12 5.02 Baseline: fuel used: kerosene I kg/day Cooking Appliance(s) used I EPC EPC EPC EPC EPC EPC EPC EPC EPC EPC EPC EPC EPC appliances Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 50.0 50.0 50.0 45.0 45.0 50.0 50.0 45.0 45.0 50.0 50.0 45.0 45.0 Electricity Source of electricity I Grid Grid Grid Grid Grid Grid Grid Grid Grid PV PV PV PV inputs AC or DC cooking I AC DC DC DC DC DC DC DC DC DC DC DC DC Fuel prices Electricity tariffs I $/kWh 0.04 0.25 0.04 0.25 0.04 0.25 0.55 0.85 0.25 0.38 Charcoal I $/kg 0.13 0.49 LPG I $/kg 1.08 2.07 1.08 2.07 1.25 2.40 1.08 2.07 1.25 2.40 1.08 2.07 1.25 2.40 1.08 2.07 1.25 2.40 Firewood I $/kg 0.04 0.13 Kerosene I $/kg PV sizing Lifetime I Years 20.00 21.00 20.00 21.00 Energy from solar irradiance I KWh/KWpeak 4.30 3.68 4.30 3.68 PV performance decay (and I % over life 10 10 10 10 oversize factor) Cocking directly from PV I % 20 21 20 21 Design peak power rating I Wpeak 0.63 1 1 0.67 Unit price (delivered) I $/Wpeak 0.42 0.42 0.35 0.35 Purchase cost of PV panels O $ 264.5 280.0 221.9 235.0 Charge Lifetime I Years 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 controller Charge controller capacity I Amps 6.0 6.0 6.0 6.0 45.0 45.0 45.0 45.0 (incl. safety factors) Battery charging round trip efficiency I % 90 85 90 85 90 85 90 85 Purchase cost of controller O $ 9.2 9.2 8.3 8.3 154.0 154.0 139.2 139.2 Battery Battery voltage I V 24 24 24 24 24 24.0 24 24 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Useable man capacity remaining at I % 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 replacement Capacity addition to account for decay I % 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Efficiency loss in wiring I  % 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles 3000 2000 3000 2000 3000 2000 3000 2000 Battery capacity NT kWh Unit price (delivered) I $/kWh 280 350 180 220 280 350 180 220 Purchase cost of battery O $ 777.80 972.30 500.00 611.10 622.3 777.8 400.0 488.9 Total system Initial purchase cost O $ (Low, High) 153 154 Generic (as used for Installation cases modelled in Everywhere 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 2025 year Figure 3.37 Cooking scenario Description Cooking efficiency DIRECT OR electric electric 100% DC 100% DC 100% DC 100% DC 100% DC 100% DC heavy foods heavy foods BATTERY foods only AC foods only AC only AC electric only AC electric 50% AC electric 50% AC electric 50% AC electric 50% AC electric battery- electric battery- electric battery- electric battery- electric battery- electric battery- electric 100% AC electric 100% AC electric 100% AC electric 100% AC electric 100% AC electric 100% AC electric Low or high Component Parameter Low Low Low Low Low Low Low Low Low Low High High High High High High High High High High cost case Input, INTmediate output, Output Cooking eCook electricity use for cooking I kWh/day 3.15 3.15 1.97 1.97 1.48 1.48 0.49 0.49 0.15 0.15 3.15 3.15 1.97 1.97 1.48 1.48 0.49 0.49 0.15 0.15 Fuel stacked with eCook: charcoal I kg/day Fuel stacked with eCook: LPG I kg/day Fuel stacked with eCook: fuelwood I kg/day Fuel stacked with eCook: kerosene I kg/day Baseline: fuel used: charcoal I kg/day Baseline: fuel used: LPG I kg/day Baseline: fuel used: fuelwood I kg/day Baseline: fuel used: kerosene I kg/day Cooking appliances Appliance(s) used I 4 plate & 4 plate & Hotplate, Hotplate, 1 insu- 1 insu- EPC EPC EPC EPC 4 plate & 4 plate & Hotplate, Hotplate, 1 insu- 1 insu- EPC EPC EPC EPC oven oven induction induction lated, 1 lated, 1 oven oven induction induction lated, 1 lated, 1 or infra- or infra- uninsu- uninsu- or infra- or infra- uninsu- uninsu- red red lated lated red red lated lated Lifetime of cooking appliances I Years 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Purchase cost of cooking appliances I $ 135.6 135.6 18.1 18.1 63.3 63.3 45.2 45.2 45.2 45.2 135.6 135.6 18.1 18.1 63.3 63.3 45.2 45.2 45.2 45.2 Electricity inputs Source of electricity I Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid Grid AC or DC cooking I AC AC AC AC AC AC AC AC AC AC DC DC DC DC DC DC DC DC DC DC Fuel prices Electricity tariffs I $/kWh 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 0.04 0.25 Charcoal I $/kg LPG I $/kg Fuelwood I $/kg Kerosene I $/kg PV sizing Lifetime I Years Energy from solar irradiance I KWh/KWpeak PV performance decay (and oversize I % over life factor) Cooking directly from PV I % Design peak power rating I Wpeak Unit price (delivered) I $/Wpeak Purchase cost of PV panels O $ Charge controller Lifetime I Years 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 Charge controller capacity (incl. I Amps 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 safety factors) Battery charging roundtrip efficiency I % 90 85 90 85 90 85 90 85 90 85 Purchase cost of controller O $ 9.2 9.2 9.2 9.2 9.2 9.2 9.2 9.2 9.2 9.2 Battery Battery voltage I V 24 24.0 24 24 24 24 24 24 24 24 Battery minimum depth of charge I % 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Useable max capacity remaining at I % 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 80.0 90.0 replacement Capacity addition to account for I % 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 decay Efficiency loss in wiring I % 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Cycle lifetime of battery INT Cycles 3000 2000 3000 2000 3000 2000 3000 2000 3000 2000 Battery capacity NT kWh Unit price (delivered) I $/kWh 280 350 280 350 280 350 280 350 280 350 Purchase cost of battery O $ 777.8 972.30 777.80 972.30 777.80 972.30 777.80 972.30 777.80 972.30 Total system Initial purchase cost O $ (Low, High) 18.1 18.1 63.3 63.3 45.2 45.2 45.2 45.2 922.6 1117.1 805.1 999.6 850.3 1044.8 832.2 1026.7 832.2 1026.7 ESMAP  |  Cooking with Electricity: A Cost Perspective APPENDIX G SOLAR ECOOKING CROSS-COMPARISON TABLE G.1 Key parameters and assumptions for solar eCooking techno-economic models FINANCING HORIZON (YEARS) ELECTRICITY DEMAND (KWH) 100 PERCENT OF COOKING? INITIAL CAPITAL MONTHLY CAPABLE OF COVERING HOUSEHOLD COOKING (NUMBER OF PEOPLE) COST EXPENDITURE KEY ASSUMPTIONS BATTERY STORAGE (S) (S) HOUSEHOLD SIZE MODEL YEAR APPLIANCE MONTHLY PV (WPEAK) SOURCE AC/DC HIGH HIGH LOW LOW 2020 5 DC EPC and Yes 4.2 1.92 2.2 kWh 630 1,163 1,343 25 29 Assumes 20 percent of hot plate LiFePO4 energy direct from PV, 0 days autonomy Case Study 5 2020 5 DC EPC and Yes 4.2 0.64 0.74 kWh 220 522 582 13 17 LPG LiFePO4 2025 5 DC EPC and Yes 4.2 1.92 2.2 kWh 630 863 970 19 21 hot plate LiFePO4 2025 5 DC EPC and Yes 4.2 0.64 0.74 kWh 220 409 445 11 15 LPG LiFePO4 2019 3 DC Hot plate Yes 5 1.2 1.5kWh 400 1,526 1,799 49 59 0 days autonomy, 2hours lithium-ion cooking/day Beyond Fire: ECooking 2019 3 DC Induction Yes 5 1 1.2kWh 300 1,390 1,635 44 52 lithium-ion (2019) 2019 3 DC Slow Unlikely 5 0.36 0.45kWh 100 491 572 16 20 Assumes slow cooker can cooker li-on cook all dishes in 2hours/day, 0 days autonomy 2019 3 DC EPC Unlikely 5 0.22Wh 0.36kWh 80 600 681 20 23 Assumes EPC can cook all lithium-ion dishes and only opened twice/ day, 0 days autonomy 2020 10 DC EPC Unlikely 6 0.6 2.1kWh 420 2,266 2,266 19 19 Assumes EPC can cook all LiFePO4 dishes, 2 days of autonomy, 2025 12 DC EPC Unlikely 6 0.6 2.1kWh 420 1,926 1,926 14 14 only cooking lunch and dinner, Zubi et al. LiFePO4 also powering lights and (2017) charging phone 2030 14 DC EPC Unlikely 6 0.6 2.1kWh 420 1,644 1,644 11 11 LiFePO4 2035 15 DC EPC Unlikely 6 0.6 2.1kWh 420 1,426 1,426 9 9 LiFePO4 2016 20 AC Hot plate Yes 5 5.45 Not stated Not 1,032 6,202 72 162 Based on Leach and Oduro Beyond Fire stated (2015)—calculates cost for (2016) electricity from SHS sized for 2016 20 AC Induction Yes 5 4.25 Not stated Not 1,008 6,060 56 126 cooking (0.40–0.90 EUR/kWh) stated by est. demand. 2015 20 AC Hot plate Yes 4 1.4 – 2.2–9.8 kWh 367– 1,032 6,202 10 162 AC appliances, 0 days 4.2kWh LiFePO4 1331W autonomy, wide range of Leach and Oduro (2015) 2025 20 AC Hot plate Yes 4 1.4 – 2.2–8.7 kWh 367– 718 3,550 7 70 values with high/low cooking 4.2kWh LiFePO4 1331W energy demand and optimistic/ pessimistic techno-economic scenarios Appendix G: Solar eCooking Cross-Comparison 155 APPENDIX H MULTI-TIER FRAMEWORK TABLE H.1 Multi-tier framework for measuring household electricity access ATTRIBUTES TIER O TIER 1 TIER 2 TIER 3B TIER 4 TIER 5 Capacity Power capacity Less than At least At least At least 200 W At least At least ratings 3 W 3 W 50 W 800 W 2 kW (W or daily Wh) Less than At least At least At least 1 kWh At least At least 12 Wh 12 Wh 200 Wh 3.4 kWh 8.2 kWh Services Lighting of Electrical 1,000 Imhr lighting, air per day circulation, television, and phone charging are possible Availabilitya Daily Less than At least 4 hours At least 8 hours At least At least Availability 4 hours 16 hours 23 hours Evening Less than At least At least At least 3 hours At least 4 hours Availability 1 hour 1 hour 2 hours Reliability More than 14 disruptions per week At most (> 3 to 14 At most 3 14 disruptions disruptions disruptions per week / week) or per week or At most 3 disruptions with total 3 disruptions per / week with duration of week with total > 2 hours of less than duration more outage 2 hours than 2 hours Quality Household experiences voltage problems that damage Voltage problems do not appliances affect the use of desired appliances Affordability Cost of a standard consumption package Cost of a standard consumption package of of 365 kWh per year is more than 5% of 365 kWh per year is less than 5% of household household income income Formality No bill payments made for the use of electricity Bill is paid to the utility, prepaid card seller, or authorized representative Health and Serious or fatal accidents due to electricity connection Absence of past accidents Safety a. Previously referred to as “Duration” in the 2015 Beyond Connections report, this MTF attribute is now referred to as “Availability,” examining access to electricity through levels of “Duration’ (day and evening). Aggregate tier is based on lowest tier value across all attributes * Color signifies tier categorization. Source: ESMAP (2015). 156 ESMAP  |  Cooking with Electricity: A Cost Perspective