Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Sudeshna Ghosh Banerjee Kabir Malik Andrew Tipping Juliette Besnard and John Nash © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. 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 govern- ments 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. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Sudeshna Ghosh Banerjee Kabir Malik Andrew Tipping Juliette Besnard and John Nash Table of Contents Foreword.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Abbreviations and Acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Chapter 1. Agriculture and Power Nexus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 High Potential for Agricultural Transformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Past Performance: A Missed Opportunity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Investment Funding Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 An Improving Outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Major Approaches to Agricultural Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Cluster Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Smallholder Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Agricultural Growth to Raise Rural Welfare: Reasons for Optimism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Rural Electrification Has Lagged Behind. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Agriculture as an Anchor Load for Rural Electrification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Study Purpose and Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Chapter 2. Power Needs of Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Power Needs across the Agriculture Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Irrigation Potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Primary and Secondary Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Aggregate Electricity Demand from Irrigation and Processing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Chapter 3. Power Needs in Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Selection of Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Electricity Demand and Farming Scale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Electricity Demand in the Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 iv Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Chapter 4. Lessons from Ongoing Power-Agriculture Integration Projects. . . . . . . . . . . . . . . . . . . . . . . . 38 Case Study 1. Tanzania: Sumbawanga Agriculture Cluster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Financial Viability: Extension of Main Grid from Mbeya to Sumbawanga and Rukwa. . . . . . . . . . . 42 Economic Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Case Study 2. Tanzania: Mwenga Mini-Hydro Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Power Supply Options, Commercial Arrangements, and Financial Analysis. . . . . . . . . . . . . . . . . . . . 45 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Case Study 3. Zambia: Mkushi Farming Block.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Case Study 4. Zambia: Mwomboshi Irrigation Development and Support Project. . . . . . . . . . . . . . . . . 51 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Case Study 5. Kenya: Oserian Flowers and Geothermal Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Financial Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Case Study 6. Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids. . . . . . . . . . . . . . . . 58 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Power Supply Options.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Key Conclusions from the Case Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Large Power Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Supply Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Table of Contents v Financial and Economic Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Financing of Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Chapter 5. Opportunities to Harness Agriculture Load for Rural Electrification. . . . . . . . . . . . . . . . . . 64 Simulation of Power Demand in a Stylized Agricultural Setting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Simulation Study 1. Ethiopia: Power Generation from Sugar Estates.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Simulation Study 2. Mali: Mini-Grid Expansion for Productive Users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Power Demand from Mini-Grids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Main Inferences and Institutional Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Chapter 6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Key Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Overall Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Case Study Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Recommended Actions to Promote Power-Agriculture Integration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Improve Institutional Coordination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Integrate Planning of Power, Agriculture, and Rural Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Promote Farmers’ Productivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Annexes A: Business Models for Agricultural Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 B: Agriculture Fuels for Power Generation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 C: Description of Processing Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 D: Maps of Case Study Project Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 vi Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Tables ES.1 Summary of Ongoing or Planned Cases of Power-Agriculture Integration. . . . . . . . . . . . . . . . . xvii ES.2 Summary of Simulated Cases of Power-Agriculture Integration.. . . . . . . . . . . . . . . . . . . . . . . . . . . . xix 2.1 Power Demand for Irrigation, by System Type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.2 Potential Investment Needs for Large-Scale, Dam-Based and Complementary Small-Scale Irrigation Schemes in Sub-Saharan Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3 Key Power-Intensive Agribusiness Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.4 Method for Calculating Power Demand from Irrigation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.5 Power Demand for Crop Processing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.1 Analysis of Commodity Value Chains, by Scale and Region/Country. . . . . . . . . . . . . . . . . . . . . . . 24 3.2 Comparison of Historical and Projected Commodity Growth Rates and Estimated Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.3 Countries in Sub-Saharan Africa with Similar Commodity Production and Processing Systems.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.4 Power Demand for Standard 300 ha Cultivated Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.1 Sumbawanga Agriculture Cluster at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.2 Sumbawanga Geographic and Demographic Features.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3 Total Power Demand from Agriculture by 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.4 Residential and Commercial Data to Calculate Commercial Power Demand. . . . . . . . . . . . . . . . 41 4.5 Estimated Capital and Operating Costs for Transmission and Distribution Expansion.. . . . . . . 43 4.6 Estimated Power Consumption and Transmission and Distribution Tariff Requirement. . . . . . 43 4.7 Financial Present Value of Grid Extension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.8 Economic Costs and Benefits of Sumbawanga Grid Extension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.9 Mwenga Mini-Hydro Mini-Grid at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.10 Estimated Power Demand from Mwenga Mini-Hydro Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.11 Economic Costs and Benefits of Mwenga Mini-Hydro Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 4.12 Mkushi Farming Block at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.13 Power Requirements for Irrigation and Milling in the Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.14 Electrification Rates and Power Load of Households in Mkushi Farm Block. . . . . . . . . . . . . . . . . 48 4.15 Financial Analysis of Mkushi Farming Block from the Perspective of the Utility and a Representative Farmer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.16 Net Social Benefits of Grid Extension, Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.17 Economic Costs and Benefits of Grid Extension, Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.18 Mwomboshi Irrigation Development and Support Project at a Glance.. . . . . . . . . . . . . . . . . . . . . . 51 4.19 Irrigation Power Requirements in Mwomboshi, Zambia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Table of Contents vii 4.20 Milling Power Requirements in Mwomboshi, Zambia.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.21 Financial Analysis, Mwomboshi.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.22 Economic Costs and Benefits of the IDSP Project, Mwomboshi. . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.23 Oserian Flowers and Geothermal Power Project at a Glance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.24 Financial Analysis, ODCL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 4.25 Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids at a Glance. . . . . . . . . . . . 58 4.26 Typical LCOE Values for Small-Scale Generation and Distribution Systems.. . . . . . . . . . . . . . . . 62 5.1 Assumptions for Typical Area/Agricultural Activity/Power Demand Model.. . . . . . . . . . . . . . . . . 65 5.2 Ethiopia: Power Generation from Sugar Estates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.3 Total Power Demand from Agriculture and Residential/Commercial Loads. . . . . . . . . . . . . . . . . . 67 5.4 Sugar Factory Power Generation in Years 1 and 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.5 Net Power Generation from Sugar Factory by Year 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.6 Capital Cost Assumptions for Grid Connection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.7 Financial Analysis from the Utility’s Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.8 Sugar Estate Capital Costs, Assumptions for Production Costs, and Revenues.. . . . . . . . . . . . . 69 5.9 Net Economic Benefits of Grid Extension to the Sugar Estate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 5.10 Economic Net Present Value of Extending the Grid to the Sugar Estate. . . . . . . . . . . . . . . . . . . . . 71 5.11 Mali Mini-Grid Expansion for Productive Users at a Glance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.12 Potential Addition of Small Agro-Industrial Activities and Other Businesses. . . . . . . . . . . . . . . . 74 5.13 Current Financial Situation of Koury Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.14 Financial Analysis of Capacity Expansion of Koury Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Figures ES.1 Energy Intensive Activities across Agriculture Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv ES.2 Estimated Power Demand from Agriculture in 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv 1.1 Historical Performance in Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Land and Water Resources Potential in Sub-Saharan Africa.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 B1.3.1 Changes in Irrigation Revenues from Climate Change, 2015–50 (present value). . . . . . . . . . . . . 8 1.3 Projected Value of Food Markets in Sub-Saharan Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.4 Electricity as a Constraint to Food-sector Development in Sub-Saharan Africa. . . . . . . . . . . . . . 9 1.5 Electrification Rate, by Developing Region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.1 Power Needs across Agriculture Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2 Potential New or Rehabilitated Irrigable Land in Sub-Saharan Africa.. . . . . . . . . . . . . . . . . . . . . . . 18 2.3 Estimated Electricity Demand (MW) from Agriculture for Sub-Saharan Africa in 2030. . . . 21 3.1 Potential Peak Capacity and Energy Demand for Large- and Small-scale Systems. . . . . . . . . . 29 viii Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa 3.2a Electricity Input in the Maize Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2b Electricity Input in the Rice Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2c Electricity Input in the Cassava Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2d Electricity Input in the Wheat Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2e Electricity Input in the Soybean Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2f Electricity Input in the Pineapple Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2g Electricity Input in the Sugarcane Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2h Electricity Input in the Oil Palm Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2i Electricity Input in the Dairy Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.2j Electricity Input in the Poultry Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.2k Electricity Input in the Tea Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.2l Electricity Input in the Floriculture (roses) Value Chain.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.2m Electricity Input in the Cotton (lint) Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.3 Potential Power Demand in 2030 from Processing for Small-scale Agriculture, by Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.1 Estimated Peak Load and Energy Demand, by Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2 Estimated Volume of Crops That May Utilize Electricity for Processing. . . . . . . . . . . . . . . . . . . . . 41 4.3 Comparative Cost of Power Supply Options in Sumbawanga. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.4 Total Peak Load in Mkushi, 1995–2014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.5 Power Demand from Irrigation and Milling in Mkushi, 1995–2014. . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.6 Mwomboshi IDSP Plot Sites Developed for Small-scale Farmers. . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.7 Mwomboshi Peak Load and Power Consumption Forecast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.8 Residential and Commercial Demand, Electrification Rate 2016–2031. . . . . . . . . . . . . . . . . . . . . 53 4.9 Power Uses and Sources at ODCL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.10 Output of ODCL’s Power Plants and Expected Increased Output. . . . . . . . . . . . . . . . . . . . . . . . . . 57 4.11 Electricity Output of Capacity Expansion Project and Intended Uses. . . . . . . . . . . . . . . . . . . . . . . 57 4.12 KTDA’s North Mathioya Hydropower Project: Financial Benefits and Power Sold. . . . . . . . . . 59 5.1 Power Demand and Breakdown for a Given Area Radius. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.2 Sensitivity of Power Load to Changes in Percent of Commercial Irrigation. . . . . . . . . . . . . . . . . . 66 5.3 Estimated Energy Demand and Peak Load, by Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.4 Net Social Benefits of Grid Extension to Sugar Estate (years 1–20).. . . . . . . . . . . . . . . . . . . . . . . . . 71 5.5 Koury Mini-grid: Electricity Consumption Patterns.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.6 Energy Generation Profile at Koury Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.7 Koury Mini-grid Profile: Additional Commercial and Industrial Loads. . . . . . . . . . . . . . . . . . . . . . . 75 5.8 Operating Expense and Capital Expenditure Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 A.1 Example of an Agribusiness Cluster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Table of Contents ix Boxes 1.1 Terminology Clarification: Agriculture and Agribusiness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Africa’s Vision for Agriculture: CAADP Goals.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Making Africa’s Power and Water Infrastructure Climate Resilient. . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.1 Farm Type Definitions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.2 Palm Oil and Power Integration in Uganda.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 5.1 Isolated Mini-Grid Systems in Mali: Existing and Potential Power Demand.. . . . . . . . . . . . . . . . . 72 5.2 Large-Scale Opportunities for Power-Agriculture Integration in Mali. . . . . . . . . . . . . . . . . . . . . . . 77 Maps D.1 Tanzania: Power and Agriculture in the Sumbawanga Agriculture Cluster.. . . . . . . . . . . . . . . . . . . 93 D.2 Tanzania: Mwenga Mini-Hydro Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 D.3 Zambia: Mkushi Farming Block. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 D.4 Zambia: Mwomboshi Irrigation Development and Support Project. . . . . . . . . . . . . . . . . . . . . . . . . . 96 D.5 Kenya: Oserian Flowers and Harnessing Geothermal Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 D.6 Kenya Tea Development Agency Holdings Mini-Hydro Mini-Grids. . . . . . . . . . . . . . . . . . . . . . . . . 98 D.7 Ethiopia: Sugar Estates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 D.8 Mali: Power Network and Agricultural Districts.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Foreword The greatest challenge to increasing electricity access in Sub-Saharan Africa is how to make electricity provision financially viable in low-demand rural households. The presence of commercially attractive customers—typically those that have relatively large and stable electricity demand for revenue generating purposes—can help reduce the barriers to accelerating grid and off-grid approaches to rural electrification. By aggregating anchor-load demand with that of households and businesses, it may be possible to extend the grid or create opportunities for mini-grids and other decentralized options. African agriculture has tremendous potential to raise rural welfare through agricultural transformation. It is estimated that productivity growth in agriculture—which predominates the livelihoods of the subconti- nent’s rural poor—could be several times more effective than growth in other sectors in reducing rural poverty. Furthermore, there is a growing commitment among African governments toward sustainable and inclusive agricultural development. Developing energy intensive agricultural processes, such as large-scale irrigation or milling activities, can not only increase agricultural productivity, but can also increase the commercial viability of electricity provi- sion. The large-farm, agribusiness model practiced over the past 20 years has a continuing strategic role to play in promoting growth in Africa. At the same time, subsistence, smallholder farms, which account for most of Sub-Saharan Africa’s agriculture, are key to stimulating the rural economy and uplifting the poor. Energy, along with investments in other complementary infrastructure and services (e.g., roads, transport links to markets, and access to finance), is a critical input for supporting Africa’s agricultural transformation. Without access to affordable and reliable electricity, farmers will continue to face constraints to productivity growth and thus lag behind their counterparts in more prosperous regions of the developing world. Against this backdrop, this study explores opportunities for synergy between the goals of rural elec- trification and agricultural transformation in Sub-Saharan Africa. It shows that leveraging complementary investments in agriculture and electricity can yield double dividends in terms of poverty alleviation. Aligning electricity investments with agricultural development can maximize joint benefits from the expansion of rural electricity access and the increase in value added along the agricultural value chains, both of which are directly linked to improved quality of life and poverty alleviation in rural communities. Lucio Monari Ethel Sennhauser Director Director Energy and Extractives Global Practice Agriculture Global Practice The World Bank The World Bank Acknowledgments The core team for this study included Sudeshna Ghosh Banerjee, Kabir Malik, Juliette Besnard, and John Nash. The team benefited from the background report prepared by Economic Consulting Associates (ECA) and Prorustica, a consulting consortium led by Andrew Tipping and Peter Robinson. The team wishes to thank expert consultants Douglas Barnes and Subodh Mathur, who provided valuable inputs at various stages of the study. The team is grateful to Olivier Dubois and Alessandro Flammini from the Food and Agriculture Organization (FAO) for their inputs and review. The team is appreciative of the overall guidance provided by Lucio Monari and Meike van Ginneken, man- agers in the World Bank’s Africa Energy Group. The team also wishes to thank peer reviewers Vivien Foster, Malcolm Cosgrove-Davies, Dana Rysankova, Holger Kray, Melissa Williams, and Katie Kennedy Freeman for their valuable advice and constructive inputs. The team thanks Norma Adams for editing the report. Finally, the team gratefully acknowledges the funding provided by the Africa Renewable Energy Access (AFREA) program and the Energy Sector Management Assistance Program (ESMAP). Abbreviations and Acronyms ABC Anchor Business Community AMADER Malian Agency for Development of Household Energy and Rural Electrification CAADP Comprehensive Africa Agriculture Development Programme CAGR compound annual growth rate CHP combined heat and power CSR corporate social responsibility CTC cutting, tearing, and curling DRC Democratic Republic of the Congo ECA Economic Consulting Associates EEPCO Ethiopian Electric Power Corporation EFB empty fruit bunch ESC Ethiopian Sugar Corporation EWURA Energy and Water Utilities Regulatory Authority FFB fresh fruit bunch FiT feed-in tariff GDP gross domestic product GP global practice GTAP Global Trade Analysis Project IPP independent power producer IRR internal rate of return KPLC Kenya Power and Lighting Company KTDA Kenya Tea Development Agency LCOE levelized cost of electricity MSMEs micro-, small-, and medium-sized enterprises NPV net present value ODA official development assistance PV photovoltaic REA Rural Energy Agency (Tanzania) RVE Rift Valley Energy SAGCOT Southern Agricultural Growth Corridor of Tanzania SDG Sustainable Development Goal SE4ALL Sustainable Energy for All SHS solar home system SMEs small- and medium-sized enterprises SSA Sub-Saharan Africa TANESCO Tanzania Electric Supply Company Limited ZESCO Zambia Electricity Supply Corporation Executive Summary Increasing access to modern electricity services in and commercial power demand could increase the Sub-Saharan Africa is one of the main develop- feasibility of extending the grid or creating opportu- ment challenges facing the world over the next two nities for independent power producers and mini-grid decades. Inclusion of the target to “ensure access to operators. Drawing on a suite of case studies, this affordable, reliable, sustainable, and modern energy study offers insights on what it would take to opera- for all” in the Sustainable Development Goals (goal 7) tionalize the opportunities and address the challenges has brought a sharper focus to accelerating electric- for power-agriculture integration in Africa. ity access in the historically underserved regions of the world—most notably Sub-Saharan Africa. Two out of every three people in Sub-Saharan Africa live What is the scale of opportunity without electricity, a reality that is inconsistent with of power demand from the modern world. The majority of this population agriculture? without access to electricity is rural and poor. Rural electrification efforts in the region have not achieved Historical performance of agriculture in Sub- sufficient progress in increasing electricity access as Saharan Africa has been wanting. The share of these areas are typically commercially unattractive, agriculture in GDP has declined from 20 percent in characterized by sparsely distributed customers 2000 to 14 percent in 2013.1 A very small percent- with low electricity consumption and ability to pay, age of Africa’s agricultural production undergoes and a high cost of service to extend the grid. Rural industrial processing.2 In high-income countries, enterprises and households thus must cope without processing adds about US$180 of value per ton of electricity, relying instead on expensive, poor quality agricultural produce, compared to only $40 in Sub- backups (e.g., diesel, kerosene or other oil-based Saharan Africa; this disparity is aligned with the small sources), thereby stunting productivity, limiting size of Sub-Saharan Africa’s agribusiness sector rela- development outcomes, and imposing harmful tive to on-farm agriculture. In addition, for more than environmental impacts. The rural economies are four decades, the region’s share in global agricultural overwhelmingly dependent on agriculture; in fact, export markets has been on the decline. agriculture and agribusiness comprise nearly half of There are reasons to believe that agriculture Africa’s gross domestic product (GDP). These enter- productivity could turn the tide. Trends in economic prises require electricity to grow to their potential, growth and urbanization fuel the demand for food, while the expansion of rural energy services needs as do continuing improvements in infrastructure and consumers with consistent power needs to serve as a the benefits of lower oil prices. The potential urban reliable revenue source. market for agricultural goods and commodities is Can agriculture and energy come together in projected to reach US$1 trillion by 2030. There are Sub-Saharan Africa to offer a double dividend with a number of underlying structural incentives to pro- benefits to enterprises, households, utilities, and mote agriculture. The region has 45 percent of the private-sector service providers? This is the central world’s total suitable land area for expanding sustain- question of this study. That is, can energy intensive able agricultural production. Past gains in commercial activities along the agriculture value chains pro- crops (e.g., cashews, tea, and sesame seeds) indicate vide significant revenues to the power utilities and that the region can increase its agricultural pro- increase the financial viability of rural electrification? ductivity. But seizing this opportunity will require Combining agricultural load with other household farmers and agribusinesses to ramp up production xiv Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure ES.1: Energy intensive activities Figure ES.2: Estimated power demand across agriculture value chains from agriculture in 2030 8,000 Post-harvest & 7,000 6,915 Secondary On-farm primary processing processing 6,000 5,000 MW • Irrigation • Milling, drying, • Packaging, chilling, etc. bottling, etc. 4,000 3,786 3,000 Rural Urban/peri-urban 2,084 2,000 978 1,000 and develop agriculture value chains to enhance processing, logistics, market infrastructure, and retail 0 networks. Irrigation Processing (milling) Electricity is an important enabler for the 2015 2030 agriculture sector to realize its growth potential, especially for power intensive value chains. The need for electricity is distributed across the life of the today, to about 9 GW. The estimated incremental crop—from mechanized irrigation to processing for demand between 2015 and 2030 is 4.2 GW (fig- final consumption (figure ES.1). The power demand ure ES.2). Irrigation would provide about 75 percent for irrigation primarily comes from (i) sourcing bulk of agriculture’s demand, with the rest coming from water from a water body (e.g., a dam or river) and agro-processing. The irrigation demand estimate (ii) distributing it over the cultivated area. Bulk water assumes full exploitation of economically viable, pumping is typically the major source of demand and potential areas for new or rehabilitated irrigation depends on the vertical and horizontal distances of development, totaling nearly 6.8 million ha. This the scheme from the water source. Demand from would be dominated by small-scale scheme devel- distribution systems varies by the types of irrigation opment in the Gulf of Guinea and rehabilitation of system, which range in scale from manual to surface existing schemes in the Sudano-Sahelian region. flooding and localized ones to center pivots. Post- The agro-processing demand estimate is based on harvest and primary processing (e.g., milling and the electricity requirement for a typical processing drying) and secondary processing (e.g., packaging activity (milling), and thus does not capture demand and bottling) represent a growth area. It is clear that from the potential development of other processing milling is likely to increase significantly owing to the activities or storage. expected demand growth for such grains as maize, For 13 major agriculture value chains, electricity wheat, and rice. Similarly, such staples as cassava demand could increase by 2 GW (from 3.9 GW in are expected to experience increased demand for 2013 to 6 GW in 2030). This represents nearly half processing due to their perishable nature and use of the 4.2 GW of potential increase in electricity as an industrial input in the manufacture of other demand from agriculture calculated for Sub-Saharan products (e.g., glue in the case of cassava). Creating Africa. The 13 products are maize, rice, cassava, opportunities to piggyback viable rural electrification wheat, oilseed (soybean), horticulture (pineapple), onto local agricultural development will depend on sugarcane, oil palm, dairy, poultry, tea, floriculture the scale and profitability of agricultural operations, (roses), and cotton (lint). These were selected on crops, terrain, types of processing activity, and other the basis of their nature and magnitude of power site-specific local conditions. use for irrigation and processing, growth potential, By 2030, the region’s electricity demand from and ability to serve as significant loads for electricity agriculture is estimated to double from its level systems. Of the value chains studied, the per-hectare Executive Summary xv electricity demand is largest for poultry, because hybrid mini-grid (diesel-solar PV) to serve productive the process is more intensive, using less land for a users (tables ES.1 and ES.2). much larger yield. Other value chains with significant Irrigation is typically the largest source of power per-hectare demand are floriculture (roses), tea, and demand, along with processing activities in specific sugarcane. Together, maize, rice, and cassava add to instances. Irrigation usually has a larger load require- about 83 percent of the total incremental demand in ment than agro-processing activities, especially agriculture processing to 2030. For the 13 commod- in cases of supply to a given area (e.g., Tanzania’s ities analyzed, commercial-scale irrigated farming is Mufindi Tea Estate). Irrigation development and elec- the largest source of electricity demand. Commercial trification can significantly help increase the viability irrigated agriculture, which is highly mechanized, has of rural electrification. Taken alone, the smaller loads the largest potential for developing large power loads of agro-processing activities (e.g., milling and extru- across a range of farm sizes. sion) may not be sufficient to justify rural electrifica- tion investments, except when they provide a viable source of electricity generation (e.g., sugar) or have What are the case study lessons a large and consistent load requirement (e.g., tea). on economic and financial If the volumes of produce can benefit from powered viability? irrigation, supplemented by economies of scale, the load from the production could be significantly larger. This study analyzes eight case studies—six actual Irrigation and processing are often linked. In and two simulated—in five countries of Sub-Saharan many instances, increase in yields from irrigation is Africa; these provide important lessons on the ben- an important prerequisite for the development of efits and risks of large power loads, supply options, large-scale processing activities (as seen in Zambia). and viability. In Tanzania, the first case study is the This cause-and-effect relationship between irriga- Sumbawanga Agriculture Cluster, a concept-stage tion and processing was also observed in the cluster project located in the country’s Southern Agricultural concept (e.g., SAGCOT in Tanzania). Increase in the Growth Corridor of Tanzania (SAGCOT). The scale of processing activity could lead to a significant second case in Tanzania is the successful Mwenga increase in the power demand. Mini-Hydro Mini-Grid Project, which supplies the Successful integration of agriculture and power Mufindi tea estate and surrounding households in system development requires physical and market the country’s Southern Highlands. In Zambia, the infrastructure, which facilitate market access for first case is a grid extension to the ongoing Mkushi inputs and produce. Viable rural electrification relies Farming Block Project, stretching over 176,000 ha on a healthy and profitable agriculture sector. Better of land in the country’s Central Province. The infrastructure and market access improve agriculture second case study in Zambia is the Mwomboshi revenues, spurring further expansion in produc- Irrigation Development and Support Project (IDSP), tion and associated electricity demand. In Zambia, which is developing integrated irrigation agriculture for example, the strategic location of the Mkushi based around a recently built water storage dam farming block along a major international highway on the Mwomboshi River. In Kenya, the first case (T2 Highway and Tazara Railway, which connects examines floriculture development by the Oserian Lusaka and the Copperbelt in Zambia to the port at Development Company Limited (ODCL), a pioneer Dar es Salaam) has improved its development viabil- in using heat from geothermal wells for internal ity. The location of the farming block allows access power generation and consumption. The second case to markets for both inputs and produce. In Tanzania, in Kenya focuses on the Kenya Tea Development the Sumbawanga agriculture cluster benefits from Agency (KTDA) mini-hydro mini-grids. The two access to shared infrastructure and services, including simulated case studies are in Ethiopia and Mali. market access. This helped increase the viability of The Ethiopia study centers on a sugar estate with the agriculture sector as a creditable customer for self-generated power from bagasse and the opportu- electricity suppliers. nity of selling the power surplus to the main grid. The The seasonality of power demand from the Mali study analyzes capacity expansion of an existing agriculture sector can be a significant constraint xvi Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table ES.1: Summary of Ongoing or Planned Cases of Power-Agriculture Integration Project Tanzania Tanzania Zambia Zambia Kenya Kenya Name Sumbawanga Mwenga Mkushi Mwomboshi Oserian Tea Agriculture Mini-Hydro Farming Block Irrigation Flowers and Development Cluster Mini-Grid Development Geothermal Agency and Support Power Holdings Project (IDSP) Mini-Hydro Mini-Grids Overview Expansion A 4 MW hydro Extending a Grid upgrade Expansion of Development of electricity mini-grid transmission and grid the estate’s of hydropower supply to connected to line into a extension geothermal plants support the the main grid. farming area to support generation powering tea development Main local with significant irrigation capacity and factories and of an anchor load agricultural development distribution staff housing agriculture is Mufindi potential. and household network and selling cluster and Tea Estates electrification. to power surplus power surrounding and Coffee the farm’s to the grid. households Limited; 1,300 operations and through main households distribution power grid connected in within the extension. surrounding estate. communities. Commodities Maize, Coffee, tea Wheat, Tobacco, Floriculture Tea sunflower, soybean, wheat, finger millet, tobacco, soya, poultry, maize, paddy, vegetables, sunflower, sorghum coffee horticulture (tomatoes, onions, bananas) Financial The project The financial From a purely Positive With a positive Evaluation Viability is marginally viability of financial point financial NPV, financial NPV, of a sample financially the project of view and as estimated at the planned project, North unviable as a depends a stand-alone US$1.1 million. expansion Mathioya, stand-alone critically on the project, grid project of shows that project. ability to sell extension to 0.4 MW and the project excess power Mkushi was electrification is financially to the main profitable for of 2,000 viable, with grid. Despite the farmers households a NPV of financial but not for the is financially US$3.3 viability, capital utility; sharing viable. million; subsidies were of capital costs revenues provided to was however accrue from keep local an appropriate the sale of electricity and successful power to the tariffs low. approach grid and cost to project savings by tea financing. factories. Executive Summary xvii Project Tanzania Tanzania Zambia Zambia Kenya Kenya Economic Economic Economic Thanks to Positive Positive The same Viability benefits would benefits are households’ economic economic project is be significant positive (US$9 energy cost NPV was benefits were evaluated as (US$134 million); they savings, estimated at estimated economically million), come from increased US$2 million at US$2.5 viable, with justifying households’ yields from for the power million; the a NPV of the project; energy cost irrigation on line extension, main economic US$10 million; they come savings, small-scale mainly from benefit is based direct and mainly from reduced farms, and greater on increased indirect rural households’ reliance on job creation; irrigated household electrification cost savings, diesel backup the project’s tomato electrification impacts small-scale for the tea economic and maize and thus include irrigation estate, and NPV was production. savings due to electrification benefits, and job creation positive lower energy of staff margin uplift from newly (US$46 consumption housing, from market electrified million). costs (e.g., reduced access. businesses. less use of connection kerosene costs for and no more surrounding payment for households, cell phone and charging development services and of stand-alone disposable home systems. batteries). About 30,000 households will benefit from electricity connections. to viability. Large seasonal differences in electricity clusters (e.g., Sumbawanga in Tanzania) can increase dependent agricultural activities will impact the cost the viability of rural electrification. Cluster devel- recovery of electricity supply investments. In such opment has load diversity by design and thus is less cases, it is important to consider ways to mitigate the risky than relying on a single anchor load. If there is a impact of a variable load. One option, especially in the private electricity supplier and private off-takers, any case of mini-grid or captive generation, is the ability to such risk will be priced into the supply contract, thus sell excess power to the grid (e.g., Mwenga mini-hydro increasing the price of electricity for all customers. in Tanzania and KTDA mini-hydro development in In such cases, diversified cluster development can Kenya).3 During the post-harvest season, an increase also help reduce the price of electricity. In some such in the post-harvest processing activity may comple- instances, the public sector can also help mitigate this ment electricity demand from irrigation. In addition, risk through a grid connection and a feed-in tariff irrigation itself may reduce seasonality in agricultural (FiT), subsidies to increase the customer base, or production and thus electricity demand by allowing guarantee/insurance instruments. for multi-cropping (e.g., Mkushi in Zambia). Large-scale development of irrigation-based agri- When considering agricultural anchor loads, culture and sugar estates with excess generation can it is more risky for the investment to depend on a justify a main grid connection on a purely financial single large customer since any negative shock to the basis. Requirements for this—not all of which are read- customer would negatively affect operating reve- ily available in Sub-Saharan Africa—include relatively nues of the electricity supplier. As such, agricultural clear and empty land with good quality soils, a reliable xviii Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table ES.2: Summary of Simulated Cases of Power-Agriculture Integration Ethiopia: Power Generation Mali: Mini-Grid Expansion Project from Sugar Estates for Productive Uses Overview Self-generation of power from bagasse and sale Capacity expansion of an existing hybrid of power surplus to the main grid. mini-grid (diesel-solar PV) to serve productive users. Commodity Sugar Agro-industrial activities Financial Viability From the utility’s perspective, extending the From the perspective of Yeelen Kura, the grid to the sugar estate is not financially viable current financial situation of the Koury since the net present value (NPV) is negative— mini-grid is fragile; however, the capacity because it does not benefit from sales to the expansion project is profitable thanks estate, which self-supplies; however, from to a higher payment rate, additional the standpoint of the sugar estate, it is highly revenues, and proportionally low capital profitable (US$139 million). expenditure and operating expense (NPV of €103,000). Economic Viability The economic NPV for the whole period is The economic NPV for the expansion positive (US$367 million), thus justifying project project is slightly negative (−€18,000) as development. no significant savings are expected from agro-industrial customers (currently using individual diesel generators); however, it could become economically viable if other economic, environmental, and social benefits are considered (e.g., reduction in CO2 emissions, reduced reliance on imported fuels, and reduced exposure to price fluctuations). supply of sufficient water, and high quality physical and distance. The Sumbawanga cluster (Tanzania) and the market infrastructure. Suitable commodities include Mkushi farming block (Zambia) cases show that grid those typically cultivated on large-scale farms: maize, extension is the more viable option. wheat, sugar, rice, soybeans, and barley. Despite the advantages of the main grid, mini- The main grid has certain fundamental advan- grids may still offer the least cost solution to reach tages that may make it the most viable option, even unserved consumers, overcome grid unreliability, in cases where it is located at a distance. The multiple and leverage private-sector funds to accelerate rural generation sources connected to the main grid help electrification. Case studies in mini-hydro mini-grids mitigate the risk of power failure and enable the utility developed under the Mwenga (Tanzania) and KTDA to minimize costs by balancing supply profiles to match (Kenya) projects show how unreliable grid supplies demand. In contrast, a smaller isolated system based have led to the development of alternative generation on a single generation source may not be amenable to sources. However, the more typical case is establish- different load profiles and is at a greater risk of failure ing mini-grids in greenfield areas and access-deficit due to shutdowns of the sole generation facility. In countries setting up policies and regulations to create addition, due to economies of scale in generation and a level playing field and mitigate uncertainties for the ability to spread fixed costs over a wider set of private-sector, mini-grid operators. The two main ­ consumers, electricity from the main grid tends to be concerns are (i) the ability to be financially sound, cheaper than that from a smaller system. At the same either through charging cost recovery tariffs or time, the size of electricity load required to ensure via- receiving government subsidies and (ii) having regu- bility of grid extension increases with the capital costs lations that specify what happens when the large grid incurred for the extension, which, in turn, is related to reaches the mini-grid areas. Executive Summary xix A number of options exist to make projects finan- to bridge the gap between actual retail tariffs and the cially viable. First, to benefit from economies of scale, levels required for full cost recovery. the local generation capacity can be increased beyond the level of local demand, and surplus power can be sold to the grid. This option is particularly relevant in How can complementarities countries that have introduced FiT programs set above in power and agriculture be the utility’s avoided costs. Selling excess power makes harnessed? it possible to lower the per megawatt cost, but relies on the ability to sell excess generated power. For exam- To realize the full potential of agriculture-power ple, the capacity of Tanzania’s Mwenga mini-hydro integration in Sub-Saharan Africa, the region’s pol- mini-grid is greater than what the tea estate requires; icy makers and power companies must think about therefore, the surplus is sold to the utility and nearby demand creation. Governments should coordinate rural customers. Another option, as is done for the strategies in the power sector with complementary main grid extension projects in Zambia (i.e., Mkushi strategies on rural development and agricultural and Mwomboshi), is to require beneficiaries to partially extension. The experience of agriculture corridors, finance projects and share the development costs clusters, and growth poles should be analyzed and with major customers. Farmers partially contribute to applied on a wider scale. In addition, power compa- capital costs in exchange for receiving power. A further nies should coordinate with other related agencies option is load balancing across beneficiary categories, and institutions to maximize complementarities. which enables the spread of fixed costs, especially cap- Electricity can be prioritized in areas with large irriga- ital costs, across a larger pool of customers with diverse tion potential, combined with access to markets for peak-load profiles. agricultural goods. The sale of agricultural machinery, The role of subsidies to cover some costs should including irrigation pumps and small threshers, can be be highlighted. All of the distributed schemes have promoted as part of a package to encourage elec- received subsidy payments to decrease the level of tricity use in agriculture. In the process of developing cost recovery through retail tariffs. This contributes expansion plans, power companies should account for toward ensuring maximum capacity development, the electricity needs of, and benefits to, both small- increasing the project’s net present value (NPV), holder and commercial scale farmers. improving tariff affordability for customers, and Leveraging complementarities in rural devel- attracting private-sector participation. Subsidies are opment across sectors would likely result in higher particularly necessary for most privately developed, revenues for the utility companies and deliver small-scale projects under 5 MW. By subsidiz- greater economic benefits to rural areas. While ing household connections, which also tend to be power companies can prioritize regions with existing financially unviable, developers can be encouraged or potentially high levels of agricultural production, to expand their customer base to capture additional rural development or agricultural agencies can target subsidies, prioritizing smaller customers close to each areas that are able to take advantage of the many other rather than larger ones. productive use benefits of electricity. The utilities can National policy targets based on economic net create internal units responsible for encouraging the benefits, rather than financial viability, drive invest- productive and efficient use of electricity. Productive ments in rural electrification. For all the cases stud- use units can be responsible for promoting electric ied, the estimated economic viability was high. Power machinery in agriculture, from irrigation to harvest for agricultural use enables the development of pre- and post-harvest. Banks and other financial institu- viously unviable activities, which increases yields and tions should be incentivized to set up credit lines for lowers production costs. The benefits to households farmers and agricultural entrepreneurs to purchase and businesses include savings on energy expendi- agricultural machinery. Given the high expense of tures, better health, and improved educational out- using diesel powered engines for grain processing, comes. Wider benefits accrue from higher incomes campaigns by local government could be developed and improved quality of life. However, subsidies are to promote electricity as a substitute for diesel needed to make the schemes financially viable. All of engines among farmers in areas just gaining access the distributed schemes analyzed received subsidies to electricity. xx Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Coordinated planning encompassing geospa- Supporting the financial health of key sec- tial efforts and multi-agency inputs is necessary. tor institutions, central to the World Bank policy A geospatial map with information about future dialogue in the electricity sector, is important for developments of the national grid, as well as layered this agenda as well. The weak financial status of the data on agriculture and other rural infrastructure, is utilities prevents them from being able to develop important to understand where the load clusters are. financially viable projects without external support. These are the areas where feasibility studies of mini- Furthermore, their constrained cash flows increase grid developments could be carried out for potential the risk of non-payment for the power supplied by future tendering. Clarity in site identification and the private developers, which negatively impacts project regulatory environment is also useful for mini-grid costs and tariffs and, as a result, power affordability. developers and concessioners to allay fears on what If FiTs are not capped at the utility’s avoided costs, happens when the grid arrives. Such integrated maps, this situation could worsen, further deteriorating possibly housed in a national institution, can also the utility’s viability. From the perspective of power support more transparent decision making on infra- sector regulators, the extra cost and delays result- structure expansion and integrated rural development ing from inexperience in negotiating various supply approaches. arrangements may be a hindrance to developing Policy makers can support a stable regulatory private-sector power generation, distribution, and environment for electricity suppliers. To succeed, supply. projects must be implemented within a stable legal Finally, rapid changes over the last few years in environment that imposes requirements and provides small-scale generation and distribution technology, protection. Light-handed regulation of small-scale especially solar PV, have created opportunities to electricity systems is generally more favorable to test new models for viable rural electrification and developers and operators. For example, Tanzania’s power-agriculture integration. Recent techno- small power producer (SPP) framework allows private logical advancements and reduction in small-scale operators to function as power distributors and retail- generation costs have led to heightened interest in ers, charging fully cost-reflective tariffs. This type viable isolated mini-grid development models, such of regulation should tackle the economic barriers of as those based on shared solar PV systems and DC unaffordability and uneconomic supply. Regulation distribution lines. Compatible product development must also extend beyond the power sector to tackle (e.g., TVs, refrigerators, solar pumps, and grain mills) interactions with related sectors. Tanzania’s Mwenga is enabling increased productive use of electricity and Mini-Hydro Mini-Grid Project, one of the first proj- increased aggregate electricity demand from such ects of its kind, encountered significant delays when mini-grids to further improve their viability. While negotiating regulations over water rights, land access, there is limited experience of such mini-grids in oper- import laws, and building permits. Also, information ation (which thus explains why they are not reflected about future developments of the national grid and in our findings), this is a dynamic space with tremen- concession protection is crucial for dispelling devel- dous current interest and significant future potential opers’ reluctance and avoiding potential friction from to spur greater opportunities for power-agriculture tariff differences between customers. integration. endnotes 1. Authors’ calculation from the World Development Indicators (WDI) database. 2. Korwama (2011) estimates that 30 percent of agricultural produce in Sub-Saharan Africa is processed, compared to nearly 98 percent in some developed countries. 3. Apart from the mitigating impact of seasonal variation, the ability to sell excess power to the grid also helps to invest in large generation capacity and reduce costs due to economies of scale in generation. Agriculture and Power Nexus Chapter 1 A griculture predominates the livelihoods of the rural poor in Sub-Saharan Africa; thus, higher Box 1.1: Terminology growth in the agriculture sector, especially Clarification: Agriculture through increased productivity, is instru- and Agribusiness mental in reducing the incidence of extreme poverty in the region. Diao et al. (2012) estimate that the decline Agriculture refers to on-farm production. It in national poverty rates is up to four times higher for includes crops and livestock but not floriculture, agriculture-led growth, compared to growth led by nonag- ­ fisheries, or forestry. Although much agriculture ricultural sectors (e.g., 4.3 times higher for Kenya, 3.1 for in Africa is oriented to sustaining livelihoods, this Rwanda, 1.6 for Nigeria, and 1.3 for Ethiopia). Similarly, study focuses on commercial farming, recognizing ongoing research using the Global Trade Analysis Project that commercial farmers in Sub-Saharan Africa (GTAP) model of world trade finds that productivity are overwhelmingly small and medium in scale. growth in agriculture, compared to growth in other sec- Agribusiness denotes organized firms—from small- tors, is nearly three times as effective in reducing poverty. and medium-sized enterprises to multinational Agriculture and agribusiness comprise most income corporations—involved in input supply or down- generating activities in Sub-Saharan Africa’s largely stream transformation. It includes commercial rural economies (box 1.1), together accounting for nearly agriculture involving some transformation activities half of its gross domestic product (GDP) (figure 1.1). (even if they are basic). It includes smallholders and Agricultural production is the most important sector, microenterprises in food processing and retail to averaging 24 percent of the region’s GDP. Agribusiness the extent that they are market oriented. Indeed, input supply, processing, marketing, and retailing con- these producers and enterprises comprise the bulk tribute another 20 percent (World Bank 2013). Thus, of agribusiness activity in Africa today. transformation of the agriculture sector through improved productivity and incomes can simultaneously help achieve Source: World Bank 2013. both robust economic growth and poverty reduction. In other developing regions, agricultural transformation has resulted in declining numbers of the poor. Thus, for Sub-Saharan Africa, where poverty rates have remained constrains development of on-farm and off-farm eco- stubbornly high, utilizing agricultural transformation to nomic activities, as it does for other manufacturing and tackle poverty in rural areas—where more than 70 percent services firms. Rural electrification can raise productivity of the region’s poor live—is a critical part of any poverty and income when farmers switch from manual to electric- reduction strategy. ity powered inputs and small industries begin using electric For both agricultural and nonagricultural households, tools and machinery. Access to reliable electricity supply electricity is needed to raise living standards,1 as well as can increase productivity along the agriculture value chains enable broader economic development. Lack of access to and enable increased production and income generation reliable and affordable electricity in Sub-Saharan Africa for the farm sector and the rural economy as a whole. 1 2 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa The United Nations Sustainable Development Goals Past Performance: A Missed Opportunity (SDGs), adopted in September 2015, set a target for universal access to affordable, reliable, and modern energy Agricultural growth has typically lagged behind that of other services by 2030 (SDG 7). The acknowledgment of sectors in Sub-Saharan Africa. Vulnerability to weather modern energy access as a development goal builds on shocks, limited use of modern tools and inputs, low levels of the momentum created by the Sustainable Energy for processing, poor development of rural financial markets, and All (SE4ALL) initiative, which has galvanized the interna- market access barriers have all hindered agricultural growth tional community into action to achieve concrete energy and kept agricultural productivity and incomes low. Between related targets.2 Under SE4ALL, the three goals to be 2000 and 2013, the share of agriculture in GDP declined achieved by 2030 are: (i) universal access to modern by 6 percentage points (from 20 percent to 14 percent).3 energy services, (ii) doubling the share of renewables in Only a small percentage of the region’s agricultural the global energy mix, and (iii) doubling the growth rate of production undergoes industrial processing.4 For the energy efficiency. world’s high-income countries, processing adds about US$180 of value per ton of agricultural produce, com- pared to only $40 for Sub-Saharan Africa. This is related High Potential for Agricultural to the small size of the agribusiness sector compared to Transformation on-farm agriculture in Sub-Saharan Africa relative to other regions. For developing countries, including those in Historically, agriculture in Sub-Saharan Africa has Sub-Saharan Africa, the ratio of value added in agribusi- underperformed despite the region’s comparative ness to that of farming is typically 0.6. This ratio increases advantage stemming from abundant land and water to 2.0 for transforming countries (mainly in Asia), 3.3 resources. However, recent developments have created for urbanized countries (mostly in Latin America), and more favorable conditions for an agricultural trans- 13.0 for the United States, indicating significantly more formation. Today there is an expectation that well-­ value created in the downstream agribusiness sector than informed policies and investments can put agriculture on-farm production for countries outside Africa. These on a higher growth path to achieve its vast potential comparisons reflect the positive correlation between the and raise rural welfare. relative importance of agribusiness and economic growth: both per capita GDP growth (figure 1.1a) and human development indices (da Silva et al. 2009). Figure 1.1: Historical performance in agriculture a. Ratio of food processing to agricultural value added b. Market share of global exports 9.0 Food processing value added/agriculture added 0.6 8.0 % of world agricultural exports HUN 7.0 6.0 ARG 0.4 ROM BRA MEX 5.0 4.0 ZWE ECU IRN MYS BOL ZAF 3.0 PER 0.2 SVK SEN PHI MAR TUR 2.0 MWI THA IDN 1.0 EGY NPL BGD IND 0.0 0 UGA 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 0 2,000 4,000 6,000 8,000 GDP per capita, constant 2000 US$   Brazil Thailand SSA Sources: World Bank 2008, 2013. Note: In figure 1.1a, three-letter codes represent the following countries: ARG = Argentina, BGD = Bangladesh, BOL = Bolivia, BRA = Brazil, ECU = Ecuador, EGY = Egypt, HUN = Hungary, IDN = Indonesia, IND = India, IRN = Iran, MAR = Morocco, MEX = Mexico, MWI = Malawi, MYS = Malaysia, NPL = Nepal, PER = Peru, PHI = Philippines, ROM = Romania, SEN = Senegal, SVK = Slovak Republic, THA = Thailand, TUR = Turkey, UGA = Uganda, ZAF = South Africa, ZWE = Zimbabwe. Agriculture and Power Nexus 3 For more than four decades, Sub-Saharan Africa’s An Improving Outlook share in global agricultural export markets has been on the decline. By the early 1990s, the region’s share had The high yield gap between Sub-Saharan Africa and other fallen to about 2 percent, 5–6 percentage points lower regions underscores the large potential for Africa to catch than in the 1960s. Meanwhile, other important agricul- up with the productivity frontier (World Bank 2013). The tural exporters, including Brazil and Thailand, have seized increasing prominence of the agriculture sector among market share despite having a tiny fraction of Africa’s land policy makers, the private sector, and the development area, especially in the case of Thailand (figure 1.1b). community has been driven, in part, by the recognition of African imports of agricultural products have sky- decades of prior neglect of the sector by governments and rocketed due to the gap between regional demand and donors, as well as the urgent need to mobilize small-scale supply. From the 1990s to the 2000s, the balance of farmers to increase food production in order to avoid food trade in food staples for Europe and Central Asia, South security challenges in the near term. Asia, and East Asia and the Pacific moved from deficit Over the past decade, African governments have (i.e., imports exceeding exports) to surplus; however, for demonstrated a renewed and growing commitment toward Sub-Saharan Africa, this gap greatly expanded. While food agriculture. The improving policy environment, led by trade deficits are expected in regions without a compar- the Comprehensive Africa Agriculture Development ative advantage in food production, such as the Middle Programme (CAADP) (box 1.2), high investor interest, and East and North Africa, they are symptomatic of a missed technological advances that ease implementation of neces- opportunity in Sub-Saharan Africa, which is endowed with sary reforms, particularly in land administration, have cre- abundant natural resources for efficient production. ated excellent conditions for an agricultural transformation.5 The outlook for agricultural development in Sub- Saharan Africa is improving.6 Economic growth and Investment Funding Challenges urbanization have fueled an increase in food demand in Investment funding for the agriculture sector, especially Sub-Saharan Africa. In addition, continued improvements primary production, is limited by perceived high risks and in infrastructure and the benefits of lower oil prices have low returns. Poor infrastructure on farms and along the resulted in increased domestic food production. Although supply chains, low access to credit and product markets, recent declines in agricultural prices may dampen price and other regulatory hurdles have kept returns from incentives for agriculturalists, they may further increase agricultural investments in Sub-Saharan Africa below food demand and thus induce farmers to grow food and potential. Over the past decade, the increased inflows of other agricultural commodities for the market. commercial finance, especially foreign direct investment (FDI), have been vastly inadequate. Official development assistance (ODA) has helped, in part, to fill the gap. In Major Approaches to Agricultural 2003–12, ODA for agricultural projects in Sub-Saharan Development Africa rose 121 percent (from US$1.1 billion to $2.5 bil- lion). Over the same period, the share of aid allocated to There are two major approaches to agricultural devel- the agriculture sector in Sub-Saharan Africa grew from opment in Sub-Saharan Africa. The first is a cluster 37.4 percent to 40.3 percent, the highest share increase approach, which focuses on particular areas with a high for the period (Development Initiatives 2015). level of infrastructure access and development poten- The high costs of connecting agricultural land to back- tial. This generally involves support for large farms and bone infrastructure in Sub-Saharan Africa cannot be easily commercialized agriculture as growth poles. The second absorbed by most medium-sized farming businesses, let approach is smallholder agriculture, which centers on alone small-scale farms. But without these “last-mile” infra- support for smallholder farmers to increase their produc- structure investments, the region’s farmers cannot increase tivity and access to markets. These two approaches differ their productivity. Furthermore, without access to con- in their implications for electricity supply in rural areas. cessional funding, the establishment costs of an outgrower program, especially those involving provision of infrastruc- Cluster Approach ture services to small-scale farmer organizations, can be prohibitive, explaining why so few of the nucleus farm and Over the last 20 years, one rural development trend in outgrower models have been successfully established. multiple countries across Africa has focused on integrated 4 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 1.2: Africa’s Vision for Agriculture: CAADP Goals The Comprehensive Africa Agriculture Development Programme (CAADP), initiated in 2003, strives to improve country frameworks to support agricultural development. The CAADP’s initial 2015 target, extended through 2025, envisions that the continent should achieve the following goals: ºº Attain food security in terms of both availability and affordability and ensure access of the poor to adequate food and nutrition; ºº Improve the productivity of agriculture to attain an average annual growth rate of 6 percent, with particular attention to small-scale farmers, especially focusing on women; ºº Have dynamic agricultural markets among nations and between regions; ºº Integrate farmers into the market economy, including better access to markets, with Africa to become a net exporter of agricultural products; ºº Attain more equitable wealth distribution; ºº Become a strategic player in agricultural science and technology development; and ºº Practice environmentally sound production methods, featuring a culture of sustainable management of the natural resource base (including biological resources for food and agriculture) to avoid their degradation. Source: CAADP 2012. infrastructure and social development for specific areas. are likely to induce agricultural intensification. Both This cluster or corridor development approach has signifi- large-scale and smallholder agriculture will benefit from cant implications not only for the development of agri- increased productivity induced by spillovers, greater culture, but also for how electrification and other types of connectivity, and reduced transaction costs. The ability to institutions develop plans to serve such areas (annex A). serve wider markets for their goods and services will create Clusters are geographic concentrations of intercon- greater incentives to innovate. nected companies, including intermediate goods suppliers, The cluster approach brings together agricultural service and infrastructure providers, and associated insti- research stations, nucleus large farms and ranches, com- tutions in a particular product space or sector. Clusters mercially focused farmer associations, irrigated block farm- benefit from geographical agglomeration economies that ing operations, processing and storage facilities, transport may result from the proximity between intermediate and and logistics hubs, and improved “last-mile” infrastructure final goods suppliers, labor market pooling, and knowledge to farms and local communities. When occurring in the spillovers (Marshall 1890; Krugman 1991). Despite falling same geographical area, these investments result in strong transportation and communication costs, clusters con- synergies for agricultural growth, helping create the condi- tinue to be relevant today due to the underlying benefits tions for a competitive and low-cost industry. of increased firm productivity, innovation, and formation The essential elements of a cluster approach include of new businesses (Porter 1990). Transportation growth the following: corridors, a closely related concept, places the significant ºº Having a long-term strategy for agricultural develop- economies of scale of infrastructure development at the ment, recognizing that transformation occurs over an center of the benefits from spatial agglomeration. extended period (e.g., 10–20 years); In the case of agriculture, clusters can affect develop- ºº Understanding and leveraging vertical and horizon- ment in several ways. Improved access to infrastructure tal linkages between farms and other businesses to can lead to increased productivity of farms and companies maximize value addition; within a concentrated economic area. As opposed to ºº Commissioning robust analysis of the constraints on remote rural areas, these clusters of economic activity commercial agriculture and recommending how these benefit from joint access to necessary infrastructure can be addressed; services, linkages to upstream and downstream activi- ºº Establishing an independent public-private part- ties, and connectivity to markets. Better connectivity to nership organization to help coordinate and target markets and access to infrastructure, including electricity, Agriculture and Power Nexus 5 agricultural development programs and public invest- can be an important source of competitiveness in their ments; and own right. An additional benefit of smallholder led agri- ºº Leveraging government and development partner cultural growth is the much higher level of ­ second-round resources to catalyze socially and environmentally demand effects that occur when income gains are realized optimal private investment. by smallholder households, as opposed to large commer- cial farms.” Electricity is one of the fundamental requirements for Hazel et al. (2007) make the case for development cluster or corridor development. Investments in electricity of the smallholder sector, pinpointing the importance infrastructure must adequately account for long-term of infrastructure development to support it. “The case demand growth due to increased demand from large for smallholder development as one of the main ways to farmers, small farmers, farm service businesses, and other reduce poverty remains compelling. The policy agenda, tertiary development in such growth areas. Accounting however, has changed. The challenge is to improve the for medium- to long-term demand growth will allow bene- workings of markets for outputs, inputs, and financial fits to accrue from economies of scale and thus can lower services to overcome market failures.” The point is that costs to end consumers. numerous factors can support smallholder agriculture, including the coordinated efforts of farms, the private Smallholder Agriculture sector, nongovernmental organizations (NGOs), and government. Support can take the form of agricultural Most agriculture in Sub-Saharan Africa today involves research, agricultural extension, and infrastructure devel- smallholder farms, usually characterized by landholdings opment (e.g., roads and provision of electricity). of less than 2 ha, with a subsistence orientation. While the Given the “competing barriers” to agricultural large farm, agribusiness model has an important role to development, the provision of electricity infrastruc- play in promoting agricultural growth in Africa, small- ture, by itself, is unlikely to make an appreciable differ- holder agriculture is key to revitalizing the rural economy ence. Electricity investments must be coordinated with and tackling poverty. interventions targeting agricultural development (e.g., The question is what role should smallholder or family improving agricultural inputs and technology adoption; farms play, in contrast with large farms, in striving for pro- agricultural extension services; research on smallholder ductivity transformation in Sub-Saharan Africa. In agri- farming practices; and other infrastructure, including cultural economies, which describes most of Sub-Saharan roads, markets, and water supply). The combination Africa, smallholder agriculture comprises the majority of of these inputs will increase the growth of agricultural employment and production. With rising demand for sta- production and have a multiplier impact on the rural ple food crops and high-value commodities resulting from economy. rapid urbanization in the region, an increase in smallholder In short, it is not the role of electricity institutions productivity can arise from easing constraints on access to to promote agriculture; rather, their role is to support credit, infrastructure, and markets. Targeting the develop- agriculture in conjunction with other programs. This may ment of smallholder agriculture is also an effective way to seem a daunting task from a policy perspective, given reduce rural poverty. Thus, smallholders in Sub-Saharan that, in most governments, electricity, agriculture, rural Africa have a critical role to play as a source of agriculture development, and water institutions reside in isolated competitiveness. The World Bank (2009) finds that “con- “silos.” However, in countries with successful rural electri- trary to expectations, few obvious scale economies were fication programs, electricity companies have often found found in the production systems analyzed for the CCAA ways to deal with such silos, mainly through outreach and study. Compared with those of large commercial farms, coordination (Barnes 2007). For example, in Tunisia, the family farms and emerging commercial farms were typi- main electricity company (STEG) had regular meetings cally found to have lower shipment values at the farm level with rural development agencies and coordinated expan- and/or final distribution point (shipment values reflect sion plans to provide electricity in communities that were production and delivery costs). Large commercial farms receiving other development inputs. can play an important strategic role by contributing to Coordinated planning of rural electrification would the achievement of the critical mass of product needed to require a change in the way the electricity compa- attract local and international buyers, but the value chain nies operate, taking into account expected growth in analysis shows that investments in smallholder agriculture ­ energy-intensive agricultural activities and development 6 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa programs in the pipeline. To do this, electricity companies held its own for some cash crops (e.g., cocoa, rubber, need to develop an effective information sharing mech- fruits and vegetables, and tobacco) and has even gained anism with relevant agriculture sector stakeholders. This market share for others (e.g., cashew, tea, and sesame could involve reaching out to relevant agricultural agen- seed), showing some evidence of its productive potential. cies; promoting productive uses of electricity; and under- Third, Sub-Saharan Africa is poised for demographic standing future growth and development trends, especially transition and wealth creation, reflecting the growing with regard to smallholder agriculture. Electricity access aspirations of its people. According to the United Nations, for agriculture and rural businesses could effectively be between 2013 and 2050, the region’s population will more promoted as part of an overall strategy to support small than double, from about 900 million to 2.1 billion (United farmers through a variety of activities (e.g., development Nations 2013). While one-third of its population is already of farm cooperatives to purchase and market local farm living in urban areas, this proportion should increase to goods; machine rental; and agricultural extension, includ- 50 percent by 2035. Globally, urban food markets are set ing advice on irrigation practices, seeds, and fertilizers). to increase fourfold, exceeding US$400 billion by 2030 (World Bank 2013). For Africa’s 11 biggest economies, the middle class, defined as those earning at least US$450 per Agricultural Growth to Raise Rural month, tripled between 2000 and 2014 (from fewer than Welfare: Reasons for Optimism 5 million people to 15 million). Over the next 15 years, these numbers may rise by a further 25 million (Standard There are four main reasons to believe that agriculture in Bank Research 2014). Sub-Saharan Africa is poised for growth that can con- Sub-Saharan Africa’s rapid population growth, tribute significantly to raising rural welfare. First, relative accompanied by robust economic growth, is creating to much of the rest of the world, the region’s land and a huge regional urban market for agricultural goods. A water—the major natural inputs necessary for growing recent World Bank study on agribusiness predicts that crops and raising livestock—are underutilized, creating a the market for agricultural goods and commodities could huge potential for agricultural growth (figure 1.2). Of the reach US$1 trillion by 2030 (figure 1.3). It states that world’s total land area suitable for sustainable production “the majority of the increase in food consumption will expansion—that is, non-protected, non-forested land with occur in cities. Based on the United Nations’ projections low population density—Sub-Saharan Africa has the larg- of urbanization and assuming that the per capita value of est share by far, accounting for about 45 percent.7 In the food consumption is 25 percent higher in urban areas than case of Latin America, which accounts for only 28 percent rural areas, the urban market is set to expand fourfold in of land suited for production, 73 percent of that amount is 20 years” (World Bank 2013). This expansion in regional located within six hours’ travel time to the nearest market, demand will create an enormous opportunity for African compared to just 47 percent in Sub-Saharan Africa—a agriculture and agribusiness. result of the subcontinent’s generally poor state of infra- Fourth, agriculture is critical for managing the urban structure (Sebastian 2014). Sub-Saharan Africa also has transition that Africa will undergo. To date, this process significant untapped water resources. Only 2–3 percent has been driven to a large extent by populations being of the region’s renewable water resources are being pushed out of rural areas, rather than cities attracting a utilized, compared to 5 percent worldwide (World Bank workforce by acting as growth poles. It would be a more 2013). Its irrigation intensity, one of the lowest in the positive process were it driven by improving economic world, represents only 5 percent of total cultivated area, opportunities in the cities that would gradually pull in rural compared to 37 percent for South Asia and 14 percent in residents, rather than declining conditions and periodic Latin America (World Bank 2008). Despite an absolute disasters in rural areas that push residents out. The latter abundance of water resources, lack of irrigation develop- situation often leads to conflict and waves of migration ment and storage capacity has limited the availability of that cities find difficult to absorb, typically leading to water in certain basins, resulting in water stress. Also, the expanded slums. The objective of a transition strategy—of uncertainties related to climate change raise concerns which electrification is a key element—is thus to enhance about future water availability (box 1.3). living conditions and economic opportunities in rural areas. Second, despite Africa’s overall decline in the share In this context, agriculture and agribusiness can play of agricultural exports, a recent disaggregated view tells a critical role in jump-starting the economic transfor- a more nuanced story. Since the early 1990s, Africa has mation through development of agro-based industries in Agriculture and Power Nexus 7 Figure 1.2: Land and water resources potential in Sub-Saharan Africa a. Land potential, by world region b. African countries with largest available land resources Uncultivated arable land, Million ha million ha Sudan 202 Sub-Saharan Africa DRC Madagascar 123 Mozambique Latin America Chad Zambia 52 East Europe and Angola Central Asia Tanzania CAR 14 Ethiopia East and South Asia Cameroon Kenya 54 Rest of the world Mali 112 m ha Others 0 50 100 150 200 250 0 10 20 30 40 50 Land available Area less than 6 hours to market    Cultivated Available c. Aquifer productivity in Africa Aquifer productivity Very High: >20 l/s High: 5–20 l/s Moderate: 1–5 l/s Low-Moderate: 0.5–1 l/s Low: 0.1–0.5 l/s Very Low: <0.1 l/s Sources: World Bank and Schaffnit-Chatterjee 2014 (figure 1.2 a, b); British Geological Survey (figure 1.2c) (http:// www.bgs.ac.uk/research/ groundwater/international/africanGroundwater/maps.html). 8 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 1.3: Making Africa’s Power and Water Infrastructure Climate Resilient Uncertainty over water availability for productive Figure B1.3.1: Changes in irrigation uses is a critical issue facing Sub-Saharan Africa’s revenues from climate change, 2015–50 infrastructure investments, especially long-lived (present value) infrastructure (e.g., irrigation schemes, dams, and $24.8 billion power stations). Variations in annual rainfall and 90 gain monthly rainfall distribution, along with tempera- ture changes due to drier or wetter climates, could Di erence from reference case (%) 10 $1.8 billion $0.2 billion put power and water infrastructure at risk, affecting gain $0.3 billion gain gain $3.9 billion gain $0 billion gain operation and cost over their life span. Beyond 0 impacting the technical performance of infrastruc- $42.1 billion $2.4 billion ture, uncertainty about drier or wetter futures could –10 $13.2 billion loss loss loss significantly modify its financial viability, incurring $0.8 billion $7 billion losses or gains. In a drier scenario, for example, –20 loss loss shortfalls in irrigated production could raise demand $0.9 billion loss for food imports, and thus increase food prices (fig- –30 Volta Eastern Zambezi Nile Niger Senegal ure B1.3.1). Nile Equatorial Lakes Cervigni et al. (2015) highlight significant dispari- Basin ties across Africa’s seven main river basins: Congo, Maximum relative gain due to climate change/best scenario Niger, Nile, Orange, Senegal, Volta, and Zambezi. Maximum relative reduction due to climate change/best scenario The study estimates that, in dry scenarios, loss in irrigation revenue could range between 5 and Note: The bars reflect, for each basin, the range of economic 20 percent for most basins. For wet estimates, outcomes across all climate futures; that is, the highest increase (light blue bars) and highest decrease (dark blue bars) of irrigation revenue gains could reach 90 percent for the Volta revenues (discounted at 3 percent), relative to the no-climate- basin, but would be vastly less (1–4 percent) for the change reference case. The outlier bar corresponding to the Volta other areas. Under the driest scenarios, unmet irri- basin has been trimmed to avoid distorting the scale of the chart gation demand could drop by more than 25 percent and skewing the values for the other basins. Estimates reflect the in the Zambezi basin. The magnitude of impact will range, but not the distribution, of economic outcomes across all depend on the willingness and ability of decision climate futures. Each basin’s results reflect the best and worst makers to integrate climate projections and their scenarios for that basin alone, rather than the best and worst uncertainty into the planning and design of power scenarios across all basins. The Congo and Orange basins are and water infrastructure. excluded because the effects on irrigation are minimal. Africa’s need to tap its irrigation potential represents a window of opportunity to make power and water infra- structure climate resilient. Although such a paradigm shift will take time, practical steps to integrate climate resil- ience can be undertaken now. For example, Cervigni et al. (2015) recommend defining and promoting technical standards for integrating climate change into project planning and design and launching training programs target- ing relevant stakeholders. Source: Cervigni et al. 2015. Agriculture and Power Nexus 9 Figure 1.3: Projected value of food a vibrant agricultural sector. Investments in agricultural markets in Sub-Saharan Africa productivity can spur the development of downstream agribusiness; in turn, agribusiness investments stimulate 1,000 agricultural growth through the provision of new markets and development of a vibrant input supply sector. Micro-, 800 small-, and medium-sized enterprises (MSMEs) comprise the bulk of Sub-Saharan Africa’s agriculture-related value chains. In West Africa, for example, three-fourths of $ Billion 600 agriculture-related firms are micro or small enterprises (Staatz 2011). Taking advantage of this opportunity requires that 400 both farmers and agribusinesses ramp up production, while becoming more competitive; otherwise, the balloon- 200 ing demand will be filled by imports. This requires devel- oping agriculture value chains and agribusiness to enhance processing, logistics, market infrastructure, and retail 0 networks, all of which require electricity. 2010 2030 However, electricity remains a critical constraint to Urban Rural the development of the agro-industrial sector. According Source: World Bank 2013. to data from WBG enterprise surveys, the majority of firms in many countries of Sub-Saharan Africa identify lack of electricity access as a major obstacle (figure 1.4a). Figure 1.4: Electricity as a constraint to food-sector development in Sub-Saharan Africa a. By country b. Comparison with other sectors Percent of firms identifying electricity as a major Electricity considered as a constraint to invest in constraint to develop the food sector Sub-Saharan Africa (data collected from 2006 to 2014— (Enterprise Surveys—World Bank Group) Enterprise Surveys—World Bank Group) Total average Burundi Congo, Dem. Rep. Focus on the Guinea food sector Guinea-Bissau Ghana Tanzania Senegal Uganda All sectors Mali Angola Zimbabwe Rwanda 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Zambia   Nigeria Mauritius Mauritania Namibia Kenya Mozambique Madagascar Swaziland 0.0 20.0 40.0 60.0 80.0 100.0 Source: WBG 2015 (http://www.enterprisesurveys.org/). 10 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa In fact, the fraction of firms in the food sector that competitive for regional and global markets, and consider electricity a constraint to investment is signifi- ultimately increasing the profitability of agricultural cantly higher than the average fraction in all other sectors activities. (approximately 29 percent, compared to less than 15 per- cent) (figure 1.4b). Successful commercial agriculture is typically charac- Rural Electrification Has terized by the following elements: Lagged Behind ºº Ample suitable land, with benign climate conditions A majority of Africans—nearly 600 million people—live and reliable water availability. without electricity; instead, they rely on kerosene or ºº Private-sector participation in sector development, dry-cell batteries as coping mechanisms. The latest with higher skills levels and access to international estimates peg Sub-Saharan Africa’s electrification rate at capital and markets, with strong government support 35 percent overall, with 69 percent in urban areas and just (e.g., through a favorable policy and regulatory envi- 15 percent in rural areas (figure 1.5a). Viewed from space, ronment and publicly funded research and develop- the picture of Africa’s nightlights, showing large sections ment and infrastructure). of perpetual darkness, is a stark contrast to the rest of the ºº Affordable and reliable access to supporting infra- developing world, and the evolving disparity is enormous structure, in the form of reliable electricity supply, (figure 1.5b). transport links to markets, and irrigation in drier Historically, the region’s population growth has climates (often powered by grid-based electricity). outpaced the rate of expanding electricity access, and the ºº Clusters of large-, medium-, and small-scale gap in rural areas is enormous. Amid a population increase commercial farming, processing, and services firms of 202 million, only 59 million people have received concentrated in discrete geographical areas. Taken electricity. If business as usual continues, by 2030, together, the result is a reduction in costs of produc- Sub-Saharan Africa will be the world’s only region with tion through economies of scale, making prices more an increase in the number of people without electricity Figure 1.5: Electrification rate, by developing region a. Millions of people with and without access, 2012 b. Evolution of access (%), 1990–2012 2,000 100 90 80 1,500 23 363 70 0 60 555 1,000 754 50 977 40 22 589 30 61 500 271 807 20 477 659 14 10 0 237 0 94 147 278 272 36 0 0 46 74 54 109 87 0 1990 2000 2010 2012 Oceania CCA NA WA LAC SEA SSA DEV EA SA People with access (rural) People with access (urban) EA LA NA Oceania Population without access     SA SEA SSA WA Source: IEA and World Bank 2015. Note: CCA = Caucasus and Central Asia, EA = East Asia, LA = Latin America, NA = North Africa, SA = South Asia, SEA = Southeast Asia, SSA = Sub-Saharan Africa, and WA = West Africa. Agriculture and Power Nexus 11 access. Furthermore, the urban/rural disparity in elec- agricultural value added and incomes. Generally, the most tricity access is set to widen as most expansion is likely to dramatic changes in agricultural development due to rural occur in densely populated urban areas (IEA and World electrification have resulted from increased irrigation. Bank 2015). With greater access to electricity, it is more cost-effective The biggest challenge to rural electrification in the for farmers to irrigate their fields since electric pumps Sub-Saharan Africa region is the lack of commercial via- require low maintenance and are more efficient than die- bility of expanding connections. Low population density, sel alternatives. Irrigation also allows farmers to produce coupled with the limited purchasing power of most rural multiple crops in a single year and improve the productiv- consumers, implies that, in many cases, investment in ity of existing farms. These advantages lead to higher crop rural grid extension is cost-prohibitive. This problem is yields and incomes.8 compounded by the poor financial health of the region’s This relationship has most often been documented in distribution utilities, owing to a combination of factors India, which historically has emphasized the use of irriga- (e.g., low consumer base, historical mismanagement, tion pumps and new agricultural technologies to improve inadequate tariffs, high generation costs, and high rates agricultural productivity (Barnes, Peskin, and Fitzgerald of technical and nontechnical losses). The high cost of 2003). While efforts to improve rural development supply, coupled with low tariffs, puts an inordinate strain through electrification have been relatively successful on sector finances. in some countries, the question is whether this experi- This situation, in turn, traps the sector in a self-­ ence is applicable to Africa, with its low levels of existing reinforcing cycle of low investments in expansion and irrigation. improvement, resulting in an expensive, poor quality The productive impact of rural electrification depends electricity supply, circling back to low investments. Thus, heavily on several enabling factors: government policy, many of the region’s countries are stuck in a cycle of infrastructure, and complementary development pro- low generation capacity, excess demand, and inadequate grams. Electrification is an important enabler for the mobilization of private-sector investment. Breaking this development of rural businesses (e.g., small commercial negative cycle requires a multipronged approach custom- shops, grain mills, sawmills, and brickworks); however, ized to the financial, economic, and political realities of it cannot produce an explosion of economic activity in particular countries. Least-cost grid expansion, wherever the absence of roads and access to finance and markets. viable, should be creatively complemented by a decen- If these complementary conditions are inadequate, the tralized off-grid strategy based on distributed generation growth of rural economies, especially agriculture, will in the form of mini-grids, micro-grids, or stand-alone likely remain lethargic and may, in turn, adversely impact systems. the viability of the rural electrification program.9 One potential solution to address the region’s rural electrification challenge is having an anchor load, defined Agriculture as an Anchor Load as large consumers that offer power utilities a consistent for Rural Electrification and substantial source of revenue, which offsets a portion of the fixed costs of electricity supply to rural households. In recent years, African governments, donors, and the Anchor loads help ease the constraint posed by the low private sector have been reviewing the success stories demand profile of rural customers. Guaranteed demand of such countries as Brazil and Thailand in an attempt to from anchor-load customers ensures the power producer replicate or adapt agribusiness and rural electrification or utility a certain level of revenue, and may help to defray development models that take individual country charac- the fixed costs of rural electrification through demand teristics into consideration. In the case of India, the most aggregation (along with household and commercial notable example, rural electrification was strongly linked demand in neighboring communities of the anchor load). to the promotion of high-yield crop varieties and the In short, an anchor load helps overcome the problem spread of irrigated agriculture, facilitated by electric water of low demand, which constrains the viability of rural pumps with subsidized or free electricity. Here it was clear electrification. that the financial viability and reliability of rural electrifi- In some developing countries, the Anchor Business cation were linked to promoting productive uses. Community (ABC) model is being piloted, using cell- The financial viability of agricultural anchor loads rests phone towers and mining companies as anchor loads.10 In on the ability to use electricity to generate an increase in this context, the supply options range from self-supply 12 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa by the agribusiness to intermediate arrangements with along the various agriculture value chains, aggregated with an independent power producer (IPP) to grid extension. ­ commercial/household electricity demand, can potentially A recent study that analyzed the integration options make it feasible to extend the grid or create opportuni- between power and mining established a typology of ties for small IPPs and mini-grid operators. In addition power sourcing options for mines (Banerjee, Romo, and to demand aggregation, supplying both household and McMahon et al. 2015). agro-processing demand may create a balanced daily load Agriculture can potentially fit into this category of profile, helping to disperse capital and fixed operating anchor load to sustain small-scale supply arrangements costs over a larger set of consumers. with commercial establishments (including irrigation) and In addition to providing anchor loads, agricultural households in rural areas. In this way, electricity demand production can provide fuel for off-grid solutions in rural along the agriculture value chains, as well as commercial/ areas (annex B). Agricultural by-products can serve as household electricity demand, can create opportuni- cheap sources of locally available fuel for biomass electric- ties for the IPPs and mini-grid operators. In addition ity generation; they can be derived from various types of to demand aggregation, supplying both households and processing (e.g., cotton, groundnut, soybean, wheat, and agro-processing may create load balance; the demand of other cereals), but the most common ones are rice husks households and agro-processing peak at different times and sugarcane waste (i.e., field waste and bagasse). of the day, which can help to disperse capital and mainte- Such opportunities are now being commercially har- nance costs over a larger set of consumers. nessed in various countries and regions of the world. For The development of anchor loads can benefit both example, India has created a business model to serve rural centralized and decentralized approaches to rural elec- households using husk power, whereby agricultural residue trification. In the case of grid extension, promoting the (e.g., rice husks, mustard stems, corn cobs, and certain development of relatively large anchor customers in off- grasses) is cost-effectively converted into electricity. In grid areas could tip the balance in terms of the economic this study, the scope of agriculture’s role is limited to that viability of extending the grid to connect to the anchor of an anchor load in rural areas of the Sub-Saharan Africa load and bringing the grid closer to communities without region. electricity access. In current-day industrialized econo- mies, such anchor customers as mills and factories were an integral part of the electrification experience. In Sub- Study Purpose and Methodology Saharan Africa too, national grid expansion plans tend to prioritize district commercial centers and areas with Rural electrification is at a crossroads in Sub-Saharan factories or other large commercial customers. Beyond Africa; for many countries, the challenge is overwhelming, demand from the anchor customer, grid extension can be but opportunities are also emerging. It is up to govern- made viable through the potential to sell electricity back ments, the private sector, and international communities to the grid (in cases where there is an in-house generation in the region to decide how these opportunities will be facility). harnessed for the benefit of Africans living in the dark. Grid extension may not be viable if anchor customers Recently, the WBG’s Energy and Extractives Global are not large enough or are located in relatively remote Practice in the Africa Region commissioned a series of areas. In such cases, smaller isolated grid systems or mini- studies to explore potential solutions to the challenge grids can be used to save on costs associated with trans- of bringing power to Africa. This study, which follows on mission infrastructure. Mini-grids can be developed by the recent initiatives of Banerjee, Romo, and McMahon aggregating demand from the anchor load and surround- et al. (2015), Hussain et al. (forthcoming), and Hosier ing communities, with electricity generation and distribu- et al. (forthcoming), is designed as a joint effort between tion undertaken through a context-specific combination the Energy and Extractives, Agriculture, and Trade and of a small, in-house power producer and anchor business Competitiveness Global Practices. It also complements or public utility. the ongoing analytical work of the Latin America and For both on- and off-grid access solutions, the Caribbean region on energizing agriculture. presence of an anchor-load customer greatly improves This study’s overall aim is twofold: (i) to identify the financial viability. In principle, activities along agri- potential synergies between agriculture value chains culture value chains require electricity and thus might and rural electrification expansion and (ii) to examine serve this role. The electricity consumption of activities the challenges in harnessing this potential. Its specific Agriculture and Power Nexus 13 objectives are to (i) conduct an evidence-based analy- been developed, as well as those in progress or proposed. sis of the extent of the potential of power-agriculture The cases covered a range of commodities (e.g., fruits, integration for specific case studies on agriculture value floriculture, maize, sugar, tea, vegetables, and wheat). chains; (ii) assess alternative supply arrangements (busi- Since agriculture is a dispersed activity with varied ness models) for providing electricity to the combined scales of production, results of this analysis need to be power demand of agriculture and local commercial and considered with the following caveats. First, although residential demand; (iii) analyze barriers and institutional the study provides an estimate of power demand from mechanisms that will create the enabling conditions for agriculture in 2030, it was unable to capture the location private-sector participation in this space; and (iv) iden- of this demand, the extent to which it can be met by tify operationally relevant opportunities for piloting this simply increasing the generation capacity of national grids concept. (i.e., the grids already extend to production and process- This work builds on two background studies prepared ing areas), and whether alternative power sources (e.g., by the consulting consortium of Economic Consulting isolated electricity mini-grids) are the most viable supply Associates (ECA) and Prorustica in 2014–15, which options. Second, the study was unable to capture the nec- involved field visits and stakeholder discussions in the essary financial viability of power supply with reference to countries covered. The first study analyzed the landscape the price that the agricultural activities could afford to pay for rural electrification centered on agricultural activities, for power. while the second examined a set of eight case studies on The rest of this report is organized as follows. powered agribusiness activities from across Sub-Saharan Chapter 2 presents the context of power needs from Africa (Ethiopia, Kenya, Mali, Tanzania, and Zambia). agriculture, while Chapter 3 reports on the detailed anal- The primary focus of the landscape study was on power ysis of power needs by selected value chains. Chapter 4 consumption of agricultural activities within value chains, discusses power supply arrangements for a suite of case identifying where sufficient demand from the activity studies in three countries, encompassing technical, makes it possible to provide an economic or socioeco- economic, and financial analysis. Chapter 5 reviews the nomic rationale for an electrification project that may potential for harnessing power-agriculture synergies then be extended to support surrounding communities. and provides alternative integration scenarios using two The case studies comprised both national grid-connected simulated case studies. Finally, Chapter 6 summarizes the activities and those powered by distributed generation study’s key findings and recommends actions required to systems. They included power schemes that had already promote power-­ agriculture integration. endnotes 1. Households that connect to the electricity grid benefit immediately from better household lighting. With brighter light in the home, children spend more hours studying, adults have more flexible hours for completing chores and reading books, and home-based businesses remain open longer in the evenings, producing more items for sale. Once rural families connect to the grid, television sets, fans, and an array of other household appliances gradually become more affordable (Barnes 2014). 2. The SE4ALL initiative was launched by the United Nations (UN) in 2011. It is co-chaired by the UN Secretary General and World Bank Group (WBG) President; SE4ALL helped place energy access explicitly on the global development agenda, thus filling the gap left by the Millennium Development Goals (MDGs), which did not include any energy access goals. 3. Authors’ calculation from the World Development Indicators (WDI) database. 4. Korwama (2011) estimates that 30 percent of Sub-Saharan Africa’s agricultural produce is processed, compared to nearly 98 per- cent in some developed countries. 5. Focusing on the enabling environment, WBG (2016) measures regulations that impact firms in the agribusiness sector. It collects and reports data on 18 indicators for 40 countries across the world; the indicators capture aspects related to production of inputs and market enablers to help policy makers better understand barriers to growth and transaction costs imposed by the regulatory environment. 14 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa 6. Africa’s economy has been expanding at a relatively high rate. Following a very strong decade from the beginning of this century, growth in 2015 was more modest, at 3.7 percent (World Bank 2015). Growth rates over the next several years are projected at well above 4 percent. 7. About two-thirds of this area is spread over eight countries: Angola, Democratic Republic of the Congo (DRC), Madagascar, Mozambique, South Sudan, Sudan, Tanzania, and Zambia (World Bank 2013; Deininger and Byerlee 2011). 8. The impact of electricity will be lower in areas that use gravity-fed irrigation since the value added by electricity is likely to be rela- tively minor. The main impact will be realized by farmers using agricultural pump sets or other forms of mechanized irrigation. 9. A recent WBG study states that electricity access is critical to promoting a more commercialized agriculture sector in the devel- oping world; it emphasizes the importance of rural electrification as an enabling condition for agribusiness development, and discusses indicators on electricity access, reliability, and affordability (WBG 2015). 10. In the ABC model, anchor customers are the main off-takers for the generated power. Business refers to small local businesses and shops; community refers to households, farming needs (including irrigation), and local institutions. Power Needs of Agriculture Chapter 2 A gricultural transformation in Sub-Saharan to higher value urban and export markets. An increase Africa implies a shift away from household in an irrigated area to reach its estimated potential and subsistence farming toward a more market-­ improving existing irrigation practices will require electric- oriented farming sector that is effectively ity for water pumping. The mechanization of basic milling able to supply demand across the world. Achieving this or grinding that is largely done manually will require elec- transformation involves increased use of modern farm- tricity to run machines. Storage of high-value perishables ing inputs, greater value addition through post-harvest awaiting transport to demand centers will require electric- processing, and access to markets through transportation ity for chilling; and such processing activities as pulping, and storage. drying, heating, and packaging will also demand electricity. Electricity is a key input required to create greater This chapter explores the synergy between agricul- value added in the agriculture sector through enabling tural growth and rural electrification and provides initial irrigation, processing, and storage. Growth in agricul- estimates of power demand from agriculture in 2030. tural incomes is directly dependent on farmers’ ability The value generated by agricultural activities that demand to increase their yields through irrigation, processing of electricity can help tip the scales of commercial viability produce to retain a greater proportion of the value added of rural electrification interventions. along the full supply chain, and proper storage of produce to prevent spoilage. A growing agriculture sector will thus produce greater demand for electricity along its value Power Needs across the chain, from both on- and off-farm activities. Agricultural Agriculture Value Chain transformation, through increasing rural electricity demand, can thus go hand-in-hand with an expansion in Electricity input is vital for the adoption of modern rural electricity access. productivity enhancing technologies and thereby the A structural shift in agricultural markets is set to integration of small-scale farming into high-value and induce demand for electricity from agriculture. With export-oriented value chains. The implications for elec- growing domestic and export markets for agricultural tricity demand from such a shift in the agriculture sector products, the need for increased agricultural productivity of Sub-Saharan Africa will be determined by the extent to will necessitate greater on- and off-farm mechanization which modern techniques are adopted at each stage along of agricultural and agribusiness practices. In addition, eco- the value chain and the scale of each activity. In addition nomic growth is set to create markets for new products to electricity requirements, the potential of various crops and higher value commodities for urban markets and as to gain from irrigation and processing activities can vary intermediate inputs for various industries, especially in the widely. food sector. Depending on crop characteristics and target mar- Electricity demand from agriculture stems from the kets, value chains differ in post-harvest processing and various processes along the agriculture value chain—from preservation requirements. This creates differing on- and on-farm irrigation and off-farm grain milling to larger sec- off-farm demand for electricity for each value chain. ondary processing (e.g., pulping and packaging) that caters In order to examine the nature of electricity use along 15 16 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 2.1: Power needs across Powered irrigation systems are prevalent in Southern agriculture value chains and East Africa, and are emerging in West Africa. To a large extent, West Africa and the Sudano-Sahelian region utilize small-scale irrigation systems, which tend to be Post-harvest & gravity fed. Secondary On-farm primary processing Like other powered activities in agriculture, the processing electricity requirements of powered irrigation equipment depend on system scale, form of irrigation, and specific geographic conditions—the latter factor making it difficult • Irrigation • Milling, drying, • Packaging, to develop accurate estimates of electricity use for irri- chilling, etc. bottling, etc. gation. The two primary power demands for irrigation are (i) sourcing bulk water from some water body, such as a Rural Urban/peri-urban dam or river and (ii) distributing it over the cultivated area. Irrigation systems commonly used in Sub-Saharan Africa range in scale from manual systems to surface agriculture value chains, the sources of growth in future flooding and localized systems to center pivots. Manual electricity demand can be divided into three sources, as systems, including simple buckets to support small-scale follows (figure 2.1 and annex C): farmers, require no power. Surface flooding and localized systems (e.g., stationary drip schemes and pressurized ºº The potential for expanded irrigation, which is the pri- systems, such as sprinklers1) require power to source the mary on-farm source of electricity demand. bulk water that cannot be accessed by gravity only. Center ºº The potential growth in post-harvest and primary pivots may require power for bulk water supply, as well processing activities from both new and existing as for pressurizing water for the system and possibly for production; activities include cleaning/drying, milling, system mechanics (e.g., motors to rotate the pivot span). cassava processing (chipping), chilling and cold stor- In all four cases, power demand is related to system age, meat processing, and oil extraction. scale, but will vary per unit of area covered. In each case, ºº The potential growth in secondary processing activ- pumping bulk water comprises the major demand and will ities that cater mainly to urban markets and provide depend on the vertical and horizontal distances of the intermediate inputs to other production processes; scheme from the water source (table 2.1). activities include thermal treating, canning, bottling, For irrigation systems that use gravity to redistribute and packaging. water, power may only be required for bulk water pump- These several activities are presented in decreasing ing into storage (if needed). The most efficient pumping order of rural presence. Virtually all irrigation occurs systems do this to meet infield demand, running nearly in rural areas, and post-harvest and primary processing continually. But some systems may design their capacity usually occur shortly after the rural harvest, depending with larger pumps so as to require pumping for fewer on scale. Secondary processing is more likely to take hours within a day. This design is inefficient from the view- place near trading hubs and demand centers in urban or point of electricity supply, as it would require a greater peri-urban areas, although, under appropriate conditions, peak generation load. some smaller-scale operations can be viable in rural areas. Benefits from irrigation come from increased yields The prevalence of irrigation potential in rural areas and and reduced weather-related risks. Enhanced irrigation the benefits across value chains imply that irrigation is the practices may thus result in large benefits from increased largest potential source of power demand from agriculture crop yields, leading to higher farm revenues. Giordiano in Sub-Saharan Africa. et al. (2012) find that Sub-Saharan Africa has considerable area under small-scale irrigation or improved agricultural water management. The study estimates that investments Irrigation Potential in dry-season irrigation for rice could potentially increase The irrigation intensity in Sub-Saharan Africa is the lowest yields by 70–300 percent. The same study estimates that in the world; only 6 percent of the region’s cultivated land investment in relatively low-cost motorized pumps, ben- is irrigated, compared to 44 percent in Asia (FAO 2005). efiting 185 million across the Sub-Saharan Africa region, Irrigation intensity and technique vary across the region. could generate net revenues of up to US$22 billion a year. Power Needs of Agriculture 17 Table 2.1: Power Demand for Irrigation, by System Type Cultivation Estimated Power System Methods Power Demand/Unit Typical Area Type Supported Crops Supported Components (kW/ha)a Coverageb Surface flooding Small- and large- Rice, sugarcane, Possibly bulk water, 0.5–0.9 600 m2– (furrow and paddy scale commercial. tomatoes, infield pumping. 20,000 ha systems) citrus. Micro irrigation Small-scale Floriculture, Possibly bulk water, 0.5–0.9 600 m2– (drip and trickle) and intensive horticulture, infield pumping. 20 ha schemes commercial. seedling propagation, citrus, vegetables, potatoes. Micro jet irrigation Some small-scale, Floriculture, Possibly bulk water, 0.5–0.9 5–50 ha mostly large-scale horticulture, citrus, infield pumping. commercial. macadamia, some tree crops. Portable impact Small- and large- Floriculture, Possibly bulk water, 0.5–0.9 600 m2– sprinkler systems scale commercial horticulture, grain infield pumping. 20,000 ha (drag-line and (broad-scale). crops, tobacco, hand-move) bananas, sugarcane, potatoes. Center Small- and large- Wheat, barley, Possibly bulk water, 0.7–2.2 9–150 ha pivot scale commercial soya, maize, infield pumping. (65 ha per pivot is (broad-scale). groundnuts, typical on farms of sorghum, paprika, 50–5,500 ha) tobacco, sugarcane, rice. Source: ECA and Prorustica (2015). Note: The categories provided in this table are general as no two schemes are identical. a. Assumes an average distance of 300 m from the water source to the irrigation scheme. b. Indicates the system scale commonly seen in Sub-Saharan Africa. Irrigation offers distinct seasonal advantages for crop and farming practices. Despite this, multi-cropping, along production as it can help overcome rainfall variability and with the nearly constant need of water supply for efficient even temperature extremes by maintaining adequate cropping (especially under drip irrigation), does reduce levels of soil moisture year round. In the summer, the seasonal variation to a certain extent. primary advantages are greater reliability of water supply Africa’s grossly underutilized agricultural potential (i.e., reducing the impact if rainfall is less than expected) should be tapped by significantly growing the area under and the ability to plant crops early without waiting for cultivation to cover most economically viable areas. You rains. In the winter, when rains are not expected, irrigation et al. (2009) developed estimates of potential increase is indispensable for cropping, allowing for the production in irrigable area in the region using detailed topographical of wheat and other winter crops and more crop cycles per data and economic parameters (figure 2.2). The study year for rice. Therefore, annual use of irrigation allows found that both large- and small-scale irrigation schemes year-round cropping. can be economically developed in Africa, with economic The extent of irrigation and the associated electricity internal rates of return (IRRs) exceeding 12 percent.2 is likely to be characterized by some amount of seasonal- Investments in irrigation over this cut-off could poten- ity. The magnitude of the seasonal variation in irrigation tially increase irrigated areas by 7.7 million ha, with 5.8 depends on crop choice, weather variations, and irrigation million ha coming from small-scale schemes. 18 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Countries with the greatest potential for large-scale All of these countries have more than 100,000 ha of investment are Ethiopia, Mali, Mozambique, Nigeria, potential, based on existing or projected development of Sudan, Tanzania, Zambia, and Zimbabwe (You et al. 2009). mainly multipurpose water-storage reservoirs. Except for Southern Africa, small-scale irrigation projects in Sub- Saharan Africa are generally estimated to have higher IRR Figure 2.2: Potential new or than large-scale irrigation. This implies that economically rehabilitated irrigable land in Sub- viable, small-scale irrigation projects could increase in Saharan Africa area under irrigation to a greater extent than large-scale 2,000 projects (table 2.2).3 By far, the greatest potential is found in Nigeria, which accounts for more than 2.5 million ha or nearly half of suitable hectares. Such countries as Cameroon, Chad, Irrigation potential (’000 ha) 1,500 Ethiopia, Mali, Niger, South Africa, Sudan, Tanzania, Togo, and Uganda each has at least 100,000 ha of potential. To begin to tap this potential, the CAADP Program 1,000 for Investment in Agricultural Water targets region-wide expansion of the irrigated area by 3 million ha, approxi- mately doubling the current area by 2030 (World Bank 500 2013). In some areas, this expansion could be carried out even more quickly: the World Bank’s proposed Sahel Irrigation Initiative has a goal of “doubling the irrigated 0 areas in Sahel in five years through improved public n rn rn ea policies and increased private-sector involvement.” Much l lia tra he te in he n Gu s ut Ea Ce a of this irrigation would be gravity fed, but some of it, So -S of no lf especially small-scale irrigation, would require pumping da Gu Su for transport and/or extraction. And there is an additional Large-scale Small-scale Rehabilitation synergy: the development of hydroelectric power sources Source: You et al. 2009. can often be combined with irrigation projects. Table 2.2: Potential Investment Needs for Large-Scale, Dam-Based and Complementary Small-Scale Irrigation Schemes in Sub-Saharan Africa Large-scale Irrigation Small-scale Irrigation Increase in Investment Average Increase in Average Irrigated Area Cost IRR Irrigated Area Investment Cost IRR Region (million ha) (million US$)a (%) (million ha) (million US$)a (%) Sudano-Sahelian 0.26 508 14 1.26 4,391 33 East 0.25 482 18 1.08 3,873 28 Gulf of Guinea 0.61 1,188 18 2.61 8,233 22 Central 0.00 4 12 0.30 881 29 Southern 0.23 458 16 0.19 413 13 Indian Ocean Islands 0.00 0.00 n.a. 0.00 0.00 n.a. Total 1.35 2,640 17 5.44 17,790 26 Source: You 2008. Notes: The average value for IRR was weighted by the increase in irrigated area. Benin, Chad, and Madagascar have no profitable, large-scale irrigation; n.a. = not available. a. These estimates are one-time investment costs rather than annualized figures. Power Needs of Agriculture 19 Primary and Secondary Processing Aggregate Electricity Demand from Irrigation and Processing Electricity is a vital input in value-added processing activities, such as post-harvest cleaning and drying to By 2030, we estimate that electricity demand from remove moisture and prevent spoilage (e.g., for cereals agriculture could double from today’s level, reaching and legumes), milling (e.g., of maize, rice, and cassava), about 9 GW. This is a simplified estimate as the varied and crushing. Specific processing activities for high-value nature of product value chains and associated irrigation, agricultural products also rely on electricity inputs (e.g., processing, and storage activities makes it impossible wet-processed coffee using machinery for pulping). to develop a comprehensive, region-wide estimate. The Furthermore, electricity can improve storage of pro- demand emerges from considering the potential increase duce through cold chains, thereby reducing income loss in irrigation and post-harvest activities. Assumptions from spoilage and increasing the ability to specialize in about increased development of irrigation and processing high-value perishable products (e.g., dairy, meats, fruits, potential, unit electricity use, and accompanying growth and vegetables). It is estimated that about 30 percent in crop yields underlie this estimation. Growth in agricul- of agricultural produce is wasted due to spoilage. Cold tural production catering to domestic and export demand storage and drying can reduce this figure substantially. and accompanying movement up the agriculture value Electric fans for air precooling, ice-making machines chain are expected to increase electricity demand from and hydro-coolers can improve cooling efficiency in cold irrigation and post-harvest processing. storage rooms. By 2030, about 3.1 GW in additional electricity Though difficult to estimate accurately due to the demand is expected from the development of irrigation dispersed potential, primary and secondary processing potential across Sub-Saharan Africa (figure 2.3). Given represent a significant growth area in Sub-Saharan Africa. the region’s significant underutilized water resources, The expected demand growth for grain milling is likely along with the ubiquitous benefits from irrigation across to increase significantly (e.g., maize in Nigeria, wheat in most value chains, it is expected that irrigation will Zambia, and rice in Tanzania). Similarly, increased demand account for a significant portion of electricity demand for processing of cassava—a widely produced and con- from the agriculture sector.5 The estimated demand sumed staple in many countries (e.g., Angola, Democratic from irrigation is based on fully exploiting potential areas Republic of the Congo, Mozambique, Nigeria, and for new or rehabilitated irrigable areas, totalling nearly Uganda)—is expected due to its perishable nature and use 6.8 million ha.6 This area is dominated by small-scale as an industrial input in the manufacturing of glue. scheme development in the Gulf of Guinea (with more Additional primary and post-harvest processing (if than 1.5 million ha in Nigeria alone) and rehabilitation of developed to full potential), together with the activities existing schemes in the Sudano-Sahelian region (with over discussed above, could significantly change the rural elec- 1 million ha in Sudan) (table 2.4).7 tricity markets. Table 2.3 summarizes the various activities Figure 2.3 shows that about an additional 1.1 GW is that can serve as anchor loads for rural electrification, expected from the development of the region’s agro-pro- along with the value chains they are part of and examples cessing potential. Power demand from the development of countries where they are present and likely to grow. of agricultural processing activity is based on increased The creation of opportunities for viable rural electri- growth in both primary crop production and the propor- fication on the back of local agricultural development will tion of crops that are processed. Currently, the percent- depend on various site-specific factors, including the scale age of crop production processed through electrified value and profitability of agricultural operations, crop, terrain, chains is quite low (conservatively estimated at 10 per- type of processing activity, and other local conditions. cent). By 2030, this percentage is expected to grow to Rural electrification opportunities will be best served 15 percent as a result of the increased participation of by agro-processing activities that generate electric- small-scale farmers in formal value chains. ity demand close to rural population centers, generate Given the varied nature of processing activities by enough income to cover electricity supply costs, are type, scale, location, and technology, the estimate is based sufficiently large in relation to household demand,4 and on the electricity requirement of a typical processing have relatively low seasonal variation. 20 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 2.3: Key Power-Intensive Agribusiness Activities Value SSA Countries/ Scale of Growth Potential Chains Regions Where Power Demand/ of Value Chain and Activity Supported Activity Occurs Supply Activity New large- Maize, rice, wheat, Most countries Single areas can demand Many areas that can be supported scale irrigation oilseed, sugarcane, tea, > 15 MW of capacity. likely to require farms of > 250 ha; floriculture crop choice depends on market prices. Substitute Maize, rice, cassava, Most countries In unconnected rural Many towns in agricultural areas power for oilseed towns, demand unlikely will have this demand; risky as diesel in small- to exceed 500 kW for anchor load for electrification. scale milling the whole town. New large- Maize, rice, wheat, Most countries Demand can be Widespread opportunity. Reliant scale milling oilseed, sugarcane, oil > 1 MW from a single on base supply from commercial palm, tea, cotton mill. estates; crop choice depends on market prices. Milking and Dairy Few countries > 800 kW peak demand. Small markets in SSA; climatic cold storage conditions not ideal for dairy farming. Cold storage Floriculture, export Ethiopia, Kenya, and 10 MWh/ha per year. Continued demand for floriculture vegetables Uganda (floriculture in Europe, leading to agribusiness and export vegetables); growth in select countries; Rwanda and Tanzania challenges with horticulture (export vegetables) through demand for high quality, competitive retail markets driving down margins and tariff restrictions in European markets. Biomass-fueled Rice, oil palm Many countries (rice); Can provide > 10 MW of Beyond Africa, export market for generation West Africa and East power (ha/ton). rice is challenging and unreliable and Southern Africa for palm oil. Water intensity restricts locations; depends on reliable supply of biomass from commercial estates. Bagasse-fueled Sugarcane Eastern and Southern Can provide > 10 MW of Large market for crop, but generation Africa (South Africa) power (70 kWh/MT of price-dependent. Water intensity sugarcane, or restricts locations; depends on 243 kWh/MT of reliable supply of bagasse from bagasse). commercial estates. Source: ECA and Prorustica (2015). Power Needs of Agriculture 21 Figure 2.3: Estimated electricity demand activity (milling) and thus does not capture the electricity (MW) from agriculture for Sub-Saharan demand from the potential development of other process- Africa in 2030 ing activities or storage. In 2012, the Food and Agriculture Organization of the United Nations (FAO) estimated crop production at about 852 million metric tons (MT). Assuming a growth rate of 2.4 percent annually (Alexandratos and Bruinsma 2084 2012), crop production would reach 1.3 billion MT by 2030 (table 2.5). The power demand from crop produc- tion is estimated by assuming that processed crops will consume, on average, the amount of power needed for an 978 average wheat mill in Zambia—some will have greater con- sumption and others less. This “average mill” is assumed 6915 to handle 8 MT per hour, operating year round at 16 hours per day and 6 days a week. This would result in approxi- 3786 mately 40,000 MT per year and have a power capacity demand of 400 kW. The total estimated electricity demand from agricul- 2015 2030 ture is indicative of the scale of the opportunity for rural Irrigation Processing (milling) electrification to benefit from agricultural growth poten- Source: ECA and Prorustica (2015). tial. The overall magnitude of electricity demand provides Table 2.4: Method for Calculating Power Demand from Irrigation Prominent Countries Estimated Estimated Estimated with Irrigable Area Proportions/ Power Use Power Use Category (thousand ha) Power Use (kW/ha) (MW) Large-scale Ethiopia (191) Much of East and Southern Africa 1.2 kW/ha for 1,285 Nigeria (609) requires bulk-water pumping, West area requiring Sudan (238) Africa less so; 50% requires bulk- bulk water, Zimbabwe (142) water pumping and 50% just infield 0.7 kW/ha Total = 1,352 equipment. otherwise Small-scale Cameroon (170) Most schemes are very basic in riparian 0.7 kW/ha for 1,051 Chad (231), Mali (219) areas; 40% requires power, and 60% is area using power, Nigeria (1,538) entirely gravity fed with no power. 0 otherwise Tanzania (196) Uganda (445) Total = 3,754 Rehabilitation Somalia (135) Most rehabilitation consists of gravity 0.7 kW/ha for 793 Sudan (1,064) fed, colonial-era schemes; 10% is large- area using power, Total = 1,688 scale with bulk water, 30% large-scale 0 otherwise without, 20% small-scale with power, 40% small-scale with no power. Sources: You et al. (2009); ECA and Prorustica (2015). 22 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 2.5: Power Demand for Crop Processing Primary Processed Number of Crop Production Crops Processed Production 400-kW Total Power Year (million MT) (%) (MT) Mills Required Demand (GW) 2012 852 10 85.2 2,129 0.851 2030 1,306 15 196.0 4,893 1.960 Sources: FAO; Alexandratos and Bruinsma (2012); ECA and Prorustica (2015). a sense of the investment in generation capacity that will agriculture sector. The latter informs the likely viability of be required to meet agricultural needs and the addition accounting for agricultural growth in rural electrification to rural electricity demand that is expected, owing to the strategy and planning. endnotes 1. Some sprinklers are pressurized, while others are solely gravity operated. 2. Conditional on having initial investment costs at best-practice levels and if market access, complementary inputs, extension of credit, and a supportive policy and institutional environment are in place. 3. The higher IRR for small-scale irrigation is due to the existence of large amounts high-potential rainfed cultivation located far from large-scale developments that could be profitably converted into small-scale irrigation (You et al. 2009). 4. Although even a relatively small agricultural load can potentially help to push aggregate demand in a given area over the threshold of economic and financial viability. 5. In the context of climate change, the future availability of water will depend critically on improvements in water management practices and planning (box 1.3). World Bank (2016a) predicts that, under business as usual, water management in Southern and East Africa will not experience negative effects on GDP, while other parts of Sub-Saharan Africa could experience about a 6 percent fall in GDP in 2050. 6. You et al. (2009) classifies areas based on their anticipated IRR on irrigation investment. The numbers reported here are based on an anticipated 12 percent return, which is a typical benchmark for such projects. 7. You et al. (2009) was published before the independence of South Sudan and thus classifies the whole of Sudan together. Power Needs in Selected Value Chains Chapter 3 A gricultural production in Sub-Saharan Africa The need for post-harvest electricity input varies, is fairly diversified, and no single cereal crop depending on the nature of the crop, the type of value predominates across the region. In terms chain (or targeted market) and local conditions. A case in of production quantity, maize is the most point is Kenya’s dairy sector: 86 percent of the country’s important, followed by sorghum, millet, and rice; the milk supply is driven by small-scale farmers and small- and importance of each crop varies by individual countries. medium-sized enterprises (SMEs), with milk being sold In West and Central Africa today, cereals comprise less to small-scale vendors. Parallel to this, larger dairy farms than 20 percent of agricultural value added (compared with either integrated dairy herds and/or formal links to to 35 percent for Asia prior to the Green Revolution), dairy farmer cooperatives provide pasteurized milk and with the remainder coming from other staples (especially processed dairy products via cool chains for sale to higher roots and tubers), horticulture, export crops, and livestock income urban consumers through supermarkets (World (Schaffnit-Chatterjee 2014). Bank 2013). Owing, in part, to diversity in agricultural production, This chapter examines potential electricity use along agriculture value chains also vary widely across the region 13 selected value chains. Electricity demand from on-farm and even within countries. Value chains vary by length, activities and rural processing presents an opportunity for technologies utilized, value added, and markets served.1 the development of anchor loads to spur rural electrifica- Many value chains operate in both informal and formal tion. The source of electricity may vary on a case-by-case markets, with the former catering to low-income, domes- basis, and opportunities for biomass based generation tic consumers and the latter catering to higher income for particular value chains (e.g., oil palm and sugar) are urban and export markets (World Bank 2013). highlighted. In addition, bottom-up estimates of potential The value chains for the region’s bulk commodities future electricity demand from the selected value chains (e.g., maize and rice) are primarily informal, in contrast to are presented. more market-oriented, semi-processed and consumption ready products. As a commodity moves along the value chain to the ultimate market and consumer, hygiene and Selection of Value Chains quality standards become more stringent. Such commod- ities as sugar, tea, and oil palm are processed virtually at The value chains selected for this study help illustrate the the point of primary production, while other commodities nature of electricity demand from the rural agriculture (e.g., fruits, vegetables, and livestock products) must be and agribusiness sectors, along with the power-demand processed within a relatively short period before they profile. These value chains represent both high growth deteriorate. Still others have parallel value chains; that potential and the ability to create electricity demand is, for the same commodity, some value chains focus on for irrigation and/or processing in rural areas (table 3.1). lower end consumers in domestic markets, while others The potential for agricultural electricity demand extends are more formal, with strong processing and stringent well beyond the value chains discussed here and is often quality control. driven by site- and country-specific factors that create 23 24 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 3.1: Analysis of Commodity Value Chains, by Scale and Region/Country Commodity Scale (if applicable)a Region/Country Maize Small and large East and Southern Africa Rice Small and large Tanzania (primarily) Cassava Small West, Central, East, and Southern Africa Wheat Large Southern Africa Oilseed Small (primarily) East and Southern Africa Horticulture (pineapple) Small and large West, Central, and Southern Africa Sugarcane Small and large East and Southern Africa Oil palm Small and large West and Central Africa Dairy Small and large Kenya Poultry Large East and Southern Africa Tea Large East and Southern Africa Floriculture (roses) Large East Africa Cotton Small West, East, and Southern Africa Source: FAOSTAT (http://faostat3.fao.org). a. Farming systems are defined in terms of labor type and not merely scale. Large-scale commercial farming is defined by family labor that is predominantly managerial, with full-time labor hired for specific tasks and production catering to market supply. opportunities along other crop and processing activities. large existing potential on the extensive (area expansion) The case studies presented in chapter 4 analyze examples and intensive (yield growth) margins. of such opportunities. According to future production estimates, cassava The commodity value chains shown in table 3.1 were and maize—primary staple food crops in the region—will selected according to the following criteria. Starting with remain dominant over the period until 2030. Sugarcane, a the top 20 commodities by production value for 2012 well-established industry with conducive growth con- (from FAOSTAT), the list was modified to assure the inclu- ditions, is also expected to remain dominant across the sion of (i) key export commodities (e.g., tea, cotton, and region for the foreseeable future. In addition, recent high horticulture); (ii) value chains based on assessed electricity growth rates of cotton, pineapple, and rice suggest that use; (iii) commodities with large production volume and these commodities will likely gain greater regional impor- importance for local food markets with potential for future tance in the coming decades. growth in processing requirements (e.g., cassava and maize); Cassava. In terms of production quantity, cassava is (iv) commodities that figure in the top ones by value for Sub-Saharan Africa’s most important crop, accounting for many countries in the region (e.g., tea and soybean), which more than half of global production. Nigeria is the leading may not appear on a region-wide list; (v) commodities global producer, followed by the Democratic Republic of with large irrigation schemes (e.g., irrigated wheat); and the Congo (DRC), Angola, Ghana, and Malawi.2 Cassava (vi) value chains with the potential to supply fuel for elec- is experiencing growing demand as a staple food crop and tricity generation (e.g., oil palm and sugarcane). an intermediate input into various other commercial value Table 3.2 shows the estimated production volume chains (e.g., starch and livestock feed). The crop is still for the selected commodities in 2030, along with their mainly grown under small-scale farming conditions with estimated average annual growth rates between 2013 limited use of irrigation. Owing to its drought tolerance and 2030. Future projections are calculated using the and ability to grow in relatively poor soils, production is historical growth rate (between 2009 and 2013) for fairly widespread in rural areas across the region. Further each commodity (FAOSTAT) and applying a concavity development to make the crop’s value chain more market parameter to project a declining growth rate over time. oriented can have large effects on the livelihoods of small The assumed growth rates are qualitatively more conser- farmers. Growth in cassava production depends critically vative than those assumed by Alexandratos and Bruinsma on improved processing and drying of roots to reduce bulk (2012), who predict mostly convex growth rates, owing to and prevent deterioration. Power Needs in Selected Value Chains25 Table 3.2: Comparison of Historical and Projected Commodity Growth Rates and Estimated Production Assumed Average Growth Rate, Annual Growth, Estimated Production Projected Production 2009–13 2013–30 in 2013 in 2030 Commodity (%) (%) (million MT) (million MT) Cassava 6.4 2.8 157.7 252.7 Maize 5.8 2.5 65 101.2 Sugarcane 1.7 0.8 73.9 84.6 Rice (paddy) 5.9 2.6 22.6 35.5 Wheat 5.1 2.3 7.1 10.6 Pineapple 9.5 4.2 4.4 9 Dairy 1.6 0.7 3.2 3.6 Poultry 1.5 0.6 2.7 3 Cotton (lint) 8.1 3.5 1.3 2.5 Oil palm −0.7 −0.3a 2.4 2.2 Tea 5.4 2.4 0.7 1 Oilseed (soybean) 2.6 1.2 0.5 0.6 Sources: FAOSTAT and World Bank estimates. a. The oil palm industry is now considered less attractive; some developments are proving unsustainable and are being converted to other uses. Maize. Due to its tolerance of diverse climates, South Africa and Mozambique lead in terms of area under maize is one of the world’s most widely grown crops. In cultivation (table 3.3). Eighty percent of the world’s sugar 2013, total global production was estimated at more than is produced from sugarcane, while the other 20 percent 1 billion metric tons (MT). In Sub-Saharan Africa, maize is from sugar beet (FAO 2009). The most common pro- is one of the most prevalent cereals, with more than duction model is contracting commercial and small-scale 65 million MT produced in 2013 (table 3.2). However, the outgrowers to supply the sugar refineries. region’s average yield of 1.4 MT per ha is low compared Rice (paddy). Sub-Saharan Africa has witnessed to the global average of 5 MT per ha, and 11.6 MT per ha rapid growth in rice production, driven mainly by urban- in the United States (Iowa) (2009 figures, FAO). A few ization. The compound annual growth rate (CAGR) of countries are dominant in maize production, but their domestic production has averaged about 6 percent, with market share is less pronounced. Maize’s utilization is wide more than 22 million MT reached in 2013. According to ranging; it serves as a leading food staple and important the Africa Rice Center’s analysis, the region’s rice yields feed crop, as well as an input in the processing of food, have increased in real terms by an average of 108 kg chemicals, and fuels (ethanol).3 In East and Southern per ha annually, comparable to the Green Revolution’s Africa, maize is principally a food staple, accounting for growth rates in Asia (Seck et al. 2013). Despite such rapid 30−50 percent of low-income household expenditure.4 growth, rice imports have also increased significantly; As such, growth in production is expected to increase, in 2012, 12 million MT were imported. The region has propelled by growing regional demand. considerable potential for production growth through Sugarcane. According to the FAO, sugarcane is the increasing the area under cultivation and increasing yields. world’s largest crop in terms of production quantity, with Wheat. Among all cereals, wheat is the most highly 1.83 billion MT produced in 2012. Brazil is its largest traded. As of 2013, it was the world’s third most widely pro- producer, followed by India. Sub-Saharan Africa accounts duced cereal (behind maize and rice), at a total of 713 mil- for roughly 4–5 percent of global production, with about lion MT.5 In Sub-Saharan Africa, Ethiopia and South Africa 74 million MT produced in 2013. The region’s largest are the main wheat producers. Generally, production has producers are South Africa, followed by Sudan and Kenya; not kept pace with the region’s growing demand for wheat; 26 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 3.3: Countries in Sub-Saharan Africa with Similar Commodity Production and Processing Systems Commodity Countries Maize Kenya, Malawi, Mozambique, Tanzania, Zambia, and Zimbabwe; also Burkina Faso, Ghana, Mali, and Nigeria (but not at such large commercial volumes) Rice Madagascar and Tanzania Small-scale cassava Angola, DRC, Mozambique, Nigeria, Tanzania, and Zambia Irrigated wheat Zambia and Zimbabwe Rainfed wheat Ethiopia and Kenya Commercial soya Zambia and Zimbabwe Sugarcane Ethiopia, Kenya, Malawi, Mozambique, South Africa, Swaziland, Tanzania, and Zimbabwe Oil palm Cameroon, Côte d’Ivoire, and Ghana Dairy Kenya, Ethiopia, Rwanda, South Sudan, and Uganda Poultry Kenya, Malawi, Zambia, and Zimbabwe Tea Kenya, Malawi, Rwanda, and Uganda Floriculture (roses) Ethiopia, Kenya, Tanzania, Uganda, Zambia, and Zimbabwe Cotton Benin, Burkina Faso, Côte d’Ivoire, Mali, Mozambique, Tanzania, Uganda, Zambia, and Zimbabwe Source: ECA and Prorustica (2015). thus, wheat imports have been on the rise. Among the a minimum, suggesting that dairy storage and processing region’s handful of countries that are fully self-sufficient centers are located in the vicinity of dairy farms. in wheat production, Zambia is noteworthy; that coun- Poultry. Population growth, changing diets resulting try’s annual production, mainly commercial in scale, totals from urbanization, and income growth are the major 300,000 MT (table 3.3).6 Many parts of East, Southern, drivers of Sub-Saharan Africa’s ongoing demand for and Central Africa are suitable for wheat production. poultry. During 2000–11, poultry (meat) production Pineapple. In Africa, horticulture, in the form of trop- across the African continent grew by 5 percent per year, ical fruit production, caters mainly to own consumption reaching 4.62 million MT in 2011. Major producers are and domestic markets; in some countries, it also caters to in Northern Africa: Egypt, Algeria, Morocco, Libya, and Europe and other export markets (e.g., canned fruits and Tunisia. In Sub-Saharan Africa, 2013 production totaled pulp). After banana, pineapple is Sub-Saharan Africa’s 2.75 million MT, with South Africa and Nigeria as lead most important tropical fruit. Nigeria is the region’s producers. These two countries are also the region’s major largest pineapple producer. Kenya, the second largest, egg producers; and hatcheries are usually large-scale com- ranks among the world’s top five exporters of pineapple; mercial operations. Modern poultry complexes are usually canned pineapple, exported mainly to Europe, is its largest integrated with chicken farms to reduce the costs associ- manufactured export. ated with the transport of live animals. Contract farmers Dairy. The robust growth in dairy production reported receive chicks from the hatchery, ideally housing them in many parts of Sub-Saharan Africa today is being driven in climate-controlled chicken houses. Broiler processing by economic growth and urbanization. Traditionally, milk operations are typically located on-site at poultry farms. has been produced for own consumption or local con- Cotton (lint). Cotton is one of Africa’s main cash sumption by farmers; however, growing urban demand crops among small-scale farmers. In 2013, Sub-Saharan is increasing the need for cold supply chains to maintain Africa produced 1.3 MT of cotton (lint) (table 3.2). The product quality. According to the FAO, the region’s dairy region’s major producers are Burkina Faso, Mali, Côte production totaled 3.2 million MT in 2013. Along with d’Ivoire, Benin, and Zimbabwe. In West Africa, Burkina this demand growth is the demand created for process- Faso and Mali each produce about 400,000 MT per year. ing milk-derivative products (e.g., cheese, butter, and In East and Southern Africa, Zimbabwe is the lead pro- evaporated milk). Transport of raw milk, which is prone ducer, with an annual output of 200,000–300,000 MT to spoilage, is generally uneconomical; thus, it is kept to in seed cotton (table 3.3). Power Needs in Selected Value Chains27 Oil palm. The source of palm oil, one of the world’s high-value agricultural activities, generating revenues of leading edible vegetable oils, oil palm constitutes 60 per- $100,000–200,000 per ha.9 cent of the global trade in vegetable oils (World Bank 2011a). Oil palm fruit yields two distinct types of oils: (i) palm oil, which is edible, used mainly in the form of Electricity Demand vegetable oil and (ii) palm kernel oil, which is extracted and Farming Scale from the seed kernel, used as an input to process other foods (e.g., biscuits and margarine), manufacture house- Electricity demand along the value chain is likely to vary hold products (e.g., soap, shampoo, and cosmetics), and by scale or type of farming operations (e.g., commercial produce biodiesel fuel. Southeast Asia (mainly Malaysia versus small-scale) due to differences in farming processes and Indonesia) produces 85 percent of the world’s palm (e.g., irrigation) and the extent and nature of post-harvest oil. In Sub-Saharan Africa, West Africa is the main processing (box 3.1). While farming in Sub-Saharan Africa producer. Nigeria is the largest producer; however, Côte is predominantly in the form of smallholder agriculture, a d’Ivoire, DRC, Ghana, Guinea, and Uganda are also estab- significant portion of the future potential rests on increas- lishing major operations. While commercial-scale farmers ing yields on such farms by employing more modern account for most production, small-scale farmers also inputs and connecting them to higher value markets and find oil palm an attractive crop since it is relatively high value chains (i.e., employing large-scale operations). yielding and requires limited labor inputs. It is useful to compare electricity needs across these Tea. Tea is one of Sub-Saharan Africa’s most impor- types of agricultural arrangements. The implication for tant export commodities, especially for East Africa. Kenya overall magnitude depends on the evolving proportions is the world’s largest exporter of black tea. In 2011, it of commercial and small-scale farming techniques in the produced 378,000 MT, about two-thirds of Sub-Saharan Africa’s output. Uganda and Malawi are the region’s next two largest producers, while Tanzania and Rwanda are experiencing steady growth in production (table 3.3).7 Box 3.1: Farm Type Definitions Tea-growing usually occurs on large plantations, with processing located either on-site or nearby. Defining farming systems in terms of labor can be Oilseed (soybean). Although Sub-Saharan Africa’s useful, given that the definitions do not depend on soybean production is fairly small by global standards, production scale or crop type. Accordingly, three contributing only 1 percent of global production, the types of farm systems are distinguished here: region’s production is growing faster than the world aver- age (ACET 2013). South Africa has the highest growth in Family farms. These small-scale farms are char- percentage terms, while Nigeria has the largest absolute acterized by the predominant use of family labor, growth.8 Soybean is grown mainly on small farms, while lack of permanent workers, and presence of sea- commercial soybean farming is prevalent in South Africa, sonal labor hired during peak production times. Zambia, and Zimbabwe. Soybean is sold for both human Small investor farms. The owners/family mem- consumption and as an animal feedstock. bers are involved primarily in management and Floriculture (roses). The introduction of rose supervisory roles, while the bulk of labor input is cultivation in Sub-Saharan Africa began in Kenya provided by hired farm workers; this group is less about three decades ago. To this day, Kenya remains defined in Africa, but most, if not all, of their well-­ the region’s main producer and exporter of roses; that crops are produced for market. country also has the highest area under rose cultiva- tion, followed by Ethiopia and Uganda. Rose production Large-scale commercial farms. Family labor for in Ethiopia has been growing rapidly, and the country these farms is exclusively or predominantly mana- is fast establishing itself as a major exporter, to some gerial. A permanent hired staff of full-time work- extent capturing market share from Kenya. Most pro- ers, specialized to a certain degree (e.g., drivers), duction is for export markets, especially Europe, which produces primarily for market. generates more than US$1 billion in export revenues for the region (International Trade Center 2014). On Source: Poulton et al. (2008). a per hectare basis, rose production is one of the most 28 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa region. For example, greater proportional growth in the operating as outgrowers for commercial estates; thus, the adoption of commercial-scale farming, which depends scale of power demand cannot be viewed independent of more heavily on power input, will induce higher overall larger commercial estates.13 The figures include dairy with electricity demand by the agriculture sector. zero values to highlight that informal dairy value chains do Examining typical electricity use for irrigation and not utilize power in Sub-Saharan Africa. processing shows that, for most of the value chains Given the economies of scale in generation capacity, analyzed, irrigation constitutes a large proportion of the commercial agricultural activities are likely to be more potential electricity demand. As small-scale farming financially viable anchor loads to support affordable rural largely relies on rainfed or gravity irrigation, electricity electricity supply to rural Sub-Saharan Africa. However, demand from commercial-scale irrigated agriculture is due to recent technological improvements, accompa- an order of magnitude greater than from smallholder nied by the creation of enabling regulatory conditions, agriculture. Figure 3.1 compares typical rates of power electricity provision in the form of mini-, micro-, and even usage for large-scale irrigated and small-scale rainfed (or pico-grids has dampened the scale economies in electric- gravity fed) irrigation for selected value chains. For the ity generation and distribution investments. Increasingly, most widely grown crops in Sub-Saharan Africa, including advances in renewable energy technologies, such as solar maize, rice, and cassava, irrigation accounts for the highest photovoltaics (PV), are allowing viable electricity infra- potential electricity load.10 structure investments catering to smallholder agriculture As shown, potential peak power loads for small-scale and rural households. Even for more conventional tech- informal production are quite small relative to loads from nologies, ubiquitous small-scale, informal agriculture can commercially irrigated production on a per unit basis enhance the viability of rural electrification on the margin. (figure 3.1b), although this is partly offset by the predom- As discussed earlier, given the diversity of conditions inance of smallholder agriculture across the region, repre- across agricultural areas, site-specific opportunities still senting over 80 percent of the cultivated area (Livingston, exist if cost-effective technologies (e.g., biomass, solar, or Schonberger, and Delaney 2011). small hydro), which may not exhibit strong economies of Though irrigation accounts for a major part of the scale in installed capacity, can be utilized. potential on a per unit basis, post-harvest processing can play a significant role in supporting rural electrification, especially in the case of some commodity value chains. Electricity Demand in the Selected Adding electricity demand for processing to that for Value Chains irrigation, commercially oriented value chains such as sug- arcane, tea, floriculture, and dairy have the overall highest The development of power profiles for each commodity, potential electricity demand (figure 3.1a). Tea is easily the region, and farm type utilized a range of information most power-intensive commodity, with demand ema- sources. Value chains were analyzed in terms of their nating primarily from processing (figure 3.1c).11 Activities nature and magnitude of power use for irrigation and with potentially large loads from processing (sugarcane, processing, growth potential, and ability to serve as an tea, and floriculture) are developed and operated mainly anchor load. by large single entities or organized groups of small-scale To enable comparison, the power profiles presented farmers (see case study 6, chapter 4).12 In such cases, the below are for (arbitrary) standardized farm sizes of power load and potential power supply are usually part of 300 ha, based on the unit electricity demand presented in the planning process; examining options and incentives for table 3.4. The 300 ha benchmark was chosen to rep- rural electrification can be integrated into the planning resent the cultivated area that might constitute a typical stage itself. project site.14 However, in Sub-Saharan Africa most agricultural pro- Maize. For the maize value chain, the input of rural duction occurs in small-scale, informal value chains. The electricity is primarily for irrigation (largely restricted to potential power demand from small-scale agriculture is large-scale farming) and milling (figure 3.2a). The gain in much less than from commercial agriculture. Lower yields value from electricity use comes from the higher yields mean that a larger area is required to produce sufficient resulting from irrigation and the saving of labor and higher production volume for processing facilities. Figures 3.1b productivity resulting from electricity powered (versus and 3.1d exclude small-scale sugarcane, oil palm, and tea manual) milling. The estimated electricity demand from since these typically occur only with small-scale farmers these two activities is about 1.17 kW per ha for large-scale Power Needs in Selected Value Chains29 Figure 3.1: Potential peak capacity and energy demand for large- and small-scale systems a. Peak capacity: Large-scale irrigated production b. Peak capacity: Small-scale rainfed production 2.5 0.12 2.0 0.09 1.5 kW/ha kW/ha 0.06 1.0 0.03 0.5 0.0 0.00 O ne e ce Ca t Pi eeds ga s alm Co a e iry y O va ric n n e va ce ds at a ple Te aiz ur ir aiz o to he ssa Ri ssa rca t Ri Da he Da ee ult t t il p M M ap W Co ils ils W Ca ne O Su Flo Irrigation Processing    Irrigation Processing c. Energy demand: Large-scale irrigated production d. Energy demand: Small-scale rainfed production 1,000 120 750 90 kWh/MT kWh/MT 500 60 250 30 0 0 O ne e ce Ca t Pi eeds ga s alm Co a e ir y y O va Flo tton n e va ce ds t a ple Te aiz ur ir aiz to a he ssa Ri ssa rca Ri Da he Da ee ult t il p M M ap W Co ils ils W Ca ric ne O Su Irrigation Processing    Irrigation Processing Note: Unit electricity demands are constructed from various sources and field observations by ECA and Prorustica. Figure 3.1a does not plot poultry as it is a significant outlier and not feasible to depict on the same scale. Figure 3.1c omits floriculture due to the incomparability of yield data. Figures 3.1b and 3.1d are restricted to those commodities with significant production on smallholder farms (thus omitting such cash crops as tea, sugarcane, floriculture, and horticulture). 30 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2a: Electricity input in the maize value chain Irrigation Drying Crushing/milling • Electricity for pumping water • Mostly solar energy • Milling in town centers as and drip irrigation • Potential for use in certain scale is required green-rated heat from CHP generator production and about 0.77 kW per ha for small-scale and milling is 1.04 kW per ha for large-scale, irrigated irrigated production, suggesting that 300 ha of cultivated production and 0.03 kW per ha for small-scale (paddy) maize will require about 250–350 kW of installed power production with no irrigation. Thus, for a cultivated area generation capacity. of 300 ha, the power demand is in a range of 9–315 kW, Rice. For rice, irrigation and milling are the primary depending on farming type. Additionally, rice husk bio- sources of rural electricity demand (figure 3.2b). Because mass provides a readily available and cost-effective fuel rice can be grown under a variety of irrigated or rainfed source to generate electricity to supply mills and poten- water regimes, electricity demand for irrigation varies tially the neighboring community.15 by type of cultivation. The value gain from electricity Cassava. For cassava, the electricity demand ranges from use is from the higher yields resulting from irrigation (an 0.02 kW per ha to 0.56 kW per ha, depending on whether increase of up to 4 MT per ha) and the value added from the land is under irrigation (figure 3.2c). For a 300 ha culti- milling. The estimated electricity demand from irrigation vated area, the power demand would be about 160 kW. Figure 3.2b: Electricity input in the rice value chain Irrigation Drying/dehusking Milling/polishing • Electricity for pumping water • Mostly manual methods and • Milling in town centers as and drip irrigation solar energy certain scale is required • Potential for use in heating from CHP generator Figure 3.2c: Electricity input in the cassava value chain Irrigation Drying/peeling/chipping Grating/milling • Electricity for pumping water • Mostly manual methods and • Milling in town centers as and drip irrigation solar energy certain scale is required • Potential for use in heating from CHP generator Processing chain: High-quality cassava flour Washing Peeling Washing Grating Pressing Flash drying Milling Power Needs in Selected Value Chains31 Wheat. For winter wheat production, powered demand for electricity. Irrigation for other horticultural activities include irrigation; on-farm drying, cleaning, and crops (e.g., beans, peas, and potatoes) is fairly limited and conveying in and out of silos; and milling (figure 3.2d). usually small in scale. Owing to perishability, electricity is The value added from electricity use is through the higher needed for cooling and to power a cold chain from farm yields from irrigation (an increase of about 4 MT per to market, although this is usually provided in the form of ha) and electric milling and processing. The total power mobile refrigeration units (reefers). The value added from demand from irrigation and post-harvest processing is electricity use in the pineapple value chain includes higher estimated at 1.1 kW per ha for large-scale production and yields resulting from irrigation, increased product value 0.52 kW per ha for small-scale production. For a 300 ha resulting from juicing and canning, and reduced wastage cultivated area, power demand would be in a range of due to cold storage (figure 3.2f).16 The electricity demand 150–230 kW, depending on the farming type. from irrigation is estimated at 0.75 kW per ha for com- Oilseed (soybean). For soybean, the value added from mercial production, implying that 225 kW of power would electricity use occurs through the higher yields made be needed for 300 ha cultivated area. In addition, the possible by irrigation and increase in value from processing by-products of post-harvest processing can potentially (figure 3.2e). The total electricity demand resulting from provide biomass for electricity and heat generation, which irrigation and milling is estimated at 1.04 kW per ha for can significantly reduce power costs.17 large-scale production and 0.64 kW per ha for small-­ scale Sugarcane. Sugarcane yields are highly responsive to production. These figures suggest power demand in a irrigation; thus, water pumping for irrigation is an impor- range of 200–300 kW for a 300 ha cultivated area. tant source of electricity demand in the sugarcane value Horticulture (pineapple). Along the pineapple value chain. In addition, sugar mills constitute considerable chain, juicing and canning activities comprise the main processing demand for electricity (figure 3.2g). The value Figure 3.2d: Electricity input in the wheat value chain Irrigation Drying/cleaning Grinding/milling Figure 3.2e: Electricity input in the soybean value chain Pressing/expelling/ Irrigation Shelling/dehusking Grinding/milling extruding • Electricity for pumping water and • Cleaning/washing, drying, and • Includes heating and grinding • Extraction, using oil expellers drip irrigation storage usually done prior to • Oil would need further processing shelling • By-product is cake used as • Solar power generally used for poultry feed drying Figure 3.2f: Electricity input in the pineapple value chain Irrigation Cutting/juicing Treating/packaging • Electricity for pumping water • Electric machines used for • Thermal treatment and cooling and drip irrigation slicing and juice extraction • Packing and canning and concentration 32 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2g: Electricity input in the sugarcane value chain Irrigation Milling Refining • Electricity for pumping water • Milling: washing, chopping, • Further refining of raw sugar and drip irrigation shredding, and crushing to produced from milling extract cane juice • Usually located near urban • Subsequent clarification, markets concentration, and crystallization to produce mill-white • Biomass by-product used for electricity and heat generation gains from electricity use are derived from the higher used to power factories, with the surplus power exported yields from electricity powered irrigation and the price to the national grid. For both the South Africa sugar differential between raw cane and partially processed industry and Uganda’s Kinyara sugar manufacturer, the sugar. The increased yields from irrigation could reach power output is approximately 30 kWh per MT of crushed 50 MT per ha and even up to 150–200 MT per ha if the sugarcane. latest drip irrigation methods are utilized. On top of the Oil palm. The processing of oil palm usually occurs on value added, maintaining processing activities close to the or nearby the farm due to its bulky nature and ability to farm helps to reduce transport costs. The combined power produce biomass used to generate the heat and electric- demand of irrigation and refining is estimated at 1.81 kW ity required for oil extraction and processing. Oil palm per ha for large-scale production and 1 kW per ha for irrigation is largely rainfed. The main sources of electricity small-scale production. These figures imply that a 300 ha demand are oil processing and extraction from the fresh cultivated area will demand 300–550 kW of power, fruit bunches (FFBs) (figure 3.2h). Though uncommon, depending on the scale of production and related farming drip irrigation can raise yields by 6 MT of FFB per ha. The practices. value gained from using electricity is through processing The biomass residue (bagasse) from sugarcane and reduced transport costs. For milling, the estimated processing has a high potential to generate electricity. electricity demand is 0.02 kW per ha, suggesting a Refineries often produce their own electricity and sell the 6 kW power requirement for a 300 ha cultivated area. excess to the grid. Bagasse generated electricity could Substantial amounts of solid palm oil waste are available become important for the rural populations of sugarcane from the palm oil mills, which are energy self-sufficient; producing nations. For example, in Ethiopia, the Wonchi, that is, they produce their own energy to operate and Metehera, and Finchaa sugar factories produce approx- use the surplus generated to supply estates, sell to the imately 300,000 tons of sugar each year, powering an grid, and possibly sell to villages and towns in the area installed electricity capacity of 62 MW. The electricity is (box 3.2).18 Figure 3.2h: Electricity input in the oil palm value chain Irrigation Oil extraction Refining • Electricity for pumping • Sterilization, stripping, • Refining of extracted crude oil water and drip irrigation digesting, and pressing used • Not necessarily nearby oil palm • Electricity-powered to extract oil extracted from plantations irrigation uncommon in the FFBs Sub-Saharan Africa Power Needs in Selected Value Chains33 processing milk-based products (e.g., butter, cheese, Box 3.2: Palm Oil and Power and evaporated milk). The value gain from electricity use Integration in Uganda results from reduced spoilage due to cold storage,19 the ability to access urban markets, and the value added from One example of an integrated palm oil/power processing milk products. For large-scale operations, the setup is Uganda’s Bugala Power Station, a 1.5 MW estimated power demand is about 0.61 kW per ha. Animal biodiesel-fired thermal power plant located on manure from dairy farms may also be used to generate Bugala Island on Lake Victoria. The power station electricity. is integrated with the palm oil processing plant Poultry. Hatcheries are usually relatively large-scale owned by Bidco Oil Refineries Ltd., which also commercial operations that require electricity input for owns a 6,500 ha palm oil plantation on Bugala a host of processes, including egg incubation and clean- Island. The oil-processing factory generates heat ing. For poultry (meat) production, processing plants use through biomass incineration, used to supply electricity to power conveyor belts, cooling and heating, superheated steam to help extract oil and also and cutting (figure 3.2j). The value added from electricity turn turbines and create electricity in the process. use results from reduced spoilage, increased egg-laying The electricity is used inside the factory, with any productivity, higher labor productivity, value addition from excess sold to neighboring towns. processing, and ability to supply higher value urban mar- kets. The estimated energy demand for ­ commercial-scale broilers (meat) and layers (eggs) is 75 kW per ha each. Dairy. Dairy production systems can potentially A typical 1–2 ha operation would generate a demand create significant electricity demand in rural areas where of about 150 kW (300 kW if the two operations are there are commercial milk producers or cooperatives. co-located). The main source of rural electricity demand from dairy Tea. For the tea value chain, electricity demand is production is cold storage, and machines for electric- from irrigation and processing activities. Irrigation is ity powered milking are also becoming more prevalent mainly rainfed since most tea is grown in areas with abun- (figure 3.2i). Another potential source is machinery for dant rainfall. Even so, there is a considerable potential Figure 3.2i: Electricity input in the dairy value chain Milking Cold storage Pasteurization • Power-driven milking • Individual solar chillers might • Requires heating machines usually used for be an option for smaller- • Centrifuging and dehydration medium- and large-scale scale dairy farmers may be required for other systems derivative products (e.g., cream and dry milk powder) Figure 3.2j: Electricity input in the poultry value chain Egg or meat Incubation Temperature control processing • Temperature-controlled • In Sub-Saharan Africa, • Electricity is generally used to egg incubators temperature controlled, power refrigeration, conveyor poultry layer houses usually belts, lighting, air conditioning, require cooling rather than pumps, compressed air, and heating (apart from egg other mechanical drives incubation) 34 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2k: Electricity input in the tea value chain Irrigation Shredding/rolling Fermenting/drying • Most tea estates are rain-fed, • Weathering required prior to • Electricity is generally used to but some use supplementary shredding power refrigeration, conveyor irrigation • Cutting, tearing, and curling belts, lighting, air conditioning, (CTC) uses electricity pumps, compressed air, and other mechanical drives Figure 3.2l: Electricity input in the floriculture (roses) value chain Irrigation Cooling • Accounts for about half of • Uses about 35 percent of the the energy consumed electricity consumed • The remainder is used for general facility needs, lighting, and other purposes value gain from irrigation (i.e., increased yields of up Floriculture (roses). In Sub-Saharan Africa, roses to 8 times from sprinkler irrigation and up to 16 times are cultivated mainly in large-scale greenhouses, and from drip irrigation) (figure 3.2k). Thus, the value gain most power demand is from irrigation and cold storage from electricity use results from both increased yields in (figure 3.2l). Electricity is usually sourced through diesel response to irrigation and the value addition from process- generation sets. All farms have on-site cold storage, and ing (including reduced transport and spoilage costs). In growing is done in temperature controlled environments. Sub-Saharan Africa, there is considerable potential for tea For large-scale production, power demand is estimated at producers to gain from increasing yields and moving fur- 2.37 kW per ha, with irrigation accounting for nearly half ther up the processing value chain. In Kenya, 88 percent of energy consumption; thus, a 300 ha cultivated area of tea production is exported raw in bulk; but in Rwanda can be expected to have about 700 kW of power demand. and Uganda, processing is rising. Electricity demand from Cotton (lint). For cotton (lint) production, electricity tea cultivation and processing is estimated at 1.91 kW powered irrigation is not prevalent. Rather, electric power per ha for large-scale plantations and 0.51 kW per ha for is used mainly for seed crushing and ginning (figure 3.2m). small-scale, rainfed facilities. For a 300 ha cultivated area, Due to perishability, cotton ginning must be done soon power demand is in a range of 150–575 kW, depending after harvest. Gins are usually located near reliable power on the scale of cultivation and associated farming and sources in rural and peri-urban towns. Moving ginning post-harvest practices. closer to farms would save on transport costs and possible Figure 3.2m: Electricity input in the cotton (lint) value chain Textile Irrigation Ginning Oil extraction manufacturing • Mostly does not use • The process of separating • Oil presses, expellers used to • Ginned cotton, spun into electricity powered irrigation cotton fibers from the seeds extract oil from seeds yarn, enters various textile value chains Power Needs in Selected Value Chains35 Table 3.4: Power Demand for Standard 300 ha Cultivated Area Per Unit Total Electricity Capacity Electricity Capacity Required for 300 ha   (kW/ha) for Irrigation and Processing Cultivated Area (kW) Agricultural Commodity Small-scale Large-scale Small-scale Large-scale Maize 0.77 1.17 230 350 Rice 0.03 1.04 9 312 Wheat 0.52 1.10 156 330 Cassavaa 0.56 168 Oilseed (soybean) 0.64 1.04 192 312 Horticulture (pineapple)b 0.75 225 Sugarcane 1.00 1.81 300 543 Oil palmb 0.02 6 Teac 0.51 1.91 153 573 Cotton (lint)b 0.03 0.03 9 9 Floriculture (roses)b 2.37 711 Poultryb 75.00 22,500 Dairyb 0.61 183 Note: Choice of the 300 ha benchmark reflects the amount of cultivated area that may constitute a typical project site. For example, this would amount to 300 households, each having 1 ha of landholdings. While this benchmark is somewhat arbitrary (i.e., project sites are likely to have a variety of crops under cultivation), it can be used to construct back-of-the-envelope estimates on electricity demand from the value chains presented. a. Cassava is small-scale only. b. Horticulture (pineapple), oil palm, cotton (lint), floriculture (roses), poultry, and dairy do not use electricity for small-scale operations or are only large-scale operations. c. Small-scale tea cultivation uses rainfed irrigation. spoilage. Cottonseed crushing is done to produce cotton- is considerable. For small-scale production, potential elec- seed oil (used in some instances as a biofuel for vehicles) tricity demand ranges from 9 kW for rice or cotton (lint) and livestock feed. The power demand from cotton to 300 kW for sugarcane. For large-scale production, it cultivation and processing is estimated at 0.03 kW per ha ranges from 6 kW for oil palm to 711 kW for floriculture for both large- and small-scale farming production. This (roses); poultry is an outlier, at 22.5 MW. These estimates implies that a 300 ha cultivated area will have about 9 kW are useful for considering whether the economics of these in power demand. values chains make them viable anchor loads for rural For each of the 13 selected value chains, table 3.4 electrification. summarizes the estimated electricity demand for a Using the forecasted production for the 13 value 300 ha cultivated area and the per-hectare electricity chains presented in table 3.2, along with the constructed demand estimates from irrigation and processing. The unit unit electricity demand for each commodity, a bottom-up estimates show that per-hectare electricity demand is estimate of the total increase in demand for electricity largest for poultry by far, followed by floriculture, tea, and stemming from the selected value chains can be con- sugarcane. The potential per-hectare demand for poultry structed. The calculations show that electricity demand (meat) is considerably higher because the process is much could increase by 2 GW (from 3.9 GW in 2013 to 6 GW more intensive, using less land for a much larger yield. in 2030). This figure represents nearly half of the total The higher per-hectare demand estimates for large-scale potential increase in electricity demand from agriculture production mainly reflects the use of commercial-scale calculated for Sub-Saharan Africa in chapter 2 (4.2 GW). irrigation and the power input required to process large To the extent that the value chains selected represent yields. The range of values for the 300 ha cultivated area the best potential of the agriculture and agribusiness 36 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.3: Potential power demand in sectors in Sub-Saharan Africa, the estimated electric- 2030 from processing for small-scale ity demand provides a good indication of the possible agriculture, by selected value chains electricity-agriculture synergies (figure 3.3). The required underlying assumption is the percentage of irrigated Tea, 0.7% and processed production. Clearly, even by 2030, not Dairy, 5.6% Oilseed, 0.5% Pineapple, 0.3% Poultry, 0.3% all production is likely to be cultivated on irrigated land Wheat, 5.9% or processed using electricity driven machinery. With little detailed data available on irrigation and processing proportions by value chain, this study makes conserva- tive assumptions for each of the value chains considered: Cassava, 19.7% only 15 percent of the land is assumed to be irrigated and 15 percent of crops are assumed to be processed.20 Rice (paddy), 26.4% Maize, 37.2% Sugarcane, 3.3% Note: The underlying calculations assume concave production growth until 2030, based on historical average growth rates (2009–13), and 15 percent of the crop being irrigated and processed—no estimate available for floriculture endnotes 1. Of course, all of these factors are correlated. A value chain catering to export markets would likely add more value to the primary product through many production and processing steps and use of greater modern inputs. 2. FAOSTAT 2013 (http://faostat3.fao.org). 3. FAOSTAT 2014 (http://faostat3.fao.org). 4. International Institute of Tropical Agriculture (IITA) (http://www.iia.org/maize). 5. FAOSTAT 2013 (http://faostat3.fao.org). 6. In Zambia, an abundance of water and access to cheap grid electricity have played a significant role in the adoption of large-scale irrigated farming systems. 7. Tea and coffee are Rwanda’s most important exports (e.g., tea exports in 2013 totalled US$55 million); see FAOSTAT 2014 (http://faostat3.fao.org). 8. Production growth in Nigeria is driven by poultry-sector demand. 9. Estimates of ECA and Prorustica (2015). 10. For further analysis of commercial irrigated agriculture’s potential, see case studies 1 and 3 (chapter 4). 11. The load from processing rainfed tea is just 0.6 kW per ha. 12. Floriculture may not demand a large load in absolute terms as estates are seldom larger than 50 ha (requiring less than 120 kW for production). Exceptions may be additional power requirements for staff housing (see case study 5, chapter 4). Power Needs in Selected Value Chains37 13. Data for horticulture (pineapple) is missing and therefore not included. 14. A complementary analysis is the ongoing work in Latin America and the Caribbean on energizing agriculture; the study estimates energy demand for processing for selected value chains, and proposes energy efficiency options and associated costs (World Bank 2016b). 15. In India, this model has had some success through husk power systems. 16. Data on the electricity requirements of post-harvest activities (juicing, cooling, and canning) were unavailable. 17. An example is Del Monte’s biogas plant in Kenya, which is based on pineapple residue. 18. The produced biomass consists of empty fruit bunches (EFBs), palm kernel shells, fibers, and possibly solids from decanters; in most cases, this biomass is used to boil water and generate (super-heated) steam. 19. According to the FAO, economic losses for the dairy sector in Kenya, Tanzania, and Uganda total up to US$56 million per year. 20. The assumption for the irrigated proportion of a crop is in the ballpark of the CAADP target of doubling the land under irrigation by 2030; considering that about 6 percent of cultivated area is currently irrigated (FAO 2005), irrigated production has dispropor- tionately greater yield, and the selected value chains are the best performing crops in the region. Lessons from Ongoing Power-Agriculture Integration Projects Chapter 4 T his chapter presents a suite of case studies on Another important consideration is the trade-off power-agriculture integration in several coun- between affordability and cost recovery in setting elec- tries of Sub-Saharan Africa. All three countries tricity tariffs. While different regulatory environments covered—Tanzania, Zambia, and Kenya—show a afford different levels of flexibility in tariff setting for high potential for on-farm and agro-processing activities individual schemes, it is instructive to assess the tariff level to contribute toward regional and, in some cases, national that can optimally balance the cost recovery objective and power-sector development. These cases offer indicative affordability, in particular with respect to the anchor cus- analysis of specific project areas in terms of their potential tomer. The case studies aim to answer two key questions: and viability for furthering rural electrification.1 The objec- (i) Up to what price is power affordable for agriculture tive is to provide a point of reference for the potential of activities? and (ii) Below what price is power uneconomic power-agriculture integration and to highlight some of the to supply? important issues to consider in trying to promote such an Each case study is organized into four sec- integration. Each case study project asks (i) whether the tions: (i) power demand (agriculture and residential/­ investment in expanding rural electrification is economi- commercial), (ii) power supply options and commercial cally viable and (ii) under what conditions private-sector arrangements, (iii) financial viability, and (iv) economic participation in electricity supply is feasible. viability. Annex D presents the maps corresponding to the A standard cost-benefit analysis reveals that most case study project areas. of the projects analyzed are economically viable and are thus worth undertaking by governments.2 The social and economic benefits generated as a result of rural electrifi- Case Study 1. Tanzania: Sumbawanga cation often outweigh the costs incurred and may justify Agriculture Cluster well-designed subsidies to improve the financial viability of the project. Indeed, if the economic value of the grid The Sumbawanga agriculture cluster is located in the extension exceeds the economic costs (due to positive Southern Agricultural Growth Corridor of Tanzania externalities), an otherwise financially unviable project (SAGCOT), on the country’s western border (map D.1). can be undertaken with subsidy financing to cover the SAGCOT focuses on the coordinated development of shortfall. small and commercial agriculture, physical and market In many cases, private-sector participation is desir- infrastructure along the transport corridor that runs from able for developing and operating electricity supply as it Dar es Salaam through to (and immediately across) the can improve supply efficiency and reduce the financial Zambian border at Tunduma.3 Small-scale farmers are and capacity burden on public-sector providers. Thus, integrated into commercial value chains as outgrowers when analyzing various supply options, it is instructive and benefit from the agglomeration economies that lower to consider their commercial viability in order to under- costs of access to shared infrastructure and inputs (e.g., stand whether private-sector participation is viable and electricity, roads, markets, labor, and extension services) the amount of subsidy that may be required to attract (table 4.1). private-sector operators and developers. 38 Lessons from Ongoing Power-Agriculture Integration Projects 39 Table 4.1: Sumbawanga Agriculture Cluster at a Glance Project overview Expansion of electricity supply to support the development of an agriculture cluster and surrounding households through main power grid extension. Commodities Maize, sunflower, finger millet, paddy, and sorghum. Description Powered irrigation and residential demand are the main drivers of increased power demand. Grid extension is a viable option given that the grid extension passes through the Sumbawanga cluster to connect other load centers beyond it. Forecasted size of the load and limited local generation potential make grid extension the most feasible option. Powered irrigation is an important concentrated source of electricity demand. In its absence, greater dispersion of electricity demand over a wider area may reduce viability; thus, a greater cultivated area will be required to have large enough demand from processing. Financial As a stand-alone project, it is marginally financially unviable. A relatively small increase in electricity viability demand from agriculture or residential consumers would increase the financial viability of the grid extension. Economic Economic benefits would be significant (US$134 million) and justify the project. The benefits come viability mainly from household cost savings, small-scale irrigation, and increased commercial sale of produce. Still at a concept stage at the time of this writing, a few farmers use petrol and diesel-powered pumps which the Sumbawanga agriculture cluster aims to integrate are inefficient in water use and costly to run. To date, small-scale and commercial farming, along with process- there has been little penetration by solar pumps. ing and storage facilities, transport, and logistics hubs, With demographic and agricultural growth, forecasted and improved ‘last mile’ infrastructure to farms and local demand for electricity is expected to far exceed the cur- communities over an area of 27,000 km². The cluster rently available capacity. To meet this future demand, the has strong natural characteristics for agricultural devel- Government of Tanzania, through the Tanzania Electric opment, including proximity to Lake Tanganyika, good Supply Company Limited (TANESCO), intends to extend quality soils, and high rainfall. However, owing mainly to a 220 kV line from Tunduma (on the Zambian border) to its geographical isolation, the area lacks both physical Sumbawanga (and beyond through Mpanda to Kigoma on infrastructure (e.g., good roads, rail access, and power) Lake Tanganyika). and market infrastructure (e.g., integrated production and processing, traders, finance, and input suppliers). Power Demand Access to reliable and affordable electricity is critical to realize the cluster’s potential. Currently, the The annual power demand in the Sumbawanga region has Sumbawanga area benefits from a power capacity of the potential to increase to an estimated 60–70 MW by 10.6 MW serving a population of just over 1 million people 2030. Irrigation and residential demand are the expected (table 4.2).4 Where it is available, farmers and agribusi- main drivers of load growth, with commercial and pro- nesses purchase power from TANESCO (including from cessing loads playing a relatively less significant role its mini-grids). There is very little powered irrigation, but (figure 4.1). Agricultural demand. The majority of growth in electricity demand from agriculture will come from devel- Table 4.2: Sumbawanga Geographic oping the region’s irrigation potential, roughly estimated and Demographic Features at 50,000 ha.5 Assuming 35,000 ha of this amount is Feature Value dedicated to small-scale agriculture implies a total energy Estimated population (2012) 1,000,000 demand of roughly 25.5 MW by 2030 from both bulk Population growth rate (%) 4.0 water pumping and in-field irrigation. Newly irrigated land, higher quality inputs, crops switching, and knowledge Electricity connection rate (% of households) 7.0 sharing are expected to increase yields from 461,000 MT Sources: SAGCOT; ECA and Prorustica (2015). to 1.09 million MT by 2030 (table 4.3). 40 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.1: Estimated peak load and energy demand, by sector Power Capacity Demand (MW) Energy Demand (MWh/year) Source of Demand 2012 2030 2012 2030 Irrigation 0.0 25.5 0 48,450 Processing 0.4 4.4 2,000 22,000 Residential 3.9 26.7 26,232 174,327 Commercial 0.2 2.6 85 1,056 Total 4.5 59.2 28,317 245,833 a. Peak load b. Energy demand 80 300 60 200 GWh MW 40 100 20 0 0 2015 2017 2019 2021 2023 2025 2027 2029 2015 2017 2019 2021 2023 2025 2027 2029 Irrigation—small-scale Irrigation—large-scale Irrigation—small-scale Irrigation—large-scale Processing Residential Processing Residential Non-residential    Non-residential Source: ECA and Prorustica (2015). Table 4.3: Total Power Demand from Agriculture by 2030 Agriculture Activity Power Capacity Demand (MW) Hours of Operation/Year Energy Demand (MWh/year) Irrigation 25.5a 1,900 48,450 Processing 4.4b 5,000 22,000 Total 29.9 6,900 70,450 Sources: SAGCOT; JICA; Rukwa District Council; WREM International; ECA and Prorustica (2015). a. Based on a potential area of 50,000 ha under irrigation and an estimated power demand for irrigation of 0.65kW/ha (0.3kW/ha for small-scale farms and 1kW/ha for commercial farms). b. Based on a processed production of 472,500 MT and an estimated 11 mills required (400 kW). Lessons from Ongoing Power-Agriculture Integration Projects 41 Figure 4.2: Estimated volume of crops that may utilize electricity for processing 500 Volume of production (’000 MT) 400 300 200 100 0 2015 2017 2019 2021 2023 2025 2027 2029 Rainfed volume processed Irrigated volume processed Source: ECA and Prorustica (2015). The power demand for post-harvest processing Together, this implies an estimated power demand of will depend on the crops produced and the volume about 4.4 MW by 2030 (figure 4.2). of production. Electricity demand is expected for Residential/commercial demand. Based on the post-harvest processing of crops (e.g., milling and oil regional population growth rate of 4 percent, Rukwa’s extrusion) such as maize, paddy rice, beans, millet, population is expected to reach 2 million by 2030, repre- sorghum and sunflower.6 Greater electricity supply senting 400,000 households.7 Considering the house- and better access to markets for farmers would boost holds’ annual consumption and anticipating that their the electrification rate of agro-processing activities. demand and consumption will likely evolve over time with An estimated 40 percent of the current crop yield the adoption of additional electric appliances, residen- and an assumed 75 percent of the increased yield due tial consumers will be the main driver of energy demand to irrigation expansion will be processed by 2030. (table 4.4). Table 4.4: Residential and Commercial Data to Calculate Power Demand Residential 2012 2030 Population 1,000,000 2,025,817 Population growth 0.04 People per household 5 5 No. of households 200,000 405,163 Household connection rate 7% 20% Households connected 14,000 81,033 Per household peak consumption (kW) 0.28 0.33 Per household energy consumption (kWh/month/HH)a 156 179 Total peak (MW) 3.9 26.7 Total energy consumption (MWh) 26,232 174,327 Commercial  No. of customers 6 75 Consumption peak (kW) 34 34 Consumption energy (kWh) 14,085 14,085 Total peak (MW) 0.2 2.6 Total energy consumption (MWh) 85 1,056 Sources: SAGCOT; ECA and Prorustica (2015). a. Assumes a daily demand of 5.13 kWh per household. 42 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.3: Comparative cost of power supply options in Sumbawanga 100 90 80 80 80 Cost of power (US¢) 60 Cost to serve 40 Tari charged 25 25 20 11 10 11 0 Diesel mini-grid Hybrid mini-grid Solar mini-grid Main grid connection Source: ECA and Prorustica (2015). Commercial demand from current loads averages The least-cost method is thus estimated to be an 85 kWh per month across six TANESCO customers, with extension of the national grid. This would allow for more a peak load of 0.21 MW. Should per-customer demand efficient generation capacity sizing for demand on the levels remain as observed when electricity was supplied in system at more competitive costs. In deciding how much other areas of comparable size (e.g., Morogoro, Iringa, and transmission capacity to invest in, it is more feasible Mwanza), the number of customers would increase to 75; to install adequate capacity to meet future projected thus, annual power consumption would rise to 1,056 MWh demand rather than upgrade capacity in response to by 2030, and power-capacity demand would reach increase in demand. The subsection below describes a 2.6 MW (table 4.4). scenario where sufficient capacity is directly incorporated into a project’s initial design. Power Supply Options and Commercial Arrangements Financial Viability: Extension of Main Grid from Mbeya to Sumbawanga The analysis considered various options for additional and Rukwa power capacity to meet projected demand. Localized generation potential from diesel, solar, hybrid, hydro, and The financial viability of grid extension is estimated from bagasse/biomass was considered, along with the option the perspective of TANESCO. To supply activities in to extend the national grid. Preliminary analysis showed Sumbawanga, both grid extension and generation capacity insufficient potential for hydro- and biomass-based gen- expansion are required. However, generation capacity eration, so these options were ruled out. expansion is on a national least-cost basis; the focus here The option to expand mini-grid capacity, based on is on the viability of the transmission and distribution diesel, solar or a hybrid of the two, was also found unviable network development (table 4.5). for the region. The cost of a diesel-based mini-grid is The costs associated with provision of grid electricity estimated at US¢90 per kWh, which is much higher than to Sumbawanga consist of the cost of electricity gener- the cost of extending the national grid.8 Even if hybrid ation and transmission and distribution costs (expansion solutions enable the lowering of generation costs (i.e., at and operation). The corresponding revenues would US¢80 per kWh), they are still much more costly than be those of electricity sales at the national tariff level grid extension. Finally, solar mini-grids are not adapted to (table 4.6). the load profiles of agro-processing and irrigation activi- ties, which would imply expensive investments in storage and backups (figure 4.3). Lessons from Ongoing Power-Agriculture Integration Projects 43 Table 4.5: Estimated Capital and Operating Costs for Transmission and Distribution Expansion Operating Expense Grid Extension Cost Total Cost Assumption AC Losses Assumptions Distance (km) (thousand US$/km) (million US$) (%) (%) 11 kV 200 15 3.3 3 4.6 33 kV 200 35 7.7 3 4.6 220 kV 350 138 53.1 3 4.6 Subtotal (million $) 64.1 Present value (million $) 61.2 18.3 3.8 Total (million $) 83.4 Sources: Ministry of Energy and Minerals (MEM), Power System Master Plan; ECA and Prorustica (2015). Table 4.6: Estimated Power Consumption At the assumed 10 percent average cost of capital, the and Transmission and Distribution project is marginally financially unviable as a stand-alone Tariff Requirement project (table 4.7). However, TANESCO’s ability to attract financing on more favorable terms or greater reve- Variable Value nues from electricity demand, would improve the project’s Cost (million US$) 83.36 financial viability. On the other hand, a larger proportion Estimated consumption (MWh) 1.2 million of consumers paying lower lifeline tariffs, lower electricity Transmission and distribution, tariff requirement demand, and/or higher costs would further reduce the (US¢/kWh) 6.9 financial viability of the investment in grid extension. Source: ECA and Prorustica (2015). Economic Viability Table 4.7: Financial Present Value Analysis of the project’s economic viability adds social of Grid Extension net benefits to the financial net benefits accruing to the developer (TANESCO). Thus, the economic analysis Value includes benefits accruing to newly connected house- Variable (million US$) holds, benefits from improvement in agricultural yields, Revenue, based on TANESCO tariff 167.34 market access, and jobs creation (table 4.8). Transmission costs (83.36) The economic analysis shows that the economic bene- Generation costs (89.83) fits significantly outweigh the associated costs. In fact, the Effective project shortfall (5.85) benefits accruing to the households alone are sufficient to Internal rate of return (%) 12 justify the investment in grid extension. Sources: ECA and Prorustica (2015); World Bank. Note: Assumes a consumption of 1.2 million MWh over 20 years. The generation cost is based on cost for the upcoming Kiwira coal plant, at US¢ 7.5 per kWh (TANESCO 2012 Power System Master Plan Update, May 2013). The coal plant near Mbeya is expected to be completed by 2020. The average retail tariff is about US¢ 14 per kWh. 44 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.8: Economic Costs and Benefits of Sumbawanga Grid Extension Present Value of Cost/Benefit Economic Cost/Benefit Beneficiaries (number) (million US$) Net financial costs (5.85) Household cost savingsa 52,671 households by 2030 42.00 Small-scale irrigation 35,000 farmers (1 ha each) 34.50 Margin uplift from market access All small-scale farmers 26.80 Import substitution Tanzania broadly 8.52 No. of jobs created by electrifying the agriculture field 3,750 24.00 No. of jobs created by electrifying the town 550 4.20 Economic net present value 134.14 Source: ECA and Prorustica (2015). a. These are the additional households that are assumed to be connected from the grid extension project—over and above the baseline (w/o project). Additional household benefits may include better health outcomes from reduced fuel use, better educational outcomes for school going children, women’s time savings, and better nutrition. Case Study 2. Tanzania: Mwenga surrounding rural communities. The project was the first Mini-Hydro Mini-Grid green-field development under the Small Power Purchase Agreement (SPPA) scheme. The SPPA was signed with The 4 MW Mwenga mini-hydro mini-grid project is TANESCO in 2009, and the plant was commissioned in located in Tanzania’s Southern Highlands, close to the 2012 (table 4.9). RVE owns and operates the distribution Mufindi Tea and Coffee Company (MTC) (map D.2). network connecting roughly 20 villages and relies on a The project is operated by the Rift Valley Energy (RVE), mobile phone based pre-paid vending system for electric- a 100 percent subsidiary of the Rift Valley Corporation, ity billing. which also owns MTC. The project came about as a result Notwithstanding its long and complex development of MTC’s need to supplement electricity from the main process, Mwenga is considered Tanzania’s most success- grid to ensure access to a reliable source of uninter- ful private mini-grid development project. For the tea rupted power. Cofinanced by the European Union (EU) factory, the mini-grid is an opportunity to switch from and the Rural Energy Agency (REA), the project was grid-based power to a more reliable supply produced by developed as an independent power producer (IPP) to renewables. Although the project was initially designed to supply power to the main grid, local tea industry, and supply only the MTC, having power lines extending from Table 4.9: Mwenga Mini-Hydro Mini-Grid at a Glance Project overview A 4 MW hydro mini-grid connected to the main grid. Main local anchor load is the Mufindi Tea Estates and Coffee Limited; 2,600 households connected in the surrounding communities. Commodities Coffee, tea. Lessons learned The tea estate is the main anchor load of the grid connected mini-grid. Given the seasonality in tea processing operations, the peak load demand more than doubles during the summer season. This impacts the choice of power supply arrangement. Excess supply was sold to the grid, which helps mitigate the impact of seasonality. While residential consumers are numerous, their power demand is not high enough, at least initially, to mitigate the impact of a seasonal anchor load. Financial The project’s financial viability depends critically on the ability to sell excess power to the main grid. viability Despite financial viability, capital subsidies were provided for the project to keep local electricity tariffs low. Economic Economic benefits are positive (US$9 million) and come from households’ energy cost savings, reduced viability reliance on diesel backup for the tea estate, and job creation from new electrified businesses. Lessons from Ongoing Power-Agriculture Integration Projects 45 the hydro plant through nearby villages facilitated the All excess power from the mini-grid (about 80 per- connection of 2,600 households, as well as other com- cent of generated power) is sold to TANESCO, in accor- munity facilities. Beyond enhancing electricity access, the dance with its SPPA and feed-in tariff (FiT) arrangement; project has replaced the use of diesel and kerosene with these have been instrumental in guaranteeing offtake and sustainable hydropower among neighboring communities. have helped justify development of a scheme of its size, thus benefiting from economies of scale. Selling power Power Demand only to local consumers would not have justified the proj- ect in terms of its scale or commercial viability. Demand for power from the Mwenga mini-grid comes from the main grid (TANESCO), commercial and com- Power Supply Options, Commercial munity users, agriculture, and residential customers. As Arrangements, and Financial Analysis local demand is expected to grow, the sales to the grid are expected to decline. Local demand growth is expected to Proximity to the Mwenga River enabled the tea plant be led by the informal and semi-formal agriculture and to access a renewable source of power, with sufficient forestry sectors, highlighting the significant economic volume and head to develop a 4 MW run-of-the-river, development potential of the project. mini-hydro plant. The project is owned and operated by Agricultural demand. In terms of power for agri- MTC’s sister company and both are held by the RVC culture, MTC mainly requires electricity for processing. parent company. Specifically, electricity is used to power large motors, The project was developed as a private-public part- fans, and vibrating sieves (used to cut to length the leaves, nership and partly funded through public funds, including and wither, dry, sort, and grade the tea). The tea factory’s elements of grant and concessional loans from the EU peak load averages about 700 kW (with a summer peak and REA.10 The use of concessional funds was necessary of 900 kW and a winter peak of 400 kW), with an annual to reduce the tariff burden on local electricity custom- power consumption of 2,880 MWh.9 ers. While the electricity regulator allowed RVE to set Community and commercial demand. In addition cost-reflective tariffs, as per Tanzania’s SPP framework, to supplying agro-processing activities, the Mwenga fairness and affordability concerns led to the tariff being mini-grid project specifically targets facilities such as set in line with the tariff on the main grid. The regulator schools and clinics, as well as small commercial businesses, has allowed recent adjustments in the tariff, which is thereby improving electricity access for productive uses. currently TZS 100 per kWh up to 75 per kWh (equivalent According to RVE, annual power consumption for com- to US¢6.25 per kWh under the pre-devaluation exchange munity and commercial users is estimated at 2,988 MWh. rate). However, since 80 percent of the generated power Residential demand. Residential customers comprise is sold to TANESCO under the SPPA and FiT, the viabil- the majority of the customer base; however, most resi- ity of Mwenga’s hydro plant is not relying on the profit- dential customers have very low demand and pay lifeline ability of selling electricity to local communities. tariffs. Annual demand from the 2,600 customers is estimated at just 936 MWh (table 4.10). Table 4.10: Estimated Power Demand from Mwenga Mini-Hydro Plant Forecast Total Monthly Usage Connections Connections Approved Tariff (all customers) Customer Group (current) (2030) (TZS/kWh) (MWh) Households 2,600 5,600 100 78 Commercial 374 557 205 114 Public/community services 468 668 205 135 Tea estate 1 1 Uncertain 240 TANESCO 1 1 189 1,922 Total monthly usage (MWh) 2,489 Source: RVE. 46 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.11: Economic Costs and Benefits of Mwenga Mini-Hydro Plant Present Value of Cost/Benefit Economic Cost/Benefit Benefits (million US$) Net financial costs 0.0 Development subsidies received by project (7.1) Household cost savings (no. of households)a 5,600 6.4 Tea company savings from reduced diesel backup requirement (hours/year)b 288 1.4 Jobs created by electrifying villages (no.)c 1,120 8.6 Economic NPV 9.3 Source: ECA and Prorustica (2015). a. Households are assumed to save $14 per month from access to electricity; b. diesel backup requirement is assumed to be 10% of the total power consumption; c. it is assumed that 65 percent of the businesses will each create 1.5 jobs. Each job created is valued at the average expected salary: $1500/year. Financial Analysis Case Study 3. Zambia: Mkushi Farming Block The financial analysis considers the Mwenga mini-hydro project from the perspective of the revenues and costs The Mkushi farming block project is located in Zambia’s incurred by the owner, RVE. However, information on Central Province (300 km northeast of Lusaka) and revenue, operating cost, and capital expenditures was stretches over 176,000 ha of land (map D.3). The Mkushi confidential and thus not available. Despite this limitation, farming block is one of Sub-Saharan Africa’s largest multi- discussions with the operator allow us to make certain farmer commercial farming areas outside South Africa. salient points: Mkushi produces the largest share of Zambia’s wheat ºº Tanzania’s SPP framework allows RVE to charge a (40 percent) and soybean (21 percent), and is its sixth tariff that should ensure full cost-recovery, including largest maize producer. Other export crops grown in the a return on capital, even if all capital is at commercial area include tobacco, soya, vegetables, and coffee (Chu rates, and adjusted for any subsidies received. 2013). Mkushi experiences distinct dry winter seasons ºº In practice, social concerns implied that the tariff (May to October) and wet summer seasons (November to was set equal to the main grid. Thus, in order to April). Irrigation is thus critical for growing winter crops, accommodate this lower tariff, subsidies for capital especially wheat (table 4.12). expenditure were sought to reduce the effective cost Electrification of the Mkushi farming block occurred recovery, such that it aligned with the tariff. over time, given the evolving demand and difficulty of raising the necessary capital. Mkushi was first connected Given that RVE, a private-sector company, continues to the grid in 1996 through a 33 kV line. This effort was to operate the facility, one can assume that the project at financed by the government and a group of 20 farmers least breaks even financially. who contributed US$10,000 per km (50 percent of the total cost), which was the policy of the Zambia Electricity Economic Analysis Supply Corporation (ZESCO) at the time. However, unreliable power supply due to inadequate feeder capacity Economic net present value (NPV) is estimated at about meant that farmers had to continue to use backup diesel US$9 million, based on a 10 percent discount rate over generators for irrigation. A subsequent grid expansion the assumed project life till 2030 (table 4.11). Benefits was undertaken in 2000, followed by a third in 2005 accrue from household energy cost saving, reduced reli- to connect all farmers and many households in the area. ance on diesel backup for the tea estate, and job creation Expansion of the national grid into the area has enabled from newly electrified businesses. the area under irrigation to expand to about 18,000 ha Lessons from Ongoing Power-Agriculture Integration Projects 47 Table 4.12: Mkushi Farming Block at a Glance Project Overview Extending a transmission line into a farming area with significant agricultural potential. Commodities Wheat, soybean, tobacco, soya, vegetables, coffee. Description Irrigation counts for more than 90 percent of total power demand. Given their interest in the project, farmers accepted to contribute to capital costs. The grid extension enables a significant increase in household connection rates (from 2 percent in 1995 to 7 percent in 2014). However, more than 30,000 households remain unconnected to the main grid. Financial From a purely financial perspective and as a stand-alone project, grid extension to Mkushi was not Viability profitable for the utility. However, in order to expand access to new farmers coming into the area, sharing of capital costs was an appropriate and successful approach to project financing. Economic Thanks to household energy cost savings, increased yields from irrigation on small-scale farms, and job Viability creation, the project’s economic NPV was positive (US$46 million). Figure 4.4: Total peak load in Mkushi, 1995–2014 25.0 Irrigation 20.0 Milling Peak load (MW) Residential 15.0 Commercial 10.0 5.0 0.0 1995 2000 2005 2010 2014 Source: ECA and Prorustica (2015). and led to the subsequent development of milling Agricultural demand. Among agriculture activities, activities. irrigation has been the main driver of power demand, Out of 150 commercial farms hosted on the farm- with milling accounting for only a small share of total ing block in 2014, 80 farms have developed irrigation agricultural power demand. Power demand for irrigation schemes to enable wheat production in winter and to grew from 0.5 MW to 18 MW between 1995 and 2014, supplement summer crops. The availability of water and with a yearly consumption of 34,200 MWh in 2014 the connection to the national grid, supported by ZESCO (figure 4.5).11 In addition to development of irrigation and the Zambia National Farmers Union, were central to schemes, two mills were installed in the area following development of these irrigation schemes and processing arrival of the grid. Power demand for milling was esti- facilities. mated at 800 kW,12 for a consumption of 4,000 MWh (table 4.13). Power Demand Residential/commercial demand. Between 1995 and 2014, household connection rates grew from 2 percent to Between 1995 and 2014, overall peak load in Mkushi 7 percent, with the corresponding number of connected (from agriculture, residential, and commercial consump- households increasing from 362 to 2,516 (table 4.14). tion) increased from 0.6 MW to 20.1 MW. Over that Over the same period, power demand from residential period, irrigation accounted for more than 89 percent of and commercial customers increased from 0.13 MW to total power demand (figure 4.4). 1.32 MW, with households representing 67 percent. 48 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.5: Power demand from irrigation and milling in Mkushi, 1995–2014 20.0 Power demand 16.0 from irrigation Power demand (MW) Power demand 12.0 from milling 8.0 4.0 0.0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Source: ECA and Prorustica (2015). Table 4.13: Power Requirements for Irrigation and Milling in the Mkushi farm block Agricultural Activity Requirement 1995 2000 2005 2014 Irrigation Irrigated land area (ha) 500 7,000 10,000 18,000 Power demand (MW) 0.5 7 10 18 Power consumption (MWh) 950 13,300 19,000 34,200 Milling Power demand (MW) 0 0.4 0.8 0.8 Power consumption (MWh) 0 2,000 4,000 4,000 Sources: Ministry of Agriculture; ECA and Prorustica (2015). Table 4.14: Electrification Rates and Power Load of Households in Mkushi Consumer Type 1995 2000 2005 2014 Residential Households (no.) 18,092 21,488 25,521 34,782 Household connection rate (%) 2 3 4 7 Households connected (no.) 362 603 1,004 2,516 Power demand per household (kW) 0.24 0.26 0.29 0.35 Peak demand (MW) 0.09 0.16 0.29 0.88 Total consumption (MWh)a 222 408 751 2,247 Commercial Total demand (MW) 0.04 0.08 0.15 0.44 Total consumption (MWh)b 190 349 642 1,921 Sources: Ministry of Agriculture; Zambia Census 2010; ECA and Prorustica (2015). a. Assuming household energy consumption of 75 kWh/month. b. Assuming that commercial consumers operate 14 hour per day 6 days a week. Lessons from Ongoing Power-Agriculture Integration Projects 49 Total power demand in 2014 was 20.1 MW, with cor- specifying their peak demand load. They were required to responding annual energy demand of 42,368 MWh. Of cofinance up to 50 percent of the cost of the line exten- this amount, 18 MW came from irrigation, 0.8 MW from sion and pay for the transformers. processing, 0.88 MW from households, and 0.44 MW from commercial customers. Financial Analysis Power Supply Options and Commercial Given the cofinancing arrangement, the financial analysis Arrangements of extending the grid to the Mkushi farming block was analyzed from the perspective of both ZESCO and a Zambia has one of the lowest electricity tariffs in Sub- representative farmer newly settled in the area. From the Saharan Africa owing to fully depreciated hydropower utility’s standpoint, even after capital costs were partially dominating the generation mix. This implies considerable paid for by customers, the revenue generated from the benefits from reliable electricity supply to farmers who grid extension remained below the costs incurred. The previously relied on backup diesel generation. This, along financial NPV was estimated at US$8.9 million, mainly with the relative proximity of the main grid, ruled out a because of the very low electricity tariffs (table 4.15). mini-grid option. The farmer was required to invest in half of the line As described above, to extend the grid to Mkushi, extension for 20 km (US$10,000 per km), a transformer farmers were initially required to apply to ZESCO, ($50,000), and irrigation capital ($2,500 per ha). Table 4.15: Financial Analysis of Mkushi Farming Block from the Perspective of the Utility and a Representative Farmer Thousands of US$ Factor 1995 2000 2005 2014 Utility   Tariff revenue 14 161 244 424   Capital costs 1,300 10,000 5,045 0   Operating costs 39 339 342 342   Net benefits −1,325 −10,178 −5,142 82   Financial NPVa −8,89 Representative Farmer (500 ha of irrigated land)b  Wheat   Extra profit 60  Maize    Extra production because of irrigation (MT) 1,250   Extra profit 199    Total extra revenue from irrigation 259   Capital costs 1,500c    Electricity consumption from irrigation (MWh) 950    Cost of electricity 33   Net benefits −1,275 226 226 226   Financial NPV 523  IRR (%) 17 Note: The financial NPV is calculated over a 20-year project life starting from the initial investment (1995–2014). a. The estimated negative NPV is over 20 years. Given the magnitude of the stream of revenues relative to the costs, considering 30-year project life will not make the project financially viable from the utility’s perspective. b. Irrigated production of 500 ha of wheat in winter and 500 ha of maize in summer. c. For a 20km connection expansion. 50 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa However, after deducting the cost of electricity and yields and job creation (table 4.16). The economic NPV capital costs from the extra profit generated by irrigation, is estimated at about US$46 million, which justifies the the financial NPV for a representative farmer was positive 130-km grid extension (table 4.17). ($522,653), showing that the representative farmer The project faced various implementation barriers. benefited from increased yields, owing to supplementary Since it was not financially profitable for the utility, the summer irrigation, as well as irrigated winter cropping shortfall had to be covered by subsidies. Other issues that (table 4.15). had to be overcome included lack of access to capital for project financing, lack of coordination between farmers, Economic Analysis and insufficient grid capacity to provide reliable power supply. Moreover, ZESCO and farmers competed over From an economy-wide perspective, between 1995 and water availability and use; the utility wanted water for its 2014, the largest benefits from access to grid electricity hydropower plant, while the farmers wanted it to irrigate accrued from savings on electricity expenditure, dis- their lands. placement of imports due to increased wheat and maize Table 4.16: Net Social Benefits of Grid Extension, Mkushi Factor 1995 2000 2005 2014 Savings on Energy Consumption Electrification rate (%) 2 3 4 7 Households electrified (no.) 362 603 1,004 2,516 Savings from grid electrification per household ($/month) 10 Total savings on energy consumption (million $) 0.04 0.07 0.12 0.30 Import Savings Wheat  Irrigation area (ha) 500 7,000 10,000 18,000   Production (MT) 3,000 42,000 60,000 108,000  Import substitution value of wheat (million $)a 0.21 2.94 4.20 7.56 Maize   Production without irrigation (MT) 2,750 38,500 55,000 99,000   Production with large-scale irrigation (MT) 4,000 56,000 80,000 144,000   Benefit of locally grown production over imports (million $)a 0.11 1.51 2.15 3.87 Revenue from Job Creation Job creation from area under irrigation 143 2,008 2,868 5,163 Extra income from irrigation (million $) 0.22 3.01 4.30 7.74 Present Value of Social Benefits over the Period 1995–2014 (million $) 65.47 Note: Assumes a 10 percent discount rate over a 20-year project life. a. Import substitution is valued at the difference between farm gate price in Zambia and import price. Table 4.17: Economic Costs and Benefits of Grid Extension, Mkushi Factor Value (million US$) Financial NPV of utility −8.90 Present value of capital cost contributions from farmers –10.83 Present value of social benefits 65.47 Economic NPV 45.74 Source: ECA and Prorustica (2015). Lessons from Ongoing Power-Agriculture Integration Projects 51 Case Study 4. Zambia: Mwomboshi Power Demand Irrigation Development and Currently, Mwomboshi’s access to grid electricity is low. Support Project The northern bank of the river has no electricity supply. Among small-scale farmers who are not connected to The Mwomboshi Irrigation Development and Support electric power, only a small portion uses petrol or diesel Project (IDSP) is situated along the banks of the pumps for irrigation purposes. Along the southern bank, Mwomboshi river in Zambia’s Central Province (World electricity from the national grid is used to power staff Bank 2011b) (map D.4). The IDSP aims to support irriga- housing, crop irrigation, processing, and other small-load tion development in order to increase agricultural yields activities (e.g., offices, water pumping, and tea drying). and incomes in the area. The project also includes support Planning for sufficient capacity to consider future for complementary infrastructure, including roads and loads from expanded farming activities includes upgrading electricity. Irrigation will be developed from water storage the current 11 kV line to a 33 kV line with a 30 km grid (via construction of small- and medium-sized dams) and extension to the north side of the river, which would pro- transport to individual farms (table 4.18). An extension vide all farmers with electricity. By 2031, it is estimated of the grid to and within the site will be funded under that the aggregated peak load from agriculture, house- the project and handed over to the utility to operate holds, and commercial activities will reach 6.4 MW, repre- (ZESCO). senting an 18.5 percent average annual increase from the Direct beneficiaries of the IDSP are the area’s 3,700 2016 peak load (figure 4.7a). Driven by irrigation, power inhabitants, along with small-scale and commercial consumption is forecasted to reach up to 15,000 MWh farmers. Commercial farms are located along the south- by 2031 (figure 4.7b). ern bank of the river, while small-scale farming is mainly Agriculture demand (irrigation). In addition to the on the north side. The connection to electricity is critical 439 ha currently underirrigated in Mwomboshi, the IDSP to enable irrigation development, which creates greater plans to add an extra 3,200 ha, distributed between opportunities to increase incomes. small-scale and commercial-scale farms. This will allow for Covering 100,000 ha, on-farm irrigation develop- the release of bulk water supplied from a water storage ment can be categorized into four tiers: (1) small par- dam through pump stations for irrigation schemes. The cels of less than 1 ha each, which utilize flood irrigation project will become the area’s major power load, requir- systems; (2) individual farms with parcels in a range of ing 2 MW to supply the southern bank of the dam and 1–5 ha, which utilize spraying irrigation schemes; (3) plots 3.1 MW for the north side. Once the first pumps are larger than 60 ha each, cultivated by a community or installed, the power consumption of pumping stations commercial farm that uses modern irrigation systems is forecasted to rise from 872 MWh in 2016 to about (e.g., center pivots); and (4) large parcels cultivated by 10,000 MWh by 2031 (table 4.19). large-scale commercial farmers that are supplied water Agriculture demand (milling). Development of the through a bulk-water storage facility (figure 4.6). region’s wheat milling capacity will evolve along with the increasing yields expected from irrigation. Total energy Table 4.18: Mwomboshi Irrigation Development and Support Project at a Glance Project overview Grid upgrade and extension to support irrigation development and household electrification. Commodities Tobacco, wheat, poultry, maize, sunflower, horticulture (tomatoes, onions, bananas). Description Electrification is mainly driven by irrigation of small-scale and commercial farming, leading to crop diversification and increased yields. The project also targets near universal residential access in the area by 2031. Proximity of the existing grid and power needs meant grid extension was the only option considered viable. Financial Positive financial NPV estimated at US$1.1 million. viability Economic Positive economic NPV estimated at US$2.0 million for the power line extension, mainly from greater viability irrigated tomato and maize production. 52 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.6: Mwomboshi IDSP plot sites developed for small-scale farmers Bulk water infrastructure—Pump and mains pipes, may include dam/reservoir Tier 3—Professionally managed farm block under pivot irrigation growing marketed food and cash-crops, purchasing produce from emergent farmers, and providing support services. Tier 2—Emergent farmers growing food and horticultural crops under sprinkler or other irrigation for sale to and supervised by the Water source, e.g. river professional farmer (5 ha each). Tier 1—Smallholder gardens on land currently farmed can grow vegetables etc. for local and subsistence consumption under some basic form of irrigation, e.g. furrow (1 ha each). Source: World Bank 2011b. Figure 4.7: Mwomboshi peak load and power consumption forecast a. Peak load 7 Peak load from irrigation 6 Peak load from milling 5 Peak load residential 4 Peak load commercial MW 3 2 1 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 b. Power consumption 15,000 Power consumption 12,000 from irrigation Power consumption 9,000 from milling MWh Power consumption 6,000 residential Power consumption 3,000 commercial 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 Source: ECA and Prorustica (2015). Lessons from Ongoing Power-Agriculture Integration Projects 53 Table 4.19: Irrigation Power consumption over this period should rise from 78 MWh Requirements in Mwomboshi, Zambia to 1,137 MWh. Nonresidential demand, led by com- mercial activities, is assumed at half of residential power Irrigation Requirement 2016 2031 consumption. Its peak consumption is thus expected to Power demand (MW) 0.5 5.1 increase from 0.015 MW in 2016 to 0.22 MW by 2031 Power consumption (MWh) 872 9,757 (figure 4.8). Source: ECA and Prorustica (2015). Power Supply Options and Commercial Arrangements Table 4.20: Milling Power Requirements in Mwomboshi, Zambia Since the southern part of the area is already connected to the national grid, no other supply option has been Milling Requirement 2016 2031 considered for improving power availability. To do so, the Power demand (MW) 0 0.6 Ministry of Agriculture and Cooperatives and ZESCO will Power consumption (MWh) 0 3,000 sign a Memorandum of Understanding (MOU) framing Source: ECA and Prorustica (2015). responsibilities for the construction and maintenance of Note: Assumes a mill operates 5,000 hours per year (16 hours a the new power line. ZESCO will own the assets and be day, 6 days per week)/mill size: 200 kW. responsible for line maintenance after construction and will recover its operating costs through tariff revenues. demand from milling is expected to be significantly lower than that from irrigation (table 4.20). The first mill is Financial Analysis expected to be installed when total production from com- From ZESCO’s perspective, the grid upgrade project in mercial farmers and the marketed portion (80 percent) Mwomboshi is financially viable, with a positive NPV of of small-scale production reaches 20,000 MT. The plan US$1.1 million. Given the current average electricity tariff is to add an additional mill for every 20,000 MT of extra of US¢3.5 per kWh and the estimated level of demand, production. the utility’s revenues are calculated as the additional reve- Residential/commercial demand. The IDSP plans to nues received by the utility due to the project (table 4.21). increase household connections from 15 percent (2014) to 97 percent (2031). Based on a per-household power demand estimate, peak load would increase by 2 percent Economic Analysis a year as the household load evolves over time. Total The IDSP is estimated to generate positive net benefits residential peak load should therefore increase from with a NPV of US$2.0 million. The economic benefits are 0.03 MW in 2016 to 0.45 MW by 2031, while electricity driven largely by the increase in yields of irrigated tomato, Figure 4.8: Residential and commercial demand, electrification rate 2016–2031 1,500 100.0% Residential and non-residential demand 1,200 80.0% Electrification rate (%) (MWh/year) 900 60.0% 600 40.0% 300 20.0% 0 0.0% 2020 2021 2022 2023 2025 2026 2027 2028 2029 2030 2031 2016 2017 2018 2019 2024 Peak load residential Non-domestic power consumption Electrification rate Source: ECA and Prorustica (2015). 54 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.21: Financial Analysis, Mwomboshi Factor Assumption Electricity tariff (US¢/kWh) 3.5 Transmission tariff (US¢/kWh) 1.0 Transmission OpEx (% of CapEx) 3 Cost of capital (%) 10 Line expansion (km) 30 Cost of grid expansion ($/km) 30,000 Total cost of transformers ($) 175,000 Net Present Value (NPV) Calculations 2016–2031 Present value of revenues (million $) 2.4 Capital costs (million $) 1.1 Present value of operating costs (million $) 0.3 Financial NPV (million $) 1.1 IRR (%) 20 Source: ECA and Prorustica (2015). wheat, and maize production (table 4.22). Irrigation will is connected to the main grid and purchases electricity allow farmers to increase production through better yields from the utility, it can generate power at a lower cost. To and crop diversification. The electrification savings to increase output by 0.4 MW, a planned upgrade of the farmers from using diesel pumps and switching to elec- generation plant aims to provide power to both industrial trified irrigation schemes will be minor since only a small activities and some 2,000 households. number of farmers are currently using these irrigation solutions. As a result, the total present value of social ben- Power Demand efits for the entire project is estimated at US$34 million. However, as these benefits are the result of the whole Currently, ODCL’s power demand is 3.2 MW, with irrigation project in Mwomboshi (not only the electrifi- 13 MWh in annual consumption. Seventy percent of the cation component), the share of the cost of power line company’s total energy consumption is for industrial use— extension is used as a benchmark to allocate the share of mainly heating, ventilation, air conditioning (HVAC), benefits accruing to the electrification investments in the refrigeration, irrigation (pumping, drip irrigation, and project area. spraying), and lighting. Except for heating directly sup- plied by steam, many other industrial processes (e.g., ven- tilation, refrigeration, and irrigation) require electricity. Case Study 5. Kenya: Oserian Part of the power generated by ODCL is distributed Flowers and within the company’s estate to the community (e.g., Geothermal Power staff housing, schools, and clinics) and sister companies (e.g., tourism lodge). Currently, 2,000 households are The Oserian Development Company Limited (ODCL) connected to electricity through a mix of power from operates a 216 ha flower farm—including roses, carna- ODCL’s own power generation (95 percent) and utility tions, and statice—situated in Kenya’s Nakuru County power (5 percent). However, 2,000 other households (map D.5). The farm produces and exports 380 million within the estate remain without an electricity connec- stems annually, and employs 4,600 people (table 4.23). tion. ODCL is planning an increase in power generation ODCL is a pioneer business in its use of heat from by improving generation efficiency (via installation of geothermal wells for internal power generation and con- a partial condenser). The improvement in efficiency is sumption; its 50 ha Geothermal Rose Project is the larg- expected to increase generating capacity by 0.4 MW. The est of its kind. In addition to geothermal heat, a 3.2 MW expansion project seeks to supply these additional house- generator is dedicated to powering the farm’s operations holds for basic electricity uses (e.g., lighting and mobile and distribution within its estate. Although the company phone charging) and to power such facilities as schools Lessons from Ongoing Power-Agriculture Integration Projects 55 Table 4.22: Economic Costs and Benefits of the IDSP Project, Mwomboshi Benefit 2016 2019 2031 Revenue from job creationa Jobs resulting from the project — 313 313 Present value of increase in employees’ income ($ million) 3.4 Increase in profit revenue Small-scale (MT) Tomato production with project 5,000 57,833 57,833 Maize production with project 1,000 6,403 6,403 Wheat production with project — 3,602 3,602 Present value of profit of extra production ($ million) 20.5 Commercial (MT) Wheat production with project 2,634 12,240 12,240 Maize production with project 3,512 16,320 16,320 Present value of profit of extra production ($ million) 3.5 Savings from import substitution Present value of wheat and maize import substitution savings ($ million) 6.7 Savings from household electrification Electrification rate (%) 15 46 97 Electrified households without project 93 101 144 Electrified households with project 93 283 598 Present value of household electrification savings ($ million) 0.2 Total present value of economic benefits ($ million) 34.0 Financial NPV of utility ($ million) 1.1 Share of line upgrade project cost to total IDSP project cost (%)a 2.6 Net social benefits ($ million) 0.9 Economic NPV ($ million) 2.0 Source: ECA and Prorustica (2015). Note: Assumes that present values are over the 15-year period (2016–31). a. Because the project has multiple complementary investments, it is hard to disentangle the benefits accruing to the power line extension without a simplifying assumption; it is thus assumed that the accrual of benefits to electricity versus other investments is in the same proportion as the accrual of costs. Table 4.23: Oserian Flowers and Geothermal Power Project at a Glance Project Overview Expansion of the estate geothermal generating capacity and its distribution network to power the farm’s operations and distribution within the estate (staff housing, community facilities, and sister companies). Commodities Floriculture. Description ODCL’s captive power generates 95 percent of its requirements internally. Industrial use (heating, ventilation, irrigation, and lighting) represents 70 percent of the company’s total energy consumption. Since no power is exported to the grid or sold beyond the estate, ODCL has a license from the Energy Regulatory Commission for captive power generation and distribution. Financial With a positive financial NPV, the planned expansion project of 0.4 MW and electrification of 2,000 Viability households is financially viable. Economic Positive economic benefits estimated at US$2.5 million. The main economic benefit is based on Viability increased household electrification and, as a result, the savings are due to lower energy consumption costs (e.g., less use of kerosene and no more payment for cell-phone charging services and disposable batteries). 56 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.9: Power uses and sources at ODCL Energy use (in GWh/a) Energy sources (in GWh/a) Losses, 1.5 Communities, 1.2 KPLC, 0.6 Other uses, 1.4 Flower farm, 8.9    Production, 12.5 Source: ODCL. and a clinic. The limited increase in capacity implies that After this generation expansion, it is expected that the monthly household consumption may be constrained; plant will generate an additional 2,500 MWh per year. This however, households willing to upgrade may get individ- will include 600 MWh to offset electricity bought from ual connections through the state-owned utility, Kenya KPLC, another 600 MWh to supply the local population that Power and Lighting Company (KPLC) (figure 4.9). does not yet have access to power, and the remaining 1,300 MWh to cover ODCL industrial processes (figure 4.11). Power Supply Options and Commercial Arrangements Financial Viability ODCL’s captive power generates 95 percent of its The planned expansion project of 0.4 MW and electrification requirements. Power is generated from a farm-operated of 2,000 households is marginally financially viable, with plant, and steam is bought from the Kenya Electricity a positive financial NPV of US$3,742. The costs incurred Generating Company (KenGen) under a 15-year purchase for generation and distribution expansion and operation are agreement. Since no power is exported to the grid or sold slightly more than offset by the revenue from cost reduc- beyond the estate, ODCL has a license from the Energy tion in electricity purchased from KPLC. An investment of Regulatory Commission for captive power generation US$1.2 million is required for expansion of generation (partial and distribution. ODCL supplies power to staff workers condenser) and the distribution network (conductors, trans- within the estate using a mix of geothermal generation former, and switchgear). Also, operating cost is not expected and the main grid supply. Households consume low levels to increase as the expansion will not consume additional of energy and are not metered individually, and KPLC bills resources (e.g., the same volume of purchased steam). In ODCL rather than individual households. Over the years, fact, the increased output will lower the per-unit cost from ODCL has developed a skilled, in-house engineering team US¢6 per kWh to US¢5 per kWh. The operating cost will dedicated to geothermal power generation. therefore amount to $125,000 (table 4.24). To meet unmet power demand and offset electricity In comparison, the savings from the reduced pur- purchased from the utility, an investment of US$1 million chases from KPLC amount to $342,000. Staff house- is planned for expanding geothermal plant capacity up to holds are to be supplied electricity free of charge. 3.6 MW (figure 4.10). An additional $0.2 million will be Charging households cost-reflective tariffs would incur required to finance the distribution network extension. additional costs due to metering and billing. Considering ODCL is considering charging electricity customers a these costs in the analysis shows that, in order to break cost-reflective tariff, but this would require an additional even, a cost-recovery tariff of US¢8 per kWh would be $0.2 million investment in individual meters. required. Lessons from Ongoing Power-Agriculture Integration Projects 57 Figure 4.10: Output of ODCL’s power plants and expected increased output 3.0 2.5 2.0 1.5 1.0 0.5 0.0 s 00 rs 00 rs 00 rs rs 18 rs 22 rs 00 rs 00 rs 00 rs 00 rs 00 rs 00 rs 10 rs 110 rs s 14 rs 15 rs 16 rs 17 rs 19 rs 20 rs 21 rs 23 rs 24 rs hr hr h h h 0h h h h h h h h h h h h h h h h h h h h 00 07 02 03 00 00 08 09 00 00 00 00 00 00 00 00 01 04 05 06 00 00 00 00 00 12 13 OW 202 power plant OW 306 power plant Additional output Source: ODCL. Figure 4.11: Electricity output of capacity expansion project and intended uses 600 MWh to connected community (o setting KPLC tari of 0.18 $/kWh) 1,300 MWh for industrial Output of capacity use of farm expansion project 2,500 MWh/a 600 MWh for new connections to 2,000 households Source: ECA and Prorustica (2015). Table 4.24: Financial Analysis, ODCL Economic Analysis Item US$ Amount The expansion project constitutes a relatively small Revenues 342,000 portion of the estate’s electricity use; most electricity is Power generation Opex costs 125,000ª used for irrigation and refrigeration. The main economic Capex costs 1,200,000b benefit from the expansion project is thus from increased Margin –983,000 household electrification and, as a result, the savings due to lower energy consumption costs (e.g., less kerosene use Discount rate (%) 10 and no more payment for cell-phone charging services Financial NPV ($ amount) 3,742 and disposable batteries). An electricity connection is Source: ECA and Prorustica (2015). estimated to save households US$11 per month, implying a. Assumes a cost per kWh of $0.05. b. Assumes $1 million for distribution and $200,000 for metering. 58 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa $2.5 million in total net economic benefit (NPV) over the projects in the small-scale tea subsector aimed at reducing life of the project. No significant impact is expected in factory operating costs, improving power supply reliability, terms of job creation or commercial development. and diversifying tea farmers’ revenue sources. The power generated from these schemes will be used primarily in the tea factories, with the surplus sold to KPLC under a Case Study 6. Kenya Tea Development power purchase agreement (PPA). KTDA is in the process Agency Holdings: Mini- of setting up several small hydropower projects for its tea Hydro Mini-Grids factories. One hydropower plant has been operational in the Imenti tea factory since 2010; an additional 17 proj- This case study analyses the mini-hydro based tea factory ects are in the pipeline, ranging from 0.5 MW to 9 MW, electrification project of the Kenya Tea Development eight of which are at an advanced stage of development, Agency (KTDA). The agency is planning the implemen- with feasibility studies completed. tation of several small-scale (≤ 15 MW) run-of-the river hydropower projects at various locations in Kenya to Power Demand serve a number of tea factories under its management (map D.6). Considering the near-term pipeline, along with the KTDA is the single largest producer and exporter of operational Imenti plant, the total installed capacity is tea in Kenya. The company was created in 2000, sub- 24.4 MW. About 40 percent of power generated will be sequent to privatization of the Kenya Tea Development used primarily for the tea factories’ self-consumption, Authority. KTDA is the holding company of a number supplying mainly tea industrial processes. The remaining of subsidiaries owned by small-scale tea companies. The 58 percent of output will be sold to KPLC under a PPA agency currently manages 63 factories in Kenya’s small-­ and feed-in-tariff (FiT) scheme. Farmers will benefit from scale tea subsector. Currently, its network covers about the electricity supplied to the factories that they partially half a million small-scale farmers, with each tea factory own, but residential electricity connections will only be owned by 5,000–10,000 tea farmers (table 4.25). provided through KPLC, and not directly though KTDA. KTDA Power Company Limited, a subsidiary of Approximately 187,500 small-scale farmers, representing KTDA, is charged with consolidation, investment, and 25 tea factories, will benefit from these power projects management of energy initiatives undertaken by tea to run their farming activities. Currently, 70 percent of factories managed by KTDA. Notably, KTDA Power neighboring households (i.e., more than 130,000 farm- Company supports the development of hydropower ers) lack access to electricity. Table 4.25: Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids at a Glance Project Overview Development of hydropower plants powering tea factories and staff housing, and selling surplus power to the grid. Commodities Tea. Description The operational power plant and eight projects have a total installed capacity of 24.4 MW. About 187,500 small-scale farmers, representing 25 tea factories, will benefit from these power projects to run their farming activities. Mini-hydro plants provide a more reliable power supply to tea factories at lower cost and avoid the need for backup generators. Financial Viability Evaluation of a sample project (North Mathioya) shows that the project is financially viable, with a NPV of US$3.3 million. Revenues accrue from the sale of power to the grid and cost savings by tea factories. Economic Viability The same sample project is evaluated as economically viable, with a NPV of US$10 million. Direct and indirect impacts on rural electrification include the following: electrification of staff housing, reduced connection costs for surrounding households, development of stand-alone home systems. About 30,000 households will benefit from electricity connections. Lessons from Ongoing Power-Agriculture Integration Projects 59 Power Supply Options Figure 4.12: KTDA’s North Mathioya hydropower project: financial benefits The KTDA tea factories have two feasible supply options and power sold for meeting their power requirements: (i) purchase from the main utility at the retail tariff or (ii) self-generate Revenues (total $3.3m) electricity through the planned hydropower projects. Grid-supplied electricity is often unreliable, with frequent outages and voltage fluctuations. The need for a reliable power supply for tea operations requires investment in backup diesel generation, which adds to the overall cost of electricity. Where feasible, a captive mini-hydro genera- tion plant, with the ability to sell excess power to the main Power sold grid, is an attractive option both financially and in terms of (total 28.4 GWh/a) increased reliability. In terms of commercial arrangements, KTDA Power Company leads the project development cycle (e.g., permitting acquisition, securing land, and raising capital) and forms special purpose vehicles (SPVs) in the form of regional power companies for each project (e.g., North Mathioya Power). The factory farmers served by the Revenues from PPA, 51% Power consumed, 36% mini-hydro plant are shareholders, and raise 35 percent of Revenues from tea Sale of power to KPLC the investment cost as equity from deductions of farm- farms, 49% under PPA, 64% ers’ tea revenues. Electricity to residential consumers in Source: ECA and Prorustica (2015). the area will be provided through KPLC and not directly through the project. of about $1 million per year, as well as pressure to reduce tariffs along with KPLC’s national rates. Given these Financial Analysis assumptions, subsidies for both capital expenditure and operating expenses would be required. The financial analysis focuses on the North Mathioya (5.6 MW) hydropower project from the perspective of Economic Analysis the SPV owners. Project revenues derive from the sale of electricity to the grid at the FiT.13 The remaining elec- Although KTDA power projects are not involved in the tricity sold to tea factories is valued at the avoided cost retail sale of electricity to neighboring communities, they of grid plus diesel backup electricity at US¢16 per kWh have several direct and indirect impacts on rural electri- (figure 4.12).14 fication. First, they provide electricity to staff housing, The costs include the capital and annual operating which represents an average of 60 households per factory. expenditures of the generation plant incurred by the Second, they may facilitate grid access for the surround- SPV, at US$22.5 million and $165,800, respectively. ing households by reducing connection costs. Third, these Comparing the present value of the stream of revenues areas will be targeted by a pilot project—led by the KTDA and costs, the project is estimated to be financially viable, subsidiary, Greenland Fedha (microfinance institution), with a NPV of $3.3 million (at 10 percent cost of capital) and the KTDA Foundation—which aims to finance solar and an IRR of 13 percent. home systems (SHSs) for farmers and support their grid Although the project does not include household or connections. community electrification, except for factory staff hous- The estimate of economic benefits is based on facil- ing, a simplified financial analysis shows that such activity itating households’ access to electricity connections. Tea would be financially unviable without subsidies. Despite factory activities remain unchanged, although they gain the relatively high margin between household retail rates access to a more reliable, cheaper source of power supply. (US¢20 per kWh) and the PPA rate (US¢9 per kWh), Approximately 30,000 households will benefit from distribution and retail would require an additional capital electricity connections, which will offset their expenditure expenditure of US$15 million and administrative expenses on traditional or more expensive forms of energy. 60 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa The project will facilitate grid connection by con- concept in Tanzania also shows this cause-and-effect necting the generation facility. Costs are estimated at relationship between irrigation and processing. Increase US$500 per grid connection, with a monthly electricity in the scale of processing activity can lead to a significant bill of $3 per household. Also, the above-mentioned increase in power demand. SHS scheme in place for farmers will further increase The seasonality of power demand from the agricul- connections,15 with an average household savings of $11 ture sector can significantly constrain a project’s viability. per month.16 Thus, development of the North Mathioya Large seasonal differences in electricity-dependent agri- hydropower project will provide households net economic cultural activity will impact the cost recovery of invest- benefits; the project’s NPV is $6.7 million, implying ments in electricity supply. In such cases, it is important $10 million in total economic NPV.17 to consider ways to mitigate the impact of a variable load. One option, especially for mini-grid or captive generation, is the ability to sell excess power to the grid, as in the Key Conclusions from cases of mini-hydro development in Tanzania (Mwenga) the Case Studies and Kenya (KTDA).18 Increased processing activities in the post-harvest season may complement electricity The six case studies discussed in this chapter offer varied demand from irrigation, and irrigation itself may reduce contexts for power-agriculture integration. Each is unique seasonality in agricultural production and thus electricity in terms of the type of anchor load and country setting; demand by allowing multi-cropping (e.g., in the case of thus, one must be cautious about generalizing from the les- Zambia’s Mkushi farming block). sons learned from any particular case. Keeping this in mind, Finally, when considering agricultural anchor loads, this section discusses key findings from the six case studies it is more risky for the investment to depend on a single in terms of large power loads, supply options, financial and large customer since any negative shock to the customer economic viability, and financing of development. would negatively affect operating revenues for the elec- tricity supplier. For this reason, agricultural clusters (e.g., Large Power Loads Sumbawanga in Tanzania) can be used to increase the viability of rural electrification. Clusters development, by The viability of providing electricity depends critically on design, has load diversity and thus involves less risk than the existence of a large and stable demand for electricity reliance on a single anchor load. While not included in the (or supply, especially if the grid is supply constrained). case studies discussed in this chapter, the presence of a In rural areas, it is likely that the largest single source of private electricity supplier and private off-takers will price power demand is either agriculture or an agriculture-­ any such risk into the supply contract, thus increasing the related commercial activity. Residential electricity could price of electricity for all customers. In such cases, diversi- also be a significant source of demand (e.g., in the case fied cluster development can also help reduce the price of of Tanzania’s Sumbawanga agriculture cluster); however, electricity. The public sector may also help mitigate this risk this demand is often relatively dispersed, which reduces its through a grid connection and FiT, subsidies to increase the viability. customer base, or various guarantee/insurance instruments. In rural agricultural areas, irrigation is often the single largest potential source of electricity demand, Supply Options as exemplified in Tanzania’s Sumbawanga agriculture cluster, Zambia’s Mkushi farming block and Mwomboshi’s Most of the grid extension projects are justified by irriga- IDSP. These projects also show that the loads for agro-­ tion development, with agro-processing as a supporting processing activities (e.g., milling and extrusion) are activity. These developments require cultivating suitable comparative smaller, suggesting that the latter activi- commodities (e.g., maize, wheat, rice, and sugar), typ- ties, taken alone, may not be sufficient to justify rural ically grown on large-scale commercial farms, enabling electrification investments. These several projects also large production volumes. Small-scale farmers can then highlight how irrigation and processing are often linked. be incorporated alongside; however, they also need other The Zambia cases show how increased yields from irriga- forms of support, including access to a reliable water tion are an important prerequisite for the development supply, good physical and market infrastructure, and clear of large-scale processing activities; the agriculture cluster land with good quality soils. Lessons from Ongoing Power-Agriculture Integration Projects 61 The case studies discussed indicate that the national if the regulation allows it. Given this difficulty, financial grid usually plays an important role in the viability of rural viability, in most cases, depends on the ability to sell electrification investments—either in the form of the bulk power and lower costs. The Oserian geothermal and main supply option for agricultural and rural electricity KTDA projects show that estate-type developments demand (e.g., the Sumbawanga cluster in Tanzania) or (floriculture and tea in these respective cases) can under- as the main off-taker of the locally generated electricity take financially viable electricity investments, benefiting from a small power producer (e.g., the Mwenga mini-­ from reduced electricity costs and selling excess electric- hydro mini-grid in Tanzania). Whether the grid is the most ity to the grid. Another example is the case of the IDSP in viable supply option depends on various factors, includ- Zambia, where a grid extension was financially viable from ing distance to the grid, size and stability of electricity the utility’s perspective, owing to proximity to the grid demand, grid reliability, and local resource potential for (i.e., lower costs) and complementary investments in a generation. large irrigation scheme that increased electricity demand. Supplying rural electricity demand though small power In contrast, grid extension to the Mkushi farming block, producers (SPPs) depends critically on local generation also in Zambia, was not financially viable for the utility, potential (e.g., for mini-hydro, geothermal, and bio- despite a capital cost-sharing arrangement with benefi- mass). Viable generation potential can be a cost-effective ciary farmers. option in cases where the grid is far away, unreliable, or The choice of optimal tariffs—such that costs are expensive. In the latter case, especially, SPPs may benefit recovered and electricity consumption is affordable to primarily from selling to the grid and supplying local agri- farmers, businesses, and other customers—depends on cultural activities and residential customers in the process the size of the financial surplus generated from electricity (e.g., Tanzania’s Mwenga mini-hydro mini-grid). consumption and the constraints on how to allocate it Companies specializing in the agriculture or agribus- across various suppliers and customers. Additional consid- iness sectors may be unwilling to enter into electricity erations, such as parity with the main grid tariff, are the generation and, especially, the distribution business. main determinants (e.g., Mwenga mini-hydro mini-grid in This would be a departure from their core activities and Tanzania). may not be financially attractive enough to change their If there is flexibility in setting tariffs, then the range business model. In this respect, a variety of arrangements of feasible tariffs would be determined by the difference are possible, depending on the context and capacity of the between the customer’s willingness to pay (WTP) and entities involved. For Kenya’s Oserian geothermal project the supplier’s willingness to accept (WTA).20 A custom- and KTDA’s mini-hydro project, the companies chose to er’s WTP will be determined by the monetary benefit develop and operate the generation plant and supply their from consuming a unit of electricity. For households, operations, preferring to sell power to the grid and leave this may be a reduction in spending on their current retail power supply to the utility. For Tanzania’s Mwenga energy supply options, which are usually more expensive mini-hydro project, by contrast, RVE manages the mini- and less reliable (e.g., kerosene lamps or batteries). For grid generation and distribution, including retail power agricultural consumers, it may be driven by a reduction sales.19 in backup energy supply and/or increased revenues from higher productivity. A supplier’s WTA will be determined Financial and Economic Viability by development and operating costs, often represented by the levelized cost of electricity (LCOE) (table 4.26). The case studies discussed show that a rural electrification Assuming the WTP is more than the WTA, an optimal project can be financially viable where there is a creditable tariff may be negotiated based on some surplus allocation large off-taker and access to concessional loans/grants rule. Otherwise, if the WTP is lower than the WTA, the for capital investments. All six projects were estimated to government must step in to provide subsidies to bridge generate economic benefits well in excess of associated the gap as long as the project remains economically viable. costs, thus implying that all were economically viable. For all six of the cases analyzed in this chapter, the Tautologically, financial viability rests on the ability economic viability was high. For projects that are not to charge cost-reflective tariffs. In the case of mini-grid financially viable, economic viability is an important cri- development, charging consumers a tariff that is much terion to determine whether subsidies should be provided higher than the grid tariff might be difficult to do, even and at what level. Even with financial viability, subsidies 62 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.26: Typical LCOE Values for Small-Scale Generation and Distribution Systems Generation System Power Plant Size Range Capital Expenditure LCOE Operating Time Technology (kW) (US$/kW) (US$/kWh) (hours/year) Diesel genset 5–300 500–1,500 0.3–0.6 Any Hydro 10–1,000 2,000–5,000 0.1–0.3 3,000–8,000 Biomass gasifier 50–150 2,000–3,000 0.1–0.3 3,000–6,000 Wind hybrid 1–100 2,000–6,000 0.2–0.4 2,000–2,500 Solar hybrid 1–150 5,000–10,000 0.4–0.6 1,000–2,000 Distribution System LCOE Distribution Type Voltage Level (US$/km) Required Length Low-voltage 400 V 5,000–8,000/km 30 customers/km Average connection cost: $350/customer; average distribution cost: $200/customer. Medium-voltage 33 kV 13,000–15,000/km Total ($/kWh) 0.25–1 Source: IED Reference Costs for Green Mini-Grids. may be incorporated into the project to achieve other the spread of fixed costs, especially capital costs, across goals, such as grid parity in terms of tariffs or greater a larger pool of customers with diverse peak-load pro- adoption of electric irrigation. files. For example, since productive users need electricity during the day and households’ peak load is in the evening, Financing of Development the system peak load should be lower than the sum of individual peak loads. However, load balancing requires All six projects analyzed shared two common issues: an analysis of load profiles to optimize supply, and the (i) making projects financially viable and (ii) providing level of additional benefit depends on the proportion of funding for viable projects. Several ways have been iden- capital costs in total costs and the load matching between tified to make projects financially viable. To benefit from customers. The utilities—owing to their larger-capacity economies of scale, capacity for local generation can be cross-subsidization and ability to spread costs over a wider increased beyond the level of local demand, and surplus customer base—are usually in a better position to do so. power can be sold to the grid. This option is particularly As detailed for the Mwenga and KTDA projects, sell- relevant in countries that have introduced FiT programs ing power to more reliable customers, such as the utilities, set above the utility’s avoided costs. Selling excess power increases a project’s viability since anchor customers are makes it possible to lower the per-megawatt cost, but assumed to be better payers. This is especially true in relies on the ability to sell excess generated power. For countries where clear schemes for renewable energy FiTs example, the capacity of Tanzania’s Mwenga mini-hydro have been introduced with dedicated funding. Although mini-grid is greater than what the tea estate requires; relying on the utility still depends on its ability to afford therefore, the surplus is sold to the utility and nearby rural payments, the anchor-customer approach has reduced customers. the risk of the utility’s non-payment by giving certainty Another option, as done for the main grid exten- on tariffs. sion projects in Zambia (Mkushi and Mwomboshi), is to Finally, the role of subsidies to cover certain costs require the beneficiaries to partially finance projects and should be highlighted. All of the distributed schemes ana- share the development costs with major customers. In lyzed in this chapter have received subsidy payments to this way, farmers partially contribute to the capital costs decrease the level of cost recovery through retail tariffs. in exchange for receiving power. A further option is load This approach contributes to ensuring maximum capacity balancing across beneficiary categories, which enables development, increasing the project’s NPV, improving Lessons from Ongoing Power-Agriculture Integration Projects 63 tariff affordability for customers, and attracting private- to be financially unviable, developers can be encouraged sector participation. Subsidies are particularly necessary to expand their customer base to capture additional sub- for most privately developed, small-scale projects under sidies, prioritizing smaller customers close to each other 5 MW. By subsidizing household connections, which tend rather than larger ones. endnotes 1. The analysis presented in these case studies is indicative only and not a comprehensive feasibility study. 2. The only exceptions are projects based on quite expensive sources of power generation for small demand loads. 3. SAGCOT aims to facilitate the development of seven agribusiness clusters along the southern corridor of Tanzania’s Southern Highlands. 4. This comprises 3 MW from a 66 kV line into Zambia, 5 MW from a mini-grid in Sumbawanga, and a 2.6 MW mini-grid in Mpanda; both are isolated, diesel based mini-grids operated by TANESCO. 5. ECA and Prorustica estimates, consistent with the SAGCOT investment blueprint, constructed from own analysis and various official sources. 6. Other products such as cassava and livestock are also likely to demand electricity for processing, but for the sake of simplicity, are not included in the calculations here. 7. According to Tanzania’s national census, Rukwa had 1 million inhabitants in 2012. 8. The cost calculations consider all capital and operating expenditures; the calculations are based on ECA analyses conducted for small-scale systems in Kenya, Tanzania, and elsewhere in Sub-Saharan Africa. 9. Assumes that the factory operates 16 hours per day, 6 days a week for 10 months out of the year. 10. EU funds were through the African, Caribbean, and Pacific Group of States facility; and REA funding was supported by the World Bank’s Tanzania Energy Development and Access Project (TEDAP). 11. Assumes that the area’s power demand from irrigation is 1 kW per ha and average irrigating hours per year are about 1,900 (with a 15 percent load factor), representing in part the seasonality in demand for irrigation. 12. Assumes that the average mill has a power demand capacity of 400 kW and operates 5,000 hours per year. 13. US¢9.29 per kWh under a FiT. 14. Assumes a diesel generation cost of US¢60 per kWh (KTDA) and an overall tariff decrease of 5 percent annually. 15. Since this analysis focuses on the impact of an anchor load on household electrification, we restrict it to grid-connected households. 16. Observed for mini-grid development in Kenya. 17. If we assume that 50 percent of the 30,000 households connected are from SHS, then the household net benefits increase to US$14 million and the overall NPV to $17.2 million. 18. Apart from the mitigating impact of seasonal variation, the ability to sell excess power to the grid also helps invest in large genera- tion capacity and reduces costs due to economies of scale in generation. 19. Enabling small-scale, private power generation and distribution requires clear regulations and purchasing processes (e.g., PPAs and FiTs); regulations in Tanzania are relatively transparent in this regard. 20. The difference between WTP and WTA is a measure of the total surplus generated by the electricity sale/consumption. Opportunities to Harness Agriculture Load for Rural Electrification Chapter 5 W hat is Sub-Saharan Africa’s potential for the power load of bulk water pumping and infield irrigation harnessing power-agriculture synergies for systems were made.1 rural electrification? This chapter considers For each farming type, the production volume was this question, using a simulation model and used to calculate the milling load for the area, based on case studies from Ethiopia and Mali—two countries that assumptions about the proportions of milled production. exhibit a range of innovative options moving forward to With total milling volumes, the total load requirement 2030. Before turning to the case studies, the chapter for milling was estimated, based on the load characteris- presents a hypothetical case illustrating the conditions tics of an assumed average mill. Household and business under which power demand from agriculture could be connections for the given area were also estimated, based economically viable. on assumptions about a consistent population density and members per household, connection rate, household power consumption, and proportion of this load for busi- Simulation of Power Demand ness consumption. in a Stylized Agricultural Setting The stylized analysis from the simulation model helps to determine the general features of power demand from A simplified simulation model was developed to analyze agricultural areas. Based on the average power demand the relationship between agricultural activity, power from agricultural sources, the results show that a fairly demand, and the geographic area that a power supply large area of coverage would be required to aggregate would serve (table 5.1). The model assumed a theoretical sufficient electricity demand from customers; based on circular area around the generation source, with electric- the model assumptions, a 50 km radius area would, on ity consumers distributed uniformly throughout. Further average, aggregate 60 MW of demand. simplifying assumptions were made about what percent- In the simulation, as in the case studies, irrigation age of this area was under cultivation and the proportions accounts for a substantial proportion of power demand split between small-scale and commercial farmers. The from agriculture (figure 5.1).2 The irrigation power load electricity demand from each of the two farmer groups is dependent on choice of crops and availability of bulk were estimated separately, with differing proportions of water. Some systems with a large body of available area under irrigation and yields (on rainfed and irrigated water nearby the infield irrigation system may require summer and winter crops). The model assumed that there little bulk water pumping; however, in cases where were two crops: summer maize and winter wheat. Across water must be pumped into storage before utiliza- Sub-Saharan Africa, maize is a common summer crop on tion, additional electricity is required. As such, total both mixed-used commercial and small-scale farms. In observed power loads for irrigation are in a range of the winter months, irrigated wheat is commonly grown. 0.5 kW–2.0 kW per ha. Based on the areas under irrigation, assumptions about 64 Opportunities to Harness Agriculture Load for Rural Electrification 65 Table 5.1: Assumptions for Typical Area/Agricultural Activity/Power Demand Model Assumption Basis Small-scale Value Commercial Value Overall Value Proportion of total land area Observations of other 25 under cultivation (%) large-scale production areas Proportion of farming type Observations of other 70 30 within cultivated area (%) large-scale production areas Proportion of irrigated land Observations of other 20 50 (%) large-scale production areas Summer crop yield (rainfed) Maize yields 1.5 6 (MT/ha) observed Summer crop yield (irrigated) Maize yields 4 8 (MT/ha) observed Winter crop yield (irrigated) Wheat yields observed (not 2 5 (MT/ha) grown without irrigation) Proportion of crop milled Observations of other 25 80 (%) production areas Irrigation load requirement Average, based on 0.3 1.0 (kW/ha) schemes observed Milled load Average mill, 200 (kW) consultant calculations Hours of operation (hrs/day) Average mill 16 Days of operation (hrs/year) Average mill 313 Population density (per km2) Comparison with other 50 countries People per household (no.) Comparison with other 5 countries Household connection rate Comparison with other 50 (%) countries Peak household consumption Various household power- 0.3 (kW) consumption studies Business load as proportion Various rural business 50 of household load (%) power-consumption studies Source: ECA and Prorustica (2015). The relatively low load for processing suggests that (figure 5.2). By contrast, the impact of the proportion of the machinery used for typical post-harvest processing crop processed is relatively low, especially as this load is operations (e.g., mills) does not require large amounts already minimal. of electricity, in part, because of the small size; also, it may be in operation for fewer hours in a year. Thus, most crop-processing loads are fairly small for the volume Simulation Study 1. Ethiopia: Power processed, with the exception of such activities as sugar Generation from Sugar Estates processing, which provides much or all of its own power. The total power load for a given area is highly sen- Sugarcane is an important crop in Ethiopia (map D.7). sitive to the assumed area under commercial irrigation, Indeed, the Ethiopian Sugar Corporation (ESC) aims reiterating the importance of irrigation to power loads to increase national annual production nearly eightfold 66 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 5.1: Power demand and Figure 5.2: Sensitivity of power load breakdown for a given area radius to changes in percent of commercial irrigation 80 70 Peak power load (MW) 60 60 Peak power load (MW) 50 40 40 30 20 20 0 10 5 10 15 20 25 30 35 40 45 50 0 Area radius (km) 0 10 20 30 40 50 Irrigation Processing Households Businesses Area radius (km) 15% 35% 50% Source: ECA and Prorustica (2015). Source: ECA and Prorustica (2015). Table 5.2: Ethiopia: Power Generation from Sugar Estates Project Overview Self-generation of power from bagasse and sale of power surplus to the main grid. Commodities Sugar. Description Sugar processing and irrigation are the largest sources of electricity demand. Irrigation makes it possible to extend the sugarcane production season and therefore smooth the annual profile of both production and processing. Processing and refining are the most power consuming activities in the sugar estate. Typically, a sugar processing plant can produce enough electricity from bagasse to meet its own electricity demand, and sell excess power to the grid. The viability of connecting such processing plants to the grid depends on the amount of excess power produced, the cost of producing it relative to other sources, and additional customers that can be connected. Financial From the utility’s perspective, extending the grid to the sugar estate is not financially viable—the net Viability present value (NPV) is negative because the utility does not benefit from sales to the estate, which self-supplies; from the sugar estate’s standpoint, the project is highly profitable (US$139 million). Economic The economic NPV for the whole period is positive ($367 million), thus justifying development of the Viability project. within five years. To do so, the government has launched Power Demand the Sugar Development Programme, with the objective of upgrading existing estates and commissioning new ones Agriculture (Irrigation). Traditional sugarcane production (table 5.2). is heavily water dependent. Irrigation ensures year-round This simulation analyzes a representative example of production of the crop and therefore a smoothing of the power-agriculture integration on sugar estates in Ethiopia. annual profile of processing activity. This means that sugar Sugar estates have the potential to generate power facilities operate throughout the year with a consistent from bagasse, a natural by-product of sugar refining. electricity demand. Hypothetically, the potential electricity generation is Irrigation is also a major source of power demand in enough to cover the electricity needs of the refinery and the sugarcane production process. In Ethiopia, irrigated associated facilities and sell the surplus to the main grid or land is expected to increase from 1,500 ha to 9,000 ha other supply schemes. over 20 years. The associated power demand from Opportunities to Harness Agriculture Load for Rural Electrification 67 irrigation over the same period is expected to rise from the sugar estate, depending highly on production volume. 0.8 MW to 4.7 MW,3 with power consumption increasing Considering forecasts in terms of yield rates and produc- from 2,340 MWh to 14,040 MWh (table 5.3).4 tion increases, the power requirements for processing irri- Agriculture (Processing and refining). Processing gated sugarcane will amount to 6,300 MWh in year 1 of and refining are the most power-consuming activities in the hypothetical model, rising to 37,800 MWh five years later. For processing rainfed production, power consump- Table 5.3: Total Power Demand tion will increase from 3,150 MWh to 18,900 MWh over from Agriculture and Residential/ the same 20-year period (figure 5.3). Commercial Loads Beyond processing, refining activities also consume power for centrifuging raw sugar and crystallization. Over Power Capacity Energy Demand the 20-year period, electricity consumption from refining Demand (MW) (MWh/year) is estimated to rise from 300 MWh to 1,800 MWh, while Demand Source Year 1 Year 20 Year 1 Year 20 power load will increase from 0.09 MW to 0.54 MW.5 Irrigation 0.8 4.7 2,340 14,040 Staff housing. In addition to agricultural needs, sugar Processing 2.9 17.5 9,450 56,700 estates also require power for staff housing and other Refining 0.1 0.5 300 1,800 supporting activities. Given that the average household electricity consumption in rural Ethiopia is about 0.10 kW Residential 0.1 1.5 384 3,783 (including staff (increasing to 0.15 kW by year 20),6 total electricity housing) demand from staff housing is estimated at 0.02 MW in Commercial 0.1 0.6 278 2,780 year 1, increasing to 0.21 MW by year 20. Residential/Commercial demand. In this model, Source: ECA and Prorustica (2015). the area is not yet connected to the grid, but a 30-km Figure 5.3: Estimated energy demand and peak load, by sector a. Energy demand 100,000 Electricity demand (MWh) Irrigation 80,000 Processing Refining 60,000 Sta housing 40,000 Rural households Nonresidential 20,000 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years b. Peak load 20 Irrigation Processing Peak load (MWh) 16 Refining 12 Sta housing 8 Rural households Nonresidential 4 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years Source: ECA and Prorustica (2015). 68 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa grid extension is finalized once the sugar factory is built. Table 5.4: Sugar Factory Power Thanks to the proximity of houses and the factory, the Generation in Years 1 and 20 electrification rate rises sharply to 85 percent by year 4. Electricity Electricity As the population grows from 28,500 to 50,386 by Generation Demand year 20,7 with household growth following the same Sugarcane Capacity (MW) (MWh/year) trend,8 the total electricity load from rural households will reach 1.3 MW by year 20. For commercial activities Processing Type Year 1 Year 20 Year 1 Year 20 surrounding the sugar estates, consumption is expected Irrigated 3.8 22.5 12,600 75,600 to increase from 278 MWh in year 1 to 2,780 MWh by Rainfed 5.8 35.0 6,300 37,800 year 20 (table 5.3).9 Total 5.8 35.0 18,900 113,400 Source: ECA and Prorustica (2015). Power Supply Options and Commercial Arrangements Bagasse is commonly used to generate electricity in sugar Beyond meeting its own power needs, the sugar factories. It is mainly used as a boiler fuel to generate steam factory can generate surplus power.10 This supports the to meet the sugar factory’s heating and power needs. The development of estate activities, especially irrigation, level of net electricity generation assumes (i) a bagasse before enough on-site bagasse has been produced. It generation potential of 29 MT for every 100 MT of sugar- also covers shortfalls in power generation during planned cane produced and (ii) a 70 kWh generation capacity for annual maintenance when the mills are not operating every MT of sugarcane. Since irrigated and rainfed process- (May–September) (table 5.5). ing of sugarcane do not occur simultaneously, the power The capital cost of extending the grid line 30 km to capacity of generation equals the maximum capacity of the the sugar estate and surrounding villages is US$2.4 million two, that is 47 MW by year 20 (table 5.4). (table 5.6). Table 5.5: Net Power Generation from Sugar Factory by Year 20 Power Capacity Hours of Energy Demand Agricultural Activity Demand (MW) Operation/Year (MWh/year) Irrigation (A) 4.7 3,000 14,040 Processing (B) 17.5 3,360 56,700 Refining (C) 0.5 3,360 1,800 Total demand 22.7 72,500 Power generated during processing (D) 35 113,400 Net power surplus D − (A + B + C) 12.3 40,900 Source: ECA and Prorustica (2015). Table 5.6: Capital Cost Assumptions for Grid Connection Cost Component No./Distance (km) Unit Cost Cost (million US$) 230 kV shunt/line/transformer (thousand $/unit) 15 25 0.4 Associated switchgear (thousand $/unit) 1 120 0.1 33 kV line (thousand $/km) 50 14 0.7 11 kV line (thousand $/km) 120 10 1.2 Total 2.4 Source: ECA and Prorustica (2015). Note: Costs estimates are based on those for similar projects in Ethiopia’s 2014 Electrification Master Plan; cost assumptions include connecting villages along the power line (i.e., 33 kV and 11 kV lines and transformers). In reality, the estate may feed back power to the villages from the substation. Opportunities to Harness Agriculture Load for Rural Electrification 69 Currently in Ethiopia, however, no sugar factory From the sugar estate’s perspective, the combina- exports its power to the grid because of the country’s tion of heating and power from bagasse combustion is (i) low electricity tariffs and (ii) unclear regulations on a fundamental asset for sugar processing and refining. conditions of exporting power to the main grid. A feed-in- The project’s financial viability depends on the following tariff (FiT) proposal, which aims to provide incentives to factors (table 5.8): private investors, is expected to become law in 2016 and ºº Capital costs, linked to development of the whole should clarify those conditions; thus, under future devel- estate, including land improvement, buildings and opment plans, power sold to the grid will be at the FiT. It equipment, and staff housing. is unlikely that sugar estates will sell directly to residential ºº Production costs, including employee wages, seeds, customers; this will be left up to the electricity utility. harvesting, loading, transport, maintenance, and electricity costs. Financial Analysis ºº Expected revenues from sugar sales and power sales. The project’s financial viability can be analyzed separately The project is highly profitable for the sugar estate, from the respective standpoints of the utility and the with a NPV of US$139 million. As mentioned above, the sugar estate. From the utility’s perspective, extending the large financial benefits for the sugar estate create ample grid to the sugar estate is not financially viable; the esti- scope for a negotiated arrangement of capital cost sharing mated NPV is negative, at US$ –1.5 million (table 5.7). to improve the utility’s financial viability. The viability is driven by the amount of power purchased by the utility, the margin between retail tariff and the price at which electricity is purchased from the sugar fac- tory (possibly the FiT), and the cost of extending the grid. Table 5.8: Sugar Estate Capital Costs, The price at which the utility purchases power from Assumptions for Production Costs, the independent power producer (IPP) is confidential. In and Revenues the absence of actual data, it is assumed that the utility tariff margin is US¢1 per kWh, which amounts to 40 per- Component Value cent of the domestic tariff.11 Capital costs (million US$) The project is not viable for the utility, in large part Land improvement ($3,500/ha) 41.9 because it does not benefit from sales to the estate, which Buildings and equipment 80.5 self-supplies. Subsidies would thus be required for project Staff housing ($5,000/house) 7.0 development. Given the significant financial benefits Present value of total capital costs 129.4 that will accrue to the sugar estate from the project, one option could be to have the sugar estate contribute to Production costs capital costs. Average wage ($/month) 100 Permanent employees (months/year) 12 Temporary employees (months/year) 7 Table 5.7: Financial Analysis from the Seeds costs ($/ha) 515 Utility’s Perspective Harvest cost ($/MT) 6 Present Value Loading cost ($/MT) 2 Component (million US$) Transport to sugar mill ($/MT) 3 Net revenue from sales 2.7 Maintenance (% of capital expenditure) 3 Expenses (Opex, losses, depreciation) 1.8 Present value of total production costs 311 Capital cost 2.4 (million US$) NPV −1.5 Revenue (million US$) IRR (%) 7.6 Present value of sugar sales 573 Present value of exported power to the grid 6 Source: ECA and Prorustica (2015). Present value of total revenues 579 Note: The discount rate is 10 percent over the 20-year period; of total capital costs, operating costs account for 3 percent, while Sources: Agritrade; ECA and Prorustica (2015); ESC; IEA; losses and depreciation each account for 5 percent. National statistics. 70 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Economic Analysis over land ownership; despite the government’s ability to make quick investment decisions regarding state-owned The project’s total economic benefits, estimated at about property, identifying large tracts of high quality agricul- US$410 million, comprise household energy cost savings, tural land is difficult in Ethiopia. Third, regulations on sugar estate profits, job creation, and import substitution exporting power to the grid must be clarified by defining (table 5.9). tariff rates that guarantee investors a price for selling The economic NPV over the period, about generated power from bagasse to the utility. Finally, US$367 million, equals the sum of the net social bene- selling power to the utility carries off-taker risk; delayed fits linked to the electrification project (figure 5.4), the payments for power sold or even payment defaults would financial NPV, and the present value of the sugar estate greatly impact the sugar factory investor. investment cost (table 5.10). Various factors could hinder the development of such agriculture-power schemes in Ethiopia. The first one is Simulation Study 2. Mali: Mini-Grid funding availability for grid extension; however, given the Expansion for Productive Users project’s associated economic benefits, funding from the government, development partners, or even cost sharing Mali is a regional success in rolling out private mini-grid with the sugar estates could be sought. Second, for green- concessions for rural electrification (map D.8). field development, investors face issues about uncertainty Table 5.9: Net Economic Benefits of Grid Extension to the Sugar Estate Benefits Year 1 Year 5 Year 20 Household energy savings Electrification rate (%) 21 85 85 Households electrified (no.) 1,479 6,752 9,964a Savings from grid electrification per household ($/month) 17 Total savings on energy consumption (million $) 0.025 0.12 0.17 Incremental income to the sugar estate Production revenues (million $) 14.8 74.2 89.2 Production costs (million $) 11.0 39.5 46.7 Sugar estate’s profit (million $) 3.8 34.7 42.5 Sugar estate jobs created Monthly salary ($/month) 100 Permanent jobs created (no.) 933 4,663b 5,595 Temporary jobs created (no.) 1,588 7,939 9,527 Total salaries (million $) 2.2 11.1 13.4 Non-sugar jobs created Jobs created (no.) 1,260 6,301 7,561 Salaries paid (million $) 0.13 0.63 0.76 Import substitution New production of sugar (MT) 42,000 210,000 252,000 Value of import substitution (million $) 1.3 6.3 7.6 Total economic benefits (million $) 7.5 52.9 64.4 Source: ECA and Prorustica (2015). a. The difference in the number of connected households between years 5 and 20 is related to population growth, which is expected to increase by 2.89 percent. b. Assumes 0.37 permanent job and 0.67 temporary job (working 7 months a year) created by hectare—Estimation based on the number of employees in Metehara sugar factory in Ethiopia. Opportunities to Harness Agriculture Load for Rural Electrification 71 Figure 5.4: Net social benefits of grid extension to sugar estate (years 1–20) 60 50 Economic benefits (million $) Household energy saving 40 Incremental income for sugar estate 30 Jobs created in sugar estate 20 Non-sugar job creation 10 Import substitution 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years Source: ECA and Prorustica (2015). Table 5.10: Economic Net Present Value In 2015, it has 255 operating concessions, with a total of Extending the Grid to the Sugar installed capacity of 22 MW. However, mini-grid opera- Estate tors face key challenges, including the saturated capacity of their schemes and low revenues, which hinder invest- Value ment in capacity expansion. Limited power-generation Item (million US$) capacity has constrained the mini-grids’ ability to supply Financial NPV of Ethiopian Electric Power −1.5 households and serve productive users. The current ser- Corporation (EEPCO) vice level—limited daily hours (typically in the evenings) Present value of investment cost of sugar −41.9 and tariffs that are higher than on-site diesel generators estate (usually above US$0.50 per kWh)—are inappropriate for Net social benefits 410.0 meeting agro-industry power requirements. As a result, Economic NPV 367 productive users in off-grid areas use their own diesel generators as a more competitive power supply option Source: ECA and Prorustica (2015). Note: The discount rate is 10 percent over the 20-year period. (table 5.11). Table 5.11: Mali Mini-Grid Expansion for Productive Users at a Glance Project Overview Capacity expansion of an existing hybrid mini-grid (diesel-solar PV) to serve productive users. Commodities Agro-industrial activities. Description The Koury mini-grid is reaching a point of near saturation as generation capacity is fully taken up by existing household demand. However, small-scale commercial and agro-industrial activities in Koury (milling, water pumping, and bakeries) present significant opportunities for supplying unmet power demand. Attracting powered small businesses as mini-grid customers would require incentives to (i) lower tariffs, (ii) supply electricity during the daytime, and (iii) replace manual equipment with electricity powered machinery. Financial From the perspective of SSD Yeelen Kura, the rural energy services company, the Koury mini-grid is Viability in a fragile financial situation. However, the capacity expansion project is profitable, thanks to a higher payment rate, additional revenues, and proportionally low capital expenditure and operating expense (with a NPV of €103,000). Economic The economic NPV for the expansion project is slightly negative (−€18,000) as no significant savings Viability are expected from agro-industrial customers, who currently use individual diesel generators. However, the project could become economically viable if other economic, environmental, and social benefits were considered (e.g., reduction in CO2 emissions, reduced reliance on imported fuels, and exposure to price fluctuations). 72 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Based on a representative example of an existing mini- private mini-grid projects, an expansion project has been grid, this simulation study analyzes how agro-­industrial designed to assess the viability of supplying agro-industrial activities may improve mini-grids’ financial viability, loads. The simulated study also evaluates the potential for while benefiting from a more sustainable and competi- adding value to agricultural activities in rural areas through tive source of electricity. Based on the potentially lower mini-grid supplied power. Powered agricultural activities costs of hybrid solar photovoltaic (PV) projects, the study can indeed improve rural communities’ revenues and explores the potential for attracting agro-industrial power therefore potentially increase mini-grid operators’ profit demand to mini-grids. Given that there is no precedent (box 5.1). for tying medium- or large-scale industrial processing to Box 5.1: Isolated Mini-Grid Systems in Mali: Existing and Potential Power DemanD In Mali, large-scale irrigation schemes are gravity fed, with electric power used only for small diesel or petrol-­ powered pumps. Four key commodities that could benefit from greater access to electricity are mango, rice, shal- lot, and shea kernel. Mango. Mali’s Bamako and Sikasso regions are particularly favorable for growing mango. But to export larger vol- umes, Mali must handle various issues related to market transport and product handling, notably reliance on cold chains (e.g., fixed and mobile chilling facilities). Considered a production hub, Sikasso would be the logical location to set up a temperature-controlled mango packing house. Areas outside Sikasso not yet connected to the main grid have limited potential for extending or replacing cold-chain packing-house facilities; such areas are mainly served by isolated mini-grids or diesel gensets. An alternative value chain to fresh mango is processing mango pulp or nectar. Mali has only lightly exploited this value chain due to the lack of transforming infrastructure, irregular sourcing from small-scale farmers, and dis- tance to markets. Excess mango production can be used for dried mango or canning. However, high start-up costs and working capital would be required; this is not economically viable, given Mali’s low margins and small scale. Rice. Mali is a net importer of rice. Its rice production system uses gravity-based irrigation without mechanized bulk water pumping or infield irrigation. On the processing end, rice milling (husking) occurs throughout small- scale private milling operations, using both diesel-powered mobile or fixed husking machines and fixed-site mills. However, Malian milled rice is of low quality, with a high volume of broken rice. In some high production areas connected to the main grid (e.g., the 100,000 ha Office du Niger), larger-scale, fixed-site mills have been devel- oped with higher quality rollers that reduce broken rice, thereby adding value to the volume of rice sold. In addition to pure processing activities, post-hulling bran-hull biomass is used to generate power for the mill and related activities, as well as lighting on the premises and for staff housing facilities. Shallot. Mali could potentially become a major West African exporter of shallot, thanks to favorable growing conditions. Shallot is grown on small-scale farms across the country, and 90 percent of production ends up in local urban markets. Shallots can be provided fresh or variously processed (e.g., dried, crushed, or machine sliced, [potentially] using solar drying panels or improved solar heaters). Electricity is required for only two processes: (i) pounding and drying and (ii) slicing and drying. Since consumers prefer the fresh form of shallot, the market for transformed shallots is limited, and higher pro- duction costs induced by processing cannot be justified. The main opportunity is extending the market season for fresh shallot, capturing value from price fluctuations due to reduced market volumes. More efficient stocking and drying techniques would make fresh shallot available 4–6 months beyond the regular growing season and Opportunities to Harness Agriculture Load for Rural Electrification 73 over a year for its dried form. Because storage and drying processes require small amounts of power, there is little opportunity for power to add value to the commodity’s value chain, especially in areas not yet connected to the main grid. Shea kernel. Mali is a minor market player in kernels and butter, capturing less than 10 percent of global demand. Penalized for poor quality and yield, unreliable supply, and higher costs, Malian kernel exporters can hardly com- pete with other West African producing countries. Vegetable oil firms in Europe, India, and Japan dominate the global market, while West Africa accounts for only a handful of industrial extraction facilities, some of which work on a toll basis for global companies. Though Malian farmers have an incentive to produce higher quality kernels, they have little incentive to expand their kernel processing capacity, given the limited potential benefits (Derks and Lusby 2006). Manual processing of shea fruit includes kernel removal from pits; drying, moulding, and grinding kernels into paste; and kneading paste into separate solids and oils. These activities could benefit from mechanization, but weighed against the required investments, the benefits are not obvious, especially given the low labor costs and limited access to capital. Sources: FAO and Authors. Power Demand from Mini-Grids schemes that require water pumping rely on decentralized pumps spread over large areas. In Mali, households consume 90 percent of mini-grid electricity, which is mainly used for lighting, with peak load occurring during evening hours. The Koury mini-grid, Power Supply Options and Commercial located in a rural community of Yorosso circle (cercle) Arrangements in the Sikasso region, is operated by SSD Yeelen Kura, The Koury mini-grid is reaching a near saturation point as a private operator that manages 21 concessions12 and generation capacity is fully taken up by current demand. has started to hybridize its mini-grids with solar PV. More than 20 percent of the generated electricity is from In 2012, Yeelen Kura added 100 kWp of solar PV to diesel generators (figure 5.6). The variable cost of thermal the existing 112 kW of thermal capacity, making power generation, at €0.40 per kWh,13 and the cost of direct available 10 hours a day (typically from 3 p.m. to 1 a.m.). consumption (below €0.20 per kWh) suggest the advan- Because of the mini-grid demand profile, the solar tages of expanding solar PV capacity. output produced by PV generators is stored in batteries, Notably, expansion of solar PV could enable the elec- which increases energy losses and capital expenditure tricity provision for productive activities since they require (figure 5.5). power mainly during the daytime. Direct consumption of The Koury mini-grid currently supplies 180 MWh per solar output would (i) avoid energy losses in the battery year, mostly for households. Out of 3,371 households bank and (ii) reduce the battery bank size relative to living in the area, 556 are already connected to the mini- capacity of the solar PV generator. grid, at an average consumption level of about 24 kWh per To attract businesses as mini-grid customers, incen- month. tives would be needed to (i) lower tariffs, (ii) supply The opportunities for supplying unmet power demand electricity during the daytime, and (iii) replace manual from small-scale commercial and agro-industrial activi- equipment with electricity powered equipment. Figure 5.7 ties in Koury are significant. Although such activities rely shows the impact of adding the daytime loads of pro- mainly on their own diesel or petrol engines or genera- ductive users, along with a 50 kWp matching capacity tors, they represent a total potential energy demand of expansion of the solar PV system (totaling 150 kWp) on 7,755 kWh per month—about a 50 percent addition to the Koury mini-grid load profile.14 the existing energy production of the mini-grid power This capacity expansion is assumed to fall under plant (table 5.12). Irrigation is not expected to play a the existing rural electrification program of the Malian significant role for the mini-grids, given that most irriga- Agency for Development of Household Energy and tion in Mali utilizes gravity fed schemes, and small-scale Rural Electrification (AMADER) and therefore benefits 74 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 5.5: Koury mini-grid: Electricity consumption patterns 100 90 80 Load 70 Solar output 60 50 kW 40 Solar output Loads supplied not directly 30 by battery bank used goes to and/or genset 20 battery bank 10 0 6:00 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00 Public, 5% Commercial & industrial, 5% Residential, 90% Source: SSD Yeelen Kura. Table 5.12: Potential Addition of Small Agro-Industrial Activities and Other Businesses Typical Energy Consumption Total Consumption Business Type Number (kWh/month) (kWh/month) Milling or grinding (maize, rice, shea kernel) 6 300 1,800 Water pumping 2 300 2,520 Bakery (electric mixer) 1 300 450 Mechanical workshop (welding, grinding, drilling) 2 1,260 300 Media center (computer, printer) 1 450 135 Petrol station (pumps) 1 150 300 Small shops (refrigerators, freezers, TV, lighting) 10 135 2,250 Total 7,755 Source: GERES and SSD Yeelen Kura. Opportunities to Harness Agriculture Load for Rural Electrification 75 Figure 5.6: Energy generation profile from capital expenditure subsidies, with ownership of at Koury site infrastructure remaining with the government and the operator regulated under contract. 600 Taking a conservative approach, it is assumed that agro-industrial customers’ willingness to pay will be 500 capped at the costs of running individual diesel gensets. This implies that the tariffs needed would be lower than 400 current household tariffs. kWh/day 300 Financial Analysis 200 From the perspective of SSD Yeelen Kura, the current 100 financial situation of the Koury mini-grid is somewhat precarious (table 5.13, figure 5.8). Although operating 0 expenses are covered by revenues, the 20 percent capital expenditure contribution of the private operator is not 4 14 14 5 5 5 t-1 1 1 -1 v- c- n- b- recovered through tariffs. In order to achieve a 10–15 per- ar Oc No De Ja Fe M cent return, the project receives up to 80 percent of Diesel genset output (kWh/day) capital expenditure subsidy from the government. Equity Average solar output (kWh/day) investment and reinvestment in capacity expansion and Max solar output (kWh/day) replacement of major parts (e.g., batteries and gensets) Source: SSD Yeelen Kura. cannot be recovered. Figure 5.7: Koury mini-grid profile: Additional commercial and industrial loads 160 140 120 Total load 100 Solar ouput New commercial/ 80 kW industrial load 60 40 20 Daytime loads supplied directly by solar 0 6:00 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00 Public, 12% Commercial & industrial, 28% Residential, 60% Source: SDD Yeelen Kura. 76 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Economies of scale, daytime energy use, and falling consumption from the mini-grid. Largely as a result of solar PV prices imply that the expansion project could be the significant capital subsidies, the expansion in gener- attractive as it allows for additional revenue with relatively ation capacity is financially viable from the perspective low capital expenditure and operating expense. The oper- of SSD Yeelen Kura, with a positive NPV (table 5.14). ating costs will marginally increase due to higher expenses However, if viewed from the perspective of AMADER or in maintenance and administration, but will be offset by a the Government of Mali, the asset owners, the financial lower level of generation losses due to direct consumption returns are negative (essentially including the subsidy of solar power (reducing the need for storage) and lower costs in the calculation). use of thermal generation. Along with the capital subsidy to the developer, this implies a lower average tariff and Table 5.14: Financial Analysis of Capacity creates the incentive for new customers to switch from Expansion of Koury Mini-Grid their current diesel generators to daytime electricity Item Amount Commercial and industrial customers served (no.) 20 Table 5.13: Current Financial Situation Average total consumption (MWh/year) 80 of Koury Mini-Grid Average retail tariff (€/kWh) 0.40 Item Amount Payment rate (%) 90 Households served (no.) 556 Additional revenues (€) 28,800 Average total consumption (MWh/year) 160 Operating costs (€)a 5,600 Average retail tariff (€/kWh) 0.55 Capital costs before subsidy (€)b 189,000 Payment rate (%) 80 Capital costs after 80% subsidy (€) 37,800 Revenues (€) 70,500 Project cash flows NPV after subsidy (€)c 103,000 Operating costs (€)a 55,400 Project IRR (%) 56 Capital costs before subsidy (€)b 831,000 a. Including the cost of fuel and increased maintenance and Capital costs after 80% subsidy (€) 166,200 administrative expenses; excluding depreciation. NPV after subsidy (€) (259,700) b. Including an additional investment of 50 kWp of solar PV; assumes no additional expense in the distribution network. a. Including corporate overhead and fuel, maintenance, and c. Additional parameters affecting cash flows and thus the administrative expenses; excluding depreciation. calculation of NPV include (i) reinvestment in batteries (every b. Including the cost of solar and diesel powered generation and 6 years) and inverters (every 12 years), which are not subsidized; battery storage, as well as costs of the distribution network, civil (ii) increased fuel costs, given a PV system degradation rate of and electrical works, and engineering; current (2015) costs are 0.5 percent per year; and (iii) a 10 percent weighted average used (i.e., €5,300 /kWp, excluding the distribution network). cost of capital (WACC). Figure 5.8: Operating expense and capital expenditure distribution a. Operating expenses b. Capital expenditures Other O&M, 12% Distribution, 30% Other, 32% Operator overhead, 52% Fuel costs, 36% Storage, 20% Solar PV, 18%     Opportunities to Harness Agriculture Load for Rural Electrification 77 Similar to most other mini-grid projects in Mali, the same. No significant benefits are expected to accrue to Koury mini-grid is not financially viable without large agro-industrial customers as most would not save sig- subsidies. While capacity expansion to integrate commer- nificantly on electricity costs by switching from margin- cial and agro-industrial loads would improve the financial ally more costly individual generators to the mini-grid. performance slightly, it is unlikely to be enough to make This is unlikely to lead to an expansion in processing the grid financially sustainable without subsidies. Some activity and thus would have little associated economic measures that could improve mini-grid performance benefits, as reflected in the slightly negative economic include implementing better load management practices NPV for the expansion project (−€18,000). However, to reduce energy storage needs, reducing administrative including additional economic, environmental, and social expenses, and enhancing revenue collection through pre- benefits that are not quantified (e.g., reduction in CO2 paid meters and remote monitoring. Despite these poten- emissions and other pollutants or reduced reliance on tial improvements, the profitability for hybrid solar-diesel imported fuels and exposure to price fluctuations) could mini-grids would require a revision of the subsidy struc- make the project economically viable with a positive ture and current tariff levels. NPV. Benefits could also accrue to the agriculture To reach financial viability while serving productive sector if it has suppressed electricity demand, which can users, capital expenditure subsidy requirements, under be met much easier through mini-grid capacity expan- assumptions for a greenfield mini-grid similar to Koury, sion rather than expansion in the size of the individual would have to reach 96 percent of a one-off capital generator. expenditure subsidy for initial development and replace- ment of major parts. With more optimistic assumptions Main Inferences and Institutional (e.g., a better load management to reduce solar PV Arrangements losses, improved revenue collection, and lower battery-­ replacement costs), the subsidy requirement could be In order for the potential large-scale opportunities to inte- reduced to 77 percent of capital investment.15 grate productive users into Mali’s mini-grids to succeed, several major barriers need to be overcome (box 5.2). Economic Analysis Available financing for rural electrification is a crucial issue for both AMADER and the mini-grid operators. Solar PV capacity expansion to supply productive users Insufficient and uncertain availability of funding for capital has limited economic benefits. For households and cost grants has limited AMADER, while private operators existing customers, the cost of supply would remain the cannot afford to scale up on their own. Box 5.2: Large-Scale Opportunities for Power-Agriculture Integration in Mali Agribusiness development in Mali could have a critical impact on job creation and poverty reduction. With over 40 million ha of arable land and an irrigation potential of 560,000 ha, Mali’s agribusiness sector could benefit from favorable agro-ecological conditions and regional food demand. But constraints along the agribusiness value chain (e.g., lack of access to energy and other basic infrastructure, lack of access to finance, and poor sector gov- ernance) limit its development. Beyond developing a value-chain strategy, a spatial approach is promoted to boost productivity growth, diversification, and value addition. Since Mali is a vast country, the creation of growth poles, clusters, and trade corridors in the agribusiness sector has real significance. In the Sikasso region, conversion of the Randgold Resources–operated Morila gold mine into an agro-industrial cluster is an example of opportunities to realize large-scale power-agriculture integration. Currently, the mine’s power demand is covered by cumulative available capacity of about 26 MW, with 187,000 MWh of potential production from 10 diesel generators. Once closed and replaced by the agropole in 2017, esti- mated power needs may drop to 8–10 MW (Randgold Resources estimate), and Randgold Resources plans (continued) 78 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 5.2: Continued to hybridize the generation plant and set up a mini-grid aiming to power medium-voltage agribusiness activities, including the following: ºº Henhouse (installed capacity of 130 kW with a monthly consumption of 21,000 kWh). ºº Juice production and packaging (installed capacity of 1 MW for 4,000 bottles per hour and 30–60 packets per minute). ºº Air-conditioned logistic facility (installed capacity of 20 kW with a monthly consumption between 700 kWh in freshness period and 1,250 kWh in peak season). ºº Slaughterhouse (installed capacity of 100 kW with a daily consumption of 2,200 kWh). ºº Fish preservation units (installed capacity of 200 kW per unit). ºº Carton packaging unit (installed capacity of 2.5 MW). ºº Other activities (e.g., aquaculture, mango production, and beekeeping). The mini-grid also aims to connect 100 small- and medium-sized enterprises (SMEs) that require low-­ voltage, unitary power below 30 kW for transforming and cooling crops (e.g., cereal, shea kernel, and vegetables). Powering SME activities will also facilitate the connection of 15,000 surrounding households and community facilities. This integrated solution optimizes the use of infrastructure to support large-scale agro-industry projects and secure raw materials and supply inputs through a partnership between smallholders and large players. It can also play a role in bringing rural power to the surrounding community. Source: Randgold. One way to improve the financial viability of mini- differentiated tariffs by customer type or time of use grid operators would be through diversification of the would allow operators to cross-subsidize between cus- service offering to include other energy solutions (e.g., tomer categories. Finally, access to capital for productive stand-alone systems).16 Also, clear regulations with scope users is critical. Indeed, agribusiness players willing to con- for tariff-setting flexibility would improve the ability and nect to mini-grids will have to invest in electric machinery incentives for supplying productive customers. In addition, to replace manual equipment. endnotes 1. The highly stylized setting of the model is thus less appropriate for considering such value chains as milk, poultry, and even floricul- ture, which have a different spatial distribution of production. While it is possible to adapt the model to these and other settings, it is considered beyond the scope of the present analysis and left for future work. 2. Based on the model assumptions, irrigation load demand is about one-and-a-half times that of all other power demand combined. 3. Assumes an average mill requires 35 kWh to process 1 MT of sugarcane. For other sugar estates in Africa, per hectare power demand could be significantly higher if the potential for gravity fed flood irrigation is not as high. 4. Assumes 3,000 irrigation hours per year. 5. Assumes that the same operating hours as for processing are applied and that a modern inverter driven batch centrifugal con- sumes about 1 kWh per MT of sugarcane processed. 6. Using a metric of 0.37 employees per ha and considering 4 workers per house (with no family), there are 233 houses in year 1, which rise to 1,399 houses from year 6 onward. 7. The area occupied (300 km2), average rural population density (94 people per km2), and average population growth rate (2.9 per- cent per year) are used to estimate the surrounding population. Opportunities to Harness Agriculture Load for Rural Electrification 79 8. On average, each household has 5 members; the number of households totals 5,700 in year 1, rising to 10,077 by year 20. 9. Assumes that total power demand is half that of residential demand and that nonresidential consumers use electricity roughly 4,368 hours a year. 10. A grid connection is essential for exporting power. 11. Domestic tariff is US¢2.3 per kWh. 12. 2015. 13. Analysis was done in Euro (€) currency since the local currency (CFA Francs) is pegged to the Euro, using a diesel price of 650 FCFA per liter (1€ per liter); consumption of 0.33 liters per kWh; and 20 percent in auxiliary losses, lubricants, and other main- tenance costs. 14. Assumes no need for further investments in the distribution network or additional diesel generators. 15. Assumes that integration of at least one-third of daytime commercial and industrial loads, 10 percent reduction in solar PV losses from current levels through better load management, 90 percent revenue collection, 20 percent reduction in administrative expenses, and a 20 percent reduction in battery replacement costs within the next 4–5 years due to battery technology development. 16. Partnerships with suppliers of solar pumps or solar mills could also be attractive since many operators are progressively building on an expertise in solar PV technologies. Conclusions Chapter 6 T his chapter highlights the study’s key findings gives a sense of the investment in generation capacity on Sub-Saharan Africa’s potential for leveraging that will be required to meet agricultural needs and the complementary investments in agriculture and addition to rural electricity demand that is expected owing electricity to contribute to the region’s rural pov- to the agriculture sector. erty reduction; these include overall results of the study For the 13 agricultural value chains selected, electric- and case studies, along with key learnings from the com- ity demand could increase by 2 GW by 2030, represent- mon challenges encountered by the case study projects ing nearly half of the 4.2 GW of potential incremental (chapters 4 and 5). It then recommends steps that can be increase in electricity demand from agriculture. Among taken to maximize the joint benefits of expanded electric- the value chains examined, poultry has the largest per ity access and increased value added along the agricultural hectare electricity demand. Together, maize, rice, and value chains. cassava account for 83 percent of total incremental demand in agro-processing to 2030. The largest source of electricity demand for the 13 commodities is commercial Key Findings irrigation, which has the greatest potential to develop large power loads across a range of farm sizes. Overall results This study finds that creating opportunities to piggyback Case Study Findings viable rural electrification onto local agricultural devel- The case studies show that power supply options for agri- opment depends on a variety of site-specific factors culture and rural electrification benefit from economies (e.g., scale and profitability of agricultural operations, of scale. Small-scale power systems (less than 5 MW), crop, terrain, type of processing activity, and other local which may provide a useful source of power service for conditions). Rural electrification opportunities will be agricultural processing and household connections, are best created by agro-processing activities that generate rarely financially viable without subsidies.1 When financial electricity demand close to rural population centers, gen- viability is not a key driver (or constraint), a full range of erate adequate income to cover electricity supply costs, activities can benefit from electric power. Once economic are sufficiently large in relation to household demand, and benefits are considered, a strong case can be made for have relatively low seasonal variation. providing effective subsidies to cover gaps in financial By 2030, electricity demand from agriculture is viability. estimated to double from its level today, to about 9 GW. The case studies also confirm that irrigation consti- Between 2016 and 2030, irrigation is expected to provide tutes the largest power demand from agriculture; without about three-fourths of the incremental demand (3.1 GW), it, demand from agricultural activities (except sugar with agro-processing accounting for the remainder processing) tends to be small. Large land areas are needed (1.1 GW). The overall magnitude of electricity demand to support a major irrigation load. Economic viability is 80 Conclusions81 likely for all except the most expensive sources of power because of untested procedures and lack of precedents, generation for small loads. Power supplies generate pro- notably concerning retail tariff approbation. portionally high economic value, primarily through social Another major barrier to development is the lack and indirect economic benefits. of clear electrification plans (e.g., Tanzania and Kenya). Among the agriculture schemes examined, only Information about future developments of the national large-scale development of irrigation-based agriculture grid and concession protection is crucial for dispelling and sugar estates could justify a large grid connection on a developers’ reluctance and avoiding potential friction purely financial basis. Their requirements—not all of which from tariff differences between customers. The case of are readily available in Sub-Saharan Africa—include rela- large-scale, mini-grid development in Mali shows how tively clear and empty land with good quality soils, reliable regulation and strong government buy-in can, despite supplies of sufficient water, and high quality physical and large subsidies, allow for development (chapter 5, case market infrastructure. Suitable commodities include those study 2). This example also illustrates that clear power typically cultivated on large-scale farms: maize, wheat, regulations are a necessary, but insufficient, condition for sugar, rice, soybean and barley. successful project development. For example, Tanzania’s The projects show that successful integration of agri- Mwenga mini-hydro mini-grid—one of the first projects culture and power system development requires physical of its kind to deal with regulations about water rights, land and market infrastructure to facilitate market access for access, import laws, and building permits—has entailed inputs and produce. In Zambia, for example, the strate- significant delays. This experience highlights the need to gic location of the Mkushi farming block has improved extend regulations beyond the power sector to include its development viability. The farming block is situated related sectors (e.g., trade, water, land, and environmental alongside the main T2 Highway and Tazara Railway, which management). connect Lusaka and the Copperbelt in Zambia to Tanzania For every case study analyzed, the technical and and on to the Dar es Salaam commercial port, providing financial capacity of key institutions—the utility, regula- access to markets for both inputs and produce (chapter 4, tor, and rural energy agency—to implement and permit case study 3). In Tanzania, the site of the Mwenga mini-­ development is perceived as a challenge. The weak finan- hydro generator is situated far from the main TANZAM cial status of the utilities prevents them from being able Highway between Dar es Salaam and the Zambian border; to develop financially viable projects without external sup- however, the Tunduma, Mufindi Tea Estates, which drove port. Furthermore, their cash-strapped situation increases the mini-grid’s development, is located only 10–15 km the risk of nonpayment for the power supplied by private from the main road (chapter 4, case study 2). developers, which negatively impacts project costs and Key learnings from common challenges. The main tariffs and, as a result, power affordability. If feed-in-­ barriers faced by the case study projects are linked to tariffs (FiTs) are not capped at the utility’s avoided costs, the regulatory environment, electrification planning, and the situation could worsen, further deteriorating the institutional and financial capacity. To succeed, projects utility’s viability. From the perspective of power-sector must be implemented within a stable legal environment regulators, the extra cost and delays resulting from inex- that imposes requirements and provides protection. The perience in negotiating various supply arrangements may right degree of regulation must then be found. Viewing be a hindrance to developing private power generation, the absence of regulations as an opportunity to reduce distribution, and supply. costs increases risks considerably because of uncertainty. In Tanzania, grid extension planning is generally a Light-handed regulation of small-scale electricity systems transparent and efficient process, largely included in is generally more favorable to developers and operators. the Power System Master Plan. Although grid densi- In Tanzania, the small power producer (SPP) framework fication is currently the priority for the Rural Energy allows private operators to function as power distributors Agency (REA),3 grid extension projects, such as the one and retailers, charging fully cost-reflective tariffs.2 This in Sumbawanga, are also part of the plan, considering type of regulation should tackle the economic barriers the potential economic benefits. However, TANESCO of unaffordability and uneconomic supply. In Kenya, (Tanzania Electric Supply Company Limited) has a fragile developers have been reluctant to pursue the opportunity financial situation, which has consequences for new proj- to implement electricity distribution and retail schemes ect investments. 82 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa As the mini-hydro project illustrates, dealing with the commercial farmers. Given the extra profits potentially social and environmental considerations that any project generated by a more reliable power connection, 10 large-­ of this nature raises (e.g., water resource management, scale farmers agreed to fund half of the capital costs. forestry, village lands, land acquisition, and environmental Beyond these key success factors, some hurdles still management) is still lacking in transparency and coordi- need to be overcome. The inability of national generation nation. Both the regulatory framework and the processes capacity to support higher peak load and the resulting for project development are open to political interference. load shedding create a major risk for farmers. In response, Coupled with transmission planning, generation capacity backup diesel solutions were bought to secure produc- must be developed sufficiently and consistently to support tion, and irrigation activities were carefully planned to grid extension. avoid under-voltage. Even though the irrigation project in Tanzania generally provides developers clear guidance Mwomboshi will increase peak load slightly, it will require on tariffs, concession security, and system registration; an increase in national capacity in order to reduce risks. however, the Mwenga experience shows that application Conscious about the critical role played by agriculture in of the SPP framework, particularly in setting tariff levels, Zambia’s economy, central authorities are actively intend- continues to place unnecessary pressure on developers. ing to expand the national installed generation capacity so For mini-grid developers, especially those that sell power as to limit shortages and load shedding.4 to TANESCO, the risk comes more from the off-taker. In Kenya, small-scale, private-sector renewable Late payments create financial pressure for the opera- energy projects have had little success, despite the large tor. Third-party support can therefore help by providing number of FiT applications, owing to their high devel- bridging loans. Land access, another obstacle for project opment and transaction costs. Although permits for developers, can be overcome by developing mutually sym- self-generation are straightforward and allow industrial biotic relationships with the local community and district firms, notably in the agribusiness sector, to lead renew- authorities and gaining their support. Project develop- able energy projects, it may take up to three years to ment is still a complex process. The developer, Rift Valley acquire licensing and securing of land. The power reg- Energy (RVE), expects to sign about 3,000 agreements ulator is working to streamline licensing procedures for to access land over which its network runs. projects relying on FiTs. Also, land and way-leave issues Tariff affordability for consumers continues as one can be mitigated thanks to the involvement of project of the most critical issues for mini-grid development. beneficiaries. Although RVE is free to set up its tariffs under the SPP A second major concern in Kenya is related to the framework, pressure from social and political interests private sector’s involvement in electricity distribution and continues to make it difficult to do so. The profitability supply. Currently, Kenya Power and Lighting Company of projects is therefore supported by significant capital (KPLC) is the only licensed company undertaking distri- subsidies. bution and supply activities. The regulatory framework In Zambia, favorable conditions have facilitated the is still unclear on whether other companies are legally design and implementation of the Mkushi farming block allowed to enter this business. Other obstacles concern and the Mwomboshi Irrigation Development and Support tariffs and subsidies. Although not explicitly required Project (chapter 4, case studies 3 and 4, respectively). At under the regulations, retail tariffs cannot be higher than a national level, the Mkushi grid-extension process was KPLC’s tariff schedule. This principle could jeopardize the efficient and transparent; the Zambia Electricity Supply financial viability of any small-scale initiative. Moreover, Corporation (ZESCO) led the feasibility study, with the subsidies are not available for private companies. support of a consulting company. Also, land management was clarified by the 1995 Land Act, which gave investors more visibility and reduced the risks of long-term projects. Recommended Actions to Promote In addition, some solutions were put in place to improve Power-Agriculture Integration the financial feasibility of both projects. To overcome the utility’s cash-strapped situation, the investment costs of Power utilities in Africa, like those elsewhere in the world, grid extension in Mkushi were shared between ZESCO and often focus exclusively on their own business, rarely Conclusions83 venturing outside their limited realm of expertise. But can work both ways; that is, electricity companies can a narrow institutional approach—focused only on wires, prioritize certain regions with existing or potentially high poles, and consumer billing—means that many of the levels of agricultural production, while rural development potential development benefits from electricity remain or agricultural agencies can also target areas that will be unrealized. When used by a combination of households, able to take advantage of the many possible productive commercial businesses, industry, and agriculture, electric- use impacts of electricity. The benefits of breaking down ity provides a wide array of benefits and revenue. Ignoring institutional barriers between power, agriculture, and rural these broader possibilities not only limits the possible development programs result in higher revenues for the benefits for communities and the country overall; most utility companies and higher levels of development for importantly, it neglects the potential revenue for power regions and countries. producers from the increased electricity sales. Promote Farmers’ Productivity Improve Institutional Coordination For their part, the electricity companies can promote In order to realize their full potential as providers of internal units responsible for demand-side management electricity service, power companies need to engage with and encourage the productive and efficient use of elec- related programs to develop complementary strategies. tricity. Productive use units can be responsible for pro- In the case of agriculture-power integration, this means moting the adoption of productivity enhancing machinery establishing electricity expansion strategies in collab- in agriculture, from planting to irrigation and harvest. oration with rural development, agriculture, and other Such units can coordinate with other organizations, such institutions and agencies. as farmer associations, nongovernmental organizations Such complementary strategies can take several (NGOs), and various other local- and regional-level orga- forms. One is to provide electricity to those rural areas nizations already working closely with farmers to increase with the most potential for commercial activities, which is productivity. typically the case. For example, electricity can be prior- The barriers to farmers’ productively using electricity itized in areas with a large irrigation potential, combined in rural areas are relatively easy to overcome. They typ- with access to markets for agricultural goods. Machinery ically include a lack of simple knowledge about available used in agricultural production, including small threshers, machinery, lack of a local vendor, and inability to purchase can be promoted as part of a package to encourage elec- machinery on credit. Given the high expense of using tricity use in agriculture. For areas receiving electricity diesel-powered engines for grain processing, campaigns for the first time, agricultural fairs can be set up by local could be developed by local governments to promote the governments to demonstrate the possible machinery that substitution of electricity for diesel engines among farm- can be used in agriculture. ers in areas just gaining access to electricity. In many countries of Sub-Saharan Africa, lines of Integrate Planning of Power, Agriculture, credit to farmers and other agricultural entrepreneurs and Rural Development could be augmented by local banks so as to enable the adoption of new machinery (e.g., irrigation pumps, mills, Coordination with related institutions and agencies can and small stationary threshers). In many cases, existing also benefit the electricity companies. Once a rural devel- lines of credit are mainly for seed and other supplies opment agency realizes that an area is to receive electric- provided at the beginning of the growing season, with ity, it may make plans to include those communities in loans paid off after harvest. The electricity companies its program, meaning that the region would have access could work with banks and other credit agencies to set to electricity in conjunction with other inputs important up credit lines specifically for the purchase of electric for rural development. Thus, institutional cooperation machinery. 84 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa endnotes 1. Exceptions may include hydropower and biomass. Under favorable geographical conditions, low-cost hydropower can be provided; also, biomass can support agricultural activities, but seldom beyond those of the agriculture estate. 2 Especially for systems under 100 kW, for which no approval is required from the Energy and Water Utilities Regulatory Authority (EWURA), Tanzania’s sector regulator. 3. Tanzania’s rural electrification planning is led by the REA, with the operational support of TANESCO and support of development partners. The July 2014 National Electrification Program Prospectus identified key development centers for connection to the main grid, which will not be effectively initiated before 2016. While the prospectus suggests that some flexibility in identifying additional centers could be considered in order to develop synergies between power and agriculture, such uncertainty can be unhelpful to plan- ners of rural electrification projects. 4. In addition to these technical issues, environmental considerations must be taken into account. The impacts of these projects on the environment, especially those that involve dam construction, have a non-negligible significance. Annexes Annex A: Business Models for Agricultural Development A ttaining productivity increases by focusing on villages can be linked to water and power supplies at low small-scale agriculture and small- and medium- marginal cost. In cases where nucleus farms and out- sized agribusiness enterprises, as compared grower schemes incorporate community-owned land on to larger scale commercial systems, is a major a leasehold basis, local residents can be given an equity challenge. Larger scale farming provides economies of scale share in the farming business, as well as access to low-cost in production and input supply, including finance. This is irrigation. Likewise, farmer producer associations can be particularly observable for relatively large, uneven invest- integrated into commercial value chains through out- ments (e.g., machinery, irrigation, and electricity instal- grower or contract farming models. lation) or working capital needs. Smaller farms tend to be Other evolving agribusiness models enable the less efficient when collateral requirements affect their “crowding in” of both public and private investment into ability to raise working capital (Collier and Dercon 2009). defined areas of a country. Due to economies of scale, However, this does not mean that one farming system farmers and agribusinesses are most likely to be success- should entirely preclude the other as there are examples ful when they are located in proximity of each other and of successful crop-specific, small-scale projects, partic- related service providers. Such programs as the Southern ularly in the higher value commodities. Meeting growing Agricultural Growth Corridor of Tanzania (SAGCOT) demand will require improved performance of informal is focusing initially on 5–6 clusters within the southern value chains and their linkage with formal value chains corridor where there is potential, over time, for profitable to gain much needed capital, knowledge and skills, and groupings of farming and processing to emerge.1 Each market contacts. Achieving this will require a more flexible cluster requires investment along the full agriculture value approach to farming systems, currently being evaluated, chain. Some of these investments are public goods (e.g., whereby farming is seen as a business, with small-scale rural infrastructure and electrification) that must come farmers and their communities forging stronger linkages from the government and its development partners; with modern agribusiness. The key is to ensure economies others can expect to earn a financial return and will come of scale around aggregated small-scale farmer models from the private sector (figure A.1). linked to larger commercial agribusiness. For example, Building on existing operations and planned invest- new integrated small-scale farmer models are being ments, the clusters are likely to bring together agricultural tested in northern Ghana with the development of a research stations, larger nucleus farms and ranches with commercially run, professionally managed maize farmers outgrower schemes, commercially focused farmer associa- association, Masara N’Arziki. Such small-scale farmers tions (like those described above), irrigated block-­farming associations are being developed with the technical help operations, processing and storage facilities, transport and and financial support of commercial inputs and commod- logistics hubs, and improved “last mile” infrastructure to ity marketing companies; Masara N’Arziki currently has farms and local communities. more than 10,000 small-scale members producing over When occurring in the same geographical area, these 100,000 MT of maize for local and regional markets. investments result in strong synergies across the agri- Other models that create scale include the nucleus culture value chain, helping create the conditions for a farm hub and outgrower models. These allow small-scale competitive, low-cost industry. Similar corridor programs and emergent farmers to benefit from access to infra- are operational in Mozambique (e.g., Beira Agricultural structure, including irrigation, lower cost inputs, process- Growth Corridor), while others, such as the Lakaji ing and storage facilities, finance, and markets. Adjacent Corridor in Nigeria, are still in the design stage. 88 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure A.1: Example of an agribusiness cluster Source: SAGCOT Investment Blueprint, AgDevCo, and Prorustica. The aim of creating simultaneous coordinated The forces driving the evolution of the design and investments can also be found in the concept of growth development of these types of programs are the demands poles. Rather than being oriented around addressing of modern agribusiness and commercial agriculture for identified market failures, growth pole projects center on new technology, finance, and logistics. To ensure their exploiting opportunities that already exist. The underlying success, larger agricultural systems are needed, be assumption about the benefits of growth poles is that they stand-alone commercial farming and agribusiness they increase market size so that it becomes profitable enterprises or those linked to business focused, integrated for firms to invest, with the resulting higher wages and small-scale organizations. All of these agricultural systems economies of scale. Notable agriculture-related growth require viable and reliable power sources. The primary pole programs include those now being developed in power requirement of commercial agribusiness clusters is Burkina Faso (e.g., Bagre Growth Pole Project) and the irrigation, which can increase yields, reduce risk, and allow Democratic Republic of the Congo (e.g., Western Growth for winter cropping and post-harvest processing and stor- Poles Project). The Western Growth Poles Project also age activities; locating these activities closer to production includes development of a special economic zone to can reduce transport costs and allow for increased value provide land equipped with critical infrastructure and a capture closer to the point of production. more conducive business environment for investors and With a focus on particular regions for agribusiness private-sector operators. development in place, the aim of governments should be Annex A: Business Models for Agricultural Development89 to encourage anchor investments that require reliable related investments into the region to exploit the sources of power. Building up a critical mass of such ­ value-chain opportunities and economies of scale. These investments should lead to a trigger point, whereby activities, in turn, will lead to opportunities to electrify investments in grid extension and cluster electrification local businesses and community customers, whose low are financially and economically feasible. Reaching this levels of power consumption would not otherwise have tipping point will allow for the “crowding in” of additional justified electrification. endnote 1. Kilimo Kwanza Executive Committee, Investment Blueprint (Dar es Salaam: SAGCOT, 2011). Annex B: Agriculture Fuels for Power Generation I n addition to providing demand for power, certain agri- crushed produces nearly 3 MT of wet bagasse. The high culture activities provide a supply of power. Agricultural moisture content of bagasse, typically 40–50 percent, is products that may be used as fuels for power genera- detrimental to its use as a fuel. For electricity production, tion can be categorized as direct burning fuels or fuels it is stored wet, and the combination of the mild exother- that are the product of chemical conversions. This annex mic reaction resulting from the degradation of residual outlines three of the more common forms of power supply sugars, along with exposure to air, light, and heat, dries the from agricultural activities. bagasse pile slightly. Bagasse is used primarily as a fuel source for sugar mills. When burned in quantity, it produces sufficient heat Biomass energy to provide both electricity and heat (including steam) to supply all the needs of a typical sugar mill, with Biomass is biological material derived from living or energy to spare. At some sites, surplus electricity is sold to decaying organisms. In the context of biomass energy, the third parties (including feeding in to main grids). term often refers to plant-based material; however, bio- mass can apply equally to animal- and vegetable-derived material. As it is growing, biomass takes carbon out of the Biogas atmosphere, and returns it as it is burned. Biomass for energy can include a wide range of materials. High-value Anaerobic digestion is a natural process, whereby plant material, such as good quality large timber, is unlikely and animal materials (biomass) are broken down by to become available for energy applications. However, microorganisms in the absence of air. The process begins resources of residues and waste could potentially become when biomass is placed inside a sealed tank or digester. available, in quantity, at relatively low cost. In the con- Naturally occurring microorganisms digest the biomass, text of Sub-Saharan Africa, the main categories include which releases a methane-rich gas (biogas) that can be agricultural residues from harvesting and processing and used to generate renewable heat and power. The remain- high-yield crops grown specifically for energy applications. ing material (digestate) is rich in nutrients, so it can be Plant-based material includes wood (sawmill waste), nut- used as a fertilizer. shells, agricultural wastes (e.g., rice husks), corn stover, A biogas plant can be fed with such crops as maize and cassava peels. silage or biodegradable wastes, including sewage sludge An assessment for the West African Economic and (animal and human) and food waste. Monetary Union (UEOMA) countries suggests that agri- Four types of technology can be used to convert cultural residues amount to about 10 metric tons (MT) of the chemical energy found in biogas into electricity. In stubble per ha of maize, 5 MT of dry matter per ha of sor- biogas conversion, the chemical energy is converted into ghum, 4 MT of straw, 2.5 MT of bran per ha of rice, and mechanical energy in a controlled combustion system. 2 MT of tops per ha of groundnut and cowpea (UEMOA The mechanical energy activates a generator, producing 2008). In many countries, these are sources for tradi- electrical power. Gas turbines and internal combustion tional, as well as modern, utilization of biomass energy. engines are the most common technologies used in this type of energy conversion. Bagasse At the village level, biogas plants can be built to con- vert livestock manure into biogas and slurry, the fermented Bagasse—the fibrous matter that remains after sugar- manure. For small-scale farmers, the technology is feasible cane or sorghum stalks are crushed to extract their for those with livestock producing 50 kg of manure per juice—is used as a biofuel in many sugar estates around day, an equivalent of about 6 pigs or 3 cows. This manure is the world. In sugar production, every 10 MT of cane collected and mixed with water and fed into the plant. Annex C: Description of Processing Activities Post-Harvest and Primary Commercial-scale mills are usually found along main Processing roads with access to national grid power supplies. Diesel power supplies are too expensive for commercial operators Cleaning drying. Many of the basic drying techniques rely to remain competitive, and other sources of power can be on solar energy through sun drying (e.g., such cereals as unreliable. In many countries, a mill may have a backup wheat and maize). Slightly more rigorous drying technolo- diesel generator to compensate for the unreliability of gies use energy input for heating boilers; this energy may national grid supplies. be in the form of electricity, but often is biomass (farm Cold storage. Control temperature storage is used waste) or liquefied petroleum gas (LPG). The latter tech- to reduce the temperature of foods and flowers post-­ niques are more common for fruits, vegetables, and meats harvest. Cooling or chilling a food product reduces the risk with a high moisture content (i.e., about 60–80 percent) of bacterial growth and allows longer storage of produce which must be reduced to a range of 10–25 percent to without spoilage.1 In principle, this process enables farmers prevent spoilage. in relatively remote locations to harvest and store pro- Milling. Mills are used for processing in the value duce for shipment to large demand centers beyond the chains of maize, wheat, and rice. Smaller mills may be local markets (including exports). A cold chain is thus a powered with diesel or electricity, and larger units with necessary asset for many high-value agricultural products electricity only. For maize, the main choice of milling is (e.g., milk and dairy products, fish and other seafood, fruit either a plate mill or hammer mill (often supplied by India and vegetables, meat and prepared foods) and high-value and China, and increasingly from local craftsmen). The horticulture and floriculture industries, especially those plate mill can grind both wet and dry products, while the that are export-oriented. Large storage hubs are often hammer mill is restricted to dry products. Hammer mills centrally located at transportation centers; however, more are the more prevalent of the two although plate mills localized facilities are often necessary since products are popular in West Africa and Sudan and operate with a deteriorate quite rapidly post-harvest and must be cooled/ greater component of shear than compression. As a rule dried or processed immediately.2 While grid power is of thumb, about 1 kW can mill 25–30 kg of produce per more cost effective, alternative energy sources, including hour. Hammer mills have a power requirement in a range solar power, can be used.3 For commodities transported of 2–50 kW, while motor-driven plate mills generally fresh to market, cooling systems are often temporary or demand less power; 0.5–12 kW is usually sufficient. Larger movable, with commodities packed straight into refriger- scale hammer mills, with a capacity of 4.5–5 MT per hour, ated reefers before being moved within days. Reefers can have a power consumption of approximately 75 kW; for be plugged into any power supply for the short term, and, fully integrated milling systems, with a capacity range once in transit, are often powered with diesel gensets. of 2.5–25 MT per hour, power demand is 120–650 kW. Cassava processing. Roots and tubers (e.g., cassava, These systems can operate year-round, often at nearly potatoes, and yams) have high moisture content, which constant rates. makes them hard to store and bulky to transport. Cassava The power demand of wheat mills ranges from 20 kW is the most perishable of the roots and tubers and can for smaller units up to 600–700 kW for larger ones. deteriorate within a couple of days of harvesting. This Small-scale rice mills can remove the hard husk and polish implies that cassava is mostly sold in processed form, and the kernel. A full rice processing production line (exclud- processing facilities and machinery need to be located at ing the polisher), with a daily output of 20–30 MT, has a relatively short distances from the agricultural lands. The total power demand of approximately 38 kW, whereas a more important traditionally processed products include processing line with polishers requires 60–90 kW. dried chips, flours/starches, and gari. Most small-scale 92 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa chippers and graters are petrol driven, with capacities of that could adversely alter food properties or deactivate 1 MT per hour and a power drive of 3.5 hp, equivalent to enzyme action and optimize the retention of certain 2.6 kW. Large-scale cassava factories are usually located quality factors at minimum cost, including such processes in the vicinity of cassava farms. as pasteurization (e.g., of milk and some fruit juices) and Meat processing. The core processing equipment sterilization. Heat exchangers are used on a wide variety consists of hoists for lifting, which can be operated manu- of products, including pasteurization of cheese, milk, ally or electrically; meat grinders; bowel cutters; cooking and other beverages; ultrahigh temperature sterilization; vats; smokehouses; and chillers. Refrigeration is generally bottled water treatment; and heating of soups, sauces, the most energy-intensive activity in meat-processing and starches. facilities. Other uses of electricity include on-site water Canning, bottling, and packaging. A growing num- pumping for washing, electrical elevators, and hoists and ber of foods are packaged to increase their shelf life. stunning guns, with scalding tanks (electrical heating) Prior to packaging (or canning or bottling), food may be for pig processing. Modern abattoirs consume energy in processed (by juicing, peeling, or slicing) to increase value livestock holding; slaughtering and processing; monitoring and prevent deterioration (through pasteurization, boiling, and testing; cleaning; and packing. refrigeration, freezing, or drying). Each of these processes Oil extraction. Oil extraction from a variety of creates demand for electricity. Packing requires electric- oilseeds (e.g., sunflower, soybean, sesame, palm oil, and ity to run machines for vacuum sealing, heat sealing, and groundnut) results in significant value addition to the final bottling; in larger facilities, electricity is needed to power product. While smaller scale extraction is done using a conveyor belts, as well as to run filling, weighing, wrapping, manual press, larger scale commercial systems use motor- boxing, coding, and printing equipment. ized presses that rely on electric input. Oil filter presses Many of Sub-Saharan Africa’s canning and bottling are used for larger, electricity-powered oil-extraction factories are situated in areas where electric power is systems for sunflower, groundnut, and soybean. Once available and reliable.4 Modern packing lines require cleaned and de-hulled, the seed is placed under increas- reliable electricity supplies to operate efficiently. As with ing pressure as it is conveyed through a tapered chamber other secondary processing plants, packaging plants are (expelling). Mini extruders, typically with a capacity of often supplied with main grid power. The power require- 125 kg per hour, require a power drive of about 10 kW, ments for juicing and canning is quite low. For example, while 400 kg per hr power requirements are approxi- a juicing machine that can process up to 5 MT of raw mately 23 kW. Capacity depends on the quality and type fruit per hour may have a peak power load of 5–22 kW. of seed (e.g., groundnut capacity is 120–180 kg per hour, A canning machine with a per-hour capacity of 250 compared to sunflower capacity of 280–320 kg per hour cans (approximately 125 kg) has a power-load range of using a similar 15–18.5 kW motor). 5.5–7.5 kW.5 Given the scale efficiencies of larger facili- ties, it is difficult to extrapolate to determine the load of a much larger commercial plant without information on the Secondary Processing capacity and power demand. Thermal treating. Thermal treating of foods (either heating or cooling) is necessary to destroy microorganisms endnotes 1. Rapid chilling—also known as flash freezing—lowers this risk even further. 2. For some products, the shelf life may be diminished by a factor of eight times the length of delay between harvesting and cooling. 3. With peak demand during daylight hours matching the generation profile of solar power, freezing systems can be switched off overnight when outside temperatures are cooler. 4. Notable canned foods prevalent in Sub-Saharan Africa include pineapple, grapefruit, and tomato. 5. References come from data on plants available for sale on Alibaba. Annex D: Maps of Case Study Project Areas Map D.1: Tanzania: Power and Agriculture in the Sumbawanga Agriculture Cluster 94 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.2: Tanzania: Mwenga Mini-Hydro Mini-Grid Annex D: Maps of Case Study Project Areas95 Map D.3: Zambia: Mkushi Farming Block 96 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.4: Zambia: Mwomboshi Irrigation Development and Support Project Annex D: Maps of Case Study Project Areas97 Map D.5: Kenya: Oserian Flowers and Harnessing Geothermal Power 98 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.6: Kenya Tea Development Agency Holdings Mini-Hydro Mini-Grids Annex D: Maps of Case Study Project Areas99 Map D.7: Ethiopia: Sugar Estates 100 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.8: Mali: Power Network and Agricultural Districts References ACET (African Center for Economic Transformation). Collier, P., and S. Dercon. 2009. “Africa Agriculture in 2013. The Soybean Agri-Processing Opportunity in Africa. 50 Years: Smallholders in a Rapidly Changing World?” Accra, Ghana and Washington, DC: African Center for Expert Meeting on How to Feed the World in 2050. Economic Transformation. Rome: Food and Agriculture Organization of the United Nations (FAO). Alexandratos, N., and J. Bruinsma. 2012. World Agriculture: Towards 2030/2050. ESA Working Paper da Silva, C., D. Baker, A. Shepherd, C. Jenane, No. 12-03. Rome: Food and Agriculture Organization of and S. Miranda-da-Cruz. 2009. Agro-Industries for the United Nations (FAO). Development. Rome: Food and Agriculture Organization of the United Nations (FAO) and United Nations Banerjee, S., Z. Romo, and G. McMahon et al. 2015. The Industrial Development Organization (UNIDO). Power of the Mine: A Transformative Opportunity for Sub- Saharan Africa. Directions in Development; Energy and Deininger, K., and D. Byerlee. 2011. Rising Global Interest Mining. Washington, DC: World Bank Group. in Farmland: Can It Yield Sustainable and Equitable Benefits? Agriculture and Rural Development. Washington, DC: Barnes, D. (ed.). 2007. The Challenge of Rural World Bank. Electrification: Strategies for Developing Countries. Washington, DC: RFF Press. Derks, E., and F. Lusby. 2006. Mali Shea Kernel: Value Chain Case Study. MicroREPORT #50. Washington, DC: ———. 2014. Electric Power for Rural Growth: How United States Agency for International Development Electricity Affects Life in Developing Countries. Second (USAID). edition. Washington, DC: Energy for Development. Development Initiatives. 2015. “Aid to the Agricultural Barnes, D., H. Peskin, and K. Fitzgerald. 2003. “The Sector in Sub-Saharan Africa Doubles, 2003–12” (http:// Benefits of Rural Electrification in India: Implications for devinit.org/#!/post/aid-agricultural-sector-sub-saharan- Education, Household Lighting, and Irrigation.” Draft africa-doubles-2003-12). paper prepared for South Asia Energy and Infrastructure. World Bank Group, Washington, DC. Diao, X., J. Thurlow, S. Benin, and S. Fan (eds). 2012. Strategies and Priorities for African Agriculture: CAADP (Comprehensive Africa Agriculture Economywide Perspectives from Country Studies. Development Programme). 2012. “About CAADP” Washington, DC: International Food Policy Research (http://www.nepad-caadp.net/about-us. Accessed Nov. 7, Institute (IFPRI). 2012). ECA (Economic Consulting Associates) and Prorustica. Cervigni, R., R. Liden, J. Neumann, and K. Strzepek. 2015. Power and Agriculture in Africa—Landscape analysis. 2015. Enhancing the Climate Resilience of Africa’s London: Economic Consulting Associates Limited. Infrastructure: The Power and Water Sectors. Africa Development Forum. Washington, DC: World Bank. FAO (Food and Agriculture Organization of the United Nations). 2005. Irrigation in Africa in Figures—AQUASTAT Chu, J. 2013. Creating a Zambian Breadbasket: “Land Survey. FAO Water Reports. FAO Land and Water grabs” and Foreign Investments in Agriculture in Mkushi Development Division. Rome: Food and Agriculture District. Land Deal Politics Initiative (LDPI) Working Organization of the United Nations (FAO). Paper 33. Brighton, UK: Institute of Development Studies. 102 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa ———. 2009. Agribusiness Handbook—Sugar Beet White Rome: International Fund for Agricultural Development Sugar. FAO Investment Centre Division. Rome: Food and (IFAD). Agriculture Organization of the United Nations (FAO). Marshall, A. 1890. Principles of Economics. London: ———. 2014. Crop Prospects and Food Situation. No 1. Macmillan and Co., Ltd. Global Information and Early Warning System on Food Porter, M. 1990. “The Competitive Advantage of and Agriculture. Trade and Markets Division. Rome: Food Nations.” Harvard Business Review, March–April. and Agriculture Organization of the United Nations (FAO). Poulton, C., G. Tyler, P. Hazell, A. Dorward, J. Kydd, and M. Stockbridge. 2008. “All-Africa Review of Experiences Giordano, M., C. de Fraiture, E. Weight, and J. van der with Commercial Agriculture: Lessons from Success Bliek (eds.). 2012. Water for Wealth and Food Security: and Failure.” Background paper for the Competitive Supporting Farmer Driven Investments in Agricultural Water Commercial Agriculture in Africa (CCAA) Study. World Management. Synthesis Report of the AgWater Solutions Bank Group (WBG) and Investment Centre of the United Project. Colombo, Sri Lanka: International Water Nations Food and Agriculture Organization (FAO). Management Institute (IWMI). Schaffnit-Chatterjee, C. 2014. Agricultural Value Hazel, P., C. Poulton, S. Wiggins, and A. Dorward. 2007. Chains in Sub-Saharan Africa. Frankfurt: Deutsche Bank The Future of Small Farms for Poverty Reduction and Research. Growth. 2020 Discussion Paper No. 42. Washington, DC: International Food Policy Research Institute (IFPRI). Sebastian, K. 2014. Atlas of African Agriculture Research and Development: Revealing Agriculture’s Place in Africa. Hosier R., et al. forthcoming. Setting the Scene for Washington, DC: International Food Policy Research Regional Dialogue: Southern Africa Energy Water Nexus Institute (IFPRI). Issues. Washington, DC: World Bank Group. Seck P., A. Touré, J. Coulibaly, A. Diagne, and Hussain, M., K. Malik, J. Kapika, and C. Etienne. forth- M. Wopereis. 2013. Africa’s Rice Economy before and after coming. Southern Africa Energy Water Nexus Issues. the 2008 Crisis. Cotonou, Benin: Africa Rice Center. Washington, DC: World Bank Group. Staatz, J. 2011. “Enhancing Agricultural Productivity.” In IEA (International Energy Agency) and World Bank. Agribusiness for Africa’s Prosperity, edited by Kandeh H. 2015. Sustainable Energy for All 2015—Progress Toward Yumkella, Patrick M. Kormawa, Torben M. Roepstorff, and Sustainable Energy (June). Global Tracking Framework Anthony M. Hawkins, 58–86. Vienna: United Nations Report. Washington, DC: World Bank (http:// Industrial Development Organization (UNIDO). trackingenergy4all.worldbank.org/~media/ GIAWB/GTF/ Documents/GTF2105-Full-Report.pdf). Standard Bank Research. 2014. “Rise of the Middle Class in Sub-Saharan Africa” (http://www.blog.standardbank International Trade Center. 2014. “Kenya’s Cut Flower .com/node/61428). Export to Reach USD 1 Billion” (http://www.intracen.org). UEMOA (West African Economic and Monetary Union). Ivanic, M., and W. Martin. 2014. Short- and Long-Run 2008. Sustainable Bioenergy Development in UEMOA Impacts of Food Price Changes on Poverty. Policy Research Member Countries. Working Paper No. WPS 7011. Washington, DC: World Bank Group. United Nations. 2013. World Population Ageing 2013. Department of Economic and Social Affairs, Population Korwama, P. 2011. “Agribusiness: Africa’s Way Out of Division. New York: United Nations. Poverty.” Making It: Industry for Development, June 15. United Nations Industrial Development Organization World Bank. 2008. World Development Report 2008: (UNIDO). Agriculture for Development. Washington, DC: World Bank. Krugman, P. 1991. Geography and Trade. Cambridge, MA: MIT Press. ———. 2009. Awakening Africa’s Sleeping Giant: Prospects for Commercial Agriculture in the Guinea Savannah Zone Livingston, G., S. Schonberger, and S. Delaney. 2011. Sub- and Beyond. Directions in Development; Agriculture and Saharan Africa: The State of Smallholders in Agriculture. Rural Development. Washington, DC: World Bank. References103 ———. 2011a. The World Bank Group Framework and WBG (World Bank Group). 2015. Enabling the Business IFC Strategy for Engagement in the Palm Oil Sector. of Agriculture: 2015. Progress Report. Washington, DC: Washington, DC: World Bank. World Bank. ———. 2011b. Irrigation Development and Support Project ———. 2016. Enabling the Business of Agriculture 2016: (P102459) Project Appraisal Document. Washington, DC: Comparing Regulatory Good Practices. Washington, DC: World Bank. World Bank. ———. 2011c. Leveraging Investments by Natural Resource You, L. 2008. “Irrigation Investment Needs in Concessionaires. Infrastructure Policy Notes. Washington, Sub-Saharan Africa.” Background Paper 9, Africa DC: World Bank. Infrastructure Country Diagnostic. World Bank Group, Washington, DC. ———. 2013. Growing Africa: Unlocking the Potential of Agribusiness. Washington, DC: World Bank. You L., C. Ringler, G. Nelson, U. Wood-Sichra, R. Robertson, S. Wood, G. Zhe, T. Zhu, and Y. Sun. ———. 2015. Africa’s Pulse (October 2015). Washington, 2009. “Torrents and Trickles: Irrigation Spending DC: World Bank Group. Needs in Africa.” Background Paper 9 (Phase II), Africa ———. 2016a. High and Dry: Climate Change, Water, and Infrastructure Country Diagnostic. World Bank Group, the Economy. Washington, DC: World Bank. Washington, DC. ———. 2016b. Energizing Agriculture: Enhancing Efficiency in Agriculture. Washington, DC: World Bank. The majority of households and enterprises in rural Africa cope without electricity, compromising socio-economic welfare and firm productivity. Africa, characterized by low electricity consumption and ability to pay, makes rural electrification commercially unviable. Agriculture as the most important value added industry in rural areas presents a significant opportunity to improve commercial viability of grid and off- grid projects. This study explores the nexus between power and agriculture, challenges in scaling-up, and recommendations to harness this opportunity. Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Sudeshna Ghosh Banerjee Kabir Malik Andrew Tipping Juliette Besnard and John Nash Table of Contents Foreword.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Abbreviations and Acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Chapter 1. Agriculture and Power Nexus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 High Potential for Agricultural Transformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Past Performance: A Missed Opportunity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Investment Funding Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 An Improving Outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Major Approaches to Agricultural Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Cluster Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Smallholder Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Agricultural Growth to Raise Rural Welfare: Reasons for Optimism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Rural Electrification Has Lagged Behind. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Agriculture as an Anchor Load for Rural Electrification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Study Purpose and Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Chapter 2. Power Needs of Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Power Needs across the Agriculture Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Irrigation Potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Primary and Secondary Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Aggregate Electricity Demand from Irrigation and Processing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Chapter 3. Power Needs in Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Selection of Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Electricity Demand and Farming Scale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Electricity Demand in the Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 iv Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Chapter 4. Lessons from Ongoing Power-Agriculture Integration Projects. . . . . . . . . . . . . . . . . . . . . . . . 38 Case Study 1. Tanzania: Sumbawanga Agriculture Cluster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Financial Viability: Extension of Main Grid from Mbeya to Sumbawanga and Rukwa. . . . . . . . . . . 42 Economic Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Case Study 2. Tanzania: Mwenga Mini-Hydro Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Power Supply Options, Commercial Arrangements, and Financial Analysis. . . . . . . . . . . . . . . . . . . . 45 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Case Study 3. Zambia: Mkushi Farming Block.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Case Study 4. Zambia: Mwomboshi Irrigation Development and Support Project. . . . . . . . . . . . . . . . . 51 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Case Study 5. Kenya: Oserian Flowers and Geothermal Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Financial Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Case Study 6. Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids. . . . . . . . . . . . . . . . 58 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Power Supply Options.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Key Conclusions from the Case Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Large Power Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Supply Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Table of Contents v Financial and Economic Viability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Financing of Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Chapter 5. Opportunities to Harness Agriculture Load for Rural Electrification. . . . . . . . . . . . . . . . . . 64 Simulation of Power Demand in a Stylized Agricultural Setting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Simulation Study 1. Ethiopia: Power Generation from Sugar Estates.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Power Demand.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Simulation Study 2. Mali: Mini-Grid Expansion for Productive Users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Power Demand from Mini-Grids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Power Supply Options and Commercial Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Financial Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Economic Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Main Inferences and Institutional Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Chapter 6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Key Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Overall Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Case Study Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Recommended Actions to Promote Power-Agriculture Integration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Improve Institutional Coordination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Integrate Planning of Power, Agriculture, and Rural Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Promote Farmers’ Productivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Annexes A: Business Models for Agricultural Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 B: Agriculture Fuels for Power Generation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 C: Description of Processing Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 D: Maps of Case Study Project Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 vi Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Tables ES.1 Summary of Ongoing or Planned Cases of Power-Agriculture Integration. . . . . . . . . . . . . . . . . xvii ES.2 Summary of Simulated Cases of Power-Agriculture Integration.. . . . . . . . . . . . . . . . . . . . . . . . . . . . xix 2.1 Power Demand for Irrigation, by System Type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.2 Potential Investment Needs for Large-Scale, Dam-Based and Complementary Small-Scale Irrigation Schemes in Sub-Saharan Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3 Key Power-Intensive Agribusiness Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.4 Method for Calculating Power Demand from Irrigation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.5 Power Demand for Crop Processing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.1 Analysis of Commodity Value Chains, by Scale and Region/Country. . . . . . . . . . . . . . . . . . . . . . . 24 3.2 Comparison of Historical and Projected Commodity Growth Rates and Estimated Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.3 Countries in Sub-Saharan Africa with Similar Commodity Production and Processing Systems.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.4 Power Demand for Standard 300 ha Cultivated Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.1 Sumbawanga Agriculture Cluster at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.2 Sumbawanga Geographic and Demographic Features.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3 Total Power Demand from Agriculture by 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.4 Residential and Commercial Data to Calculate Commercial Power Demand. . . . . . . . . . . . . . . . 41 4.5 Estimated Capital and Operating Costs for Transmission and Distribution Expansion.. . . . . . . 43 4.6 Estimated Power Consumption and Transmission and Distribution Tariff Requirement. . . . . . 43 4.7 Financial Present Value of Grid Extension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.8 Economic Costs and Benefits of Sumbawanga Grid Extension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.9 Mwenga Mini-Hydro Mini-Grid at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.10 Estimated Power Demand from Mwenga Mini-Hydro Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.11 Economic Costs and Benefits of Mwenga Mini-Hydro Plant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 4.12 Mkushi Farming Block at a Glance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.13 Power Requirements for Irrigation and Milling in the Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.14 Electrification Rates and Power Load of Households in Mkushi Farm Block. . . . . . . . . . . . . . . . . 48 4.15 Financial Analysis of Mkushi Farming Block from the Perspective of the Utility and a Representative Farmer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.16 Net Social Benefits of Grid Extension, Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.17 Economic Costs and Benefits of Grid Extension, Mkushi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.18 Mwomboshi Irrigation Development and Support Project at a Glance.. . . . . . . . . . . . . . . . . . . . . . 51 4.19 Irrigation Power Requirements in Mwomboshi, Zambia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Table of Contents vii 4.20 Milling Power Requirements in Mwomboshi, Zambia.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.21 Financial Analysis, Mwomboshi.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.22 Economic Costs and Benefits of the IDSP Project, Mwomboshi. . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.23 Oserian Flowers and Geothermal Power Project at a Glance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.24 Financial Analysis, ODCL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 4.25 Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids at a Glance. . . . . . . . . . . . 58 4.26 Typical LCOE Values for Small-Scale Generation and Distribution Systems.. . . . . . . . . . . . . . . . 62 5.1 Assumptions for Typical Area/Agricultural Activity/Power Demand Model.. . . . . . . . . . . . . . . . . 65 5.2 Ethiopia: Power Generation from Sugar Estates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.3 Total Power Demand from Agriculture and Residential/Commercial Loads. . . . . . . . . . . . . . . . . . 67 5.4 Sugar Factory Power Generation in Years 1 and 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.5 Net Power Generation from Sugar Factory by Year 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.6 Capital Cost Assumptions for Grid Connection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.7 Financial Analysis from the Utility’s Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.8 Sugar Estate Capital Costs, Assumptions for Production Costs, and Revenues.. . . . . . . . . . . . . 69 5.9 Net Economic Benefits of Grid Extension to the Sugar Estate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 5.10 Economic Net Present Value of Extending the Grid to the Sugar Estate. . . . . . . . . . . . . . . . . . . . . 71 5.11 Mali Mini-Grid Expansion for Productive Users at a Glance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.12 Potential Addition of Small Agro-Industrial Activities and Other Businesses. . . . . . . . . . . . . . . . 74 5.13 Current Financial Situation of Koury Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.14 Financial Analysis of Capacity Expansion of Koury Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Figures ES.1 Energy Intensive Activities across Agriculture Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv ES.2 Estimated Power Demand from Agriculture in 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv 1.1 Historical Performance in Agriculture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Land and Water Resources Potential in Sub-Saharan Africa.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 B1.3.1 Changes in Irrigation Revenues from Climate Change, 2015–50 (present value). . . . . . . . . . . . . 8 1.3 Projected Value of Food Markets in Sub-Saharan Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.4 Electricity as a Constraint to Food-sector Development in Sub-Saharan Africa. . . . . . . . . . . . . . 9 1.5 Electrification Rate, by Developing Region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.1 Power Needs across Agriculture Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2 Potential New or Rehabilitated Irrigable Land in Sub-Saharan Africa.. . . . . . . . . . . . . . . . . . . . . . . 18 2.3 Estimated Electricity Demand (MW) from Agriculture for Sub-Saharan Africa in 2030. . . . 21 3.1 Potential Peak Capacity and Energy Demand for Large- and Small-scale Systems. . . . . . . . . . 29 viii Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa 3.2a Electricity Input in the Maize Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2b Electricity Input in the Rice Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2c Electricity Input in the Cassava Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2d Electricity Input in the Wheat Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2e Electricity Input in the Soybean Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2f Electricity Input in the Pineapple Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2g Electricity Input in the Sugarcane Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2h Electricity Input in the Oil Palm Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2i Electricity Input in the Dairy Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.2j Electricity Input in the Poultry Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.2k Electricity Input in the Tea Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.2l Electricity Input in the Floriculture (roses) Value Chain.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.2m Electricity Input in the Cotton (lint) Value Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.3 Potential Power Demand in 2030 from Processing for Small-scale Agriculture, by Selected Value Chains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.1 Estimated Peak Load and Energy Demand, by Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2 Estimated Volume of Crops That May Utilize Electricity for Processing. . . . . . . . . . . . . . . . . . . . . 41 4.3 Comparative Cost of Power Supply Options in Sumbawanga. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.4 Total Peak Load in Mkushi, 1995–2014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.5 Power Demand from Irrigation and Milling in Mkushi, 1995–2014. . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.6 Mwomboshi IDSP Plot Sites Developed for Small-scale Farmers. . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.7 Mwomboshi Peak Load and Power Consumption Forecast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.8 Residential and Commercial Demand, Electrification Rate 2016–2031. . . . . . . . . . . . . . . . . . . . . 53 4.9 Power Uses and Sources at ODCL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.10 Output of ODCL’s Power Plants and Expected Increased Output. . . . . . . . . . . . . . . . . . . . . . . . . . 57 4.11 Electricity Output of Capacity Expansion Project and Intended Uses. . . . . . . . . . . . . . . . . . . . . . . 57 4.12 KTDA’s North Mathioya Hydropower Project: Financial Benefits and Power Sold. . . . . . . . . . 59 5.1 Power Demand and Breakdown for a Given Area Radius. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.2 Sensitivity of Power Load to Changes in Percent of Commercial Irrigation. . . . . . . . . . . . . . . . . . 66 5.3 Estimated Energy Demand and Peak Load, by Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.4 Net Social Benefits of Grid Extension to Sugar Estate (years 1–20).. . . . . . . . . . . . . . . . . . . . . . . . . 71 5.5 Koury Mini-grid: Electricity Consumption Patterns.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.6 Energy Generation Profile at Koury Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.7 Koury Mini-grid Profile: Additional Commercial and Industrial Loads. . . . . . . . . . . . . . . . . . . . . . . 75 5.8 Operating Expense and Capital Expenditure Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 A.1 Example of an Agribusiness Cluster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Table of Contents ix Boxes 1.1 Terminology Clarification: Agriculture and Agribusiness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Africa’s Vision for Agriculture: CAADP Goals.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Making Africa’s Power and Water Infrastructure Climate Resilient. . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.1 Farm Type Definitions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.2 Palm Oil and Power Integration in Uganda.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 5.1 Isolated Mini-Grid Systems in Mali: Existing and Potential Power Demand.. . . . . . . . . . . . . . . . . 72 5.2 Large-Scale Opportunities for Power-Agriculture Integration in Mali. . . . . . . . . . . . . . . . . . . . . . . 77 Maps D.1 Tanzania: Power and Agriculture in the Sumbawanga Agriculture Cluster.. . . . . . . . . . . . . . . . . . . 93 D.2 Tanzania: Mwenga Mini-Hydro Mini-Grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 D.3 Zambia: Mkushi Farming Block. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 D.4 Zambia: Mwomboshi Irrigation Development and Support Project. . . . . . . . . . . . . . . . . . . . . . . . . . 96 D.5 Kenya: Oserian Flowers and Harnessing Geothermal Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 D.6 Kenya Tea Development Agency Holdings Mini-Hydro Mini-Grids. . . . . . . . . . . . . . . . . . . . . . . . . 98 D.7 Ethiopia: Sugar Estates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 D.8 Mali: Power Network and Agricultural Districts.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Foreword The greatest challenge to increasing electricity access in Sub-Saharan Africa is how to make electricity provision financially viable in low-demand rural households. The presence of commercially attractive customers—typically those that have relatively large and stable electricity demand for revenue generating purposes—can help reduce the barriers to accelerating grid and off-grid approaches to rural electrification. By aggregating anchor-load demand with that of households and businesses, it may be possible to extend the grid or create opportunities for mini-grids and other decentralized options. African agriculture has tremendous potential to raise rural welfare through agricultural transformation. It is estimated that productivity growth in agriculture—which predominates the livelihoods of the subconti- nent’s rural poor—could be several times more effective than growth in other sectors in reducing rural poverty. Furthermore, there is a growing commitment among African governments toward sustainable and inclusive agricultural development. Developing energy intensive agricultural processes, such as large-scale irrigation or milling activities, can not only increase agricultural productivity, but can also increase the commercial viability of electricity provi- sion. The large-farm, agribusiness model practiced over the past 20 years has a continuing strategic role to play in promoting growth in Africa. At the same time, subsistence, smallholder farms, which account for most of Sub-Saharan Africa’s agriculture, are key to stimulating the rural economy and uplifting the poor. Energy, along with investments in other complementary infrastructure and services (e.g., roads, transport links to markets, and access to finance), is a critical input for supporting Africa’s agricultural transformation. Without access to affordable and reliable electricity, farmers will continue to face constraints to productivity growth and thus lag behind their counterparts in more prosperous regions of the developing world. Against this backdrop, this study explores opportunities for synergy between the goals of rural elec- trification and agricultural transformation in Sub-Saharan Africa. It shows that leveraging complementary investments in agriculture and electricity can yield double dividends in terms of poverty alleviation. Aligning electricity investments with agricultural development can maximize joint benefits from the expansion of rural electricity access and the increase in value added along the agricultural value chains, both of which are directly linked to improved quality of life and poverty alleviation in rural communities. Lucio Monari Ethel Sennhauser Director Director Energy and Extractives Global Practice Agriculture Global Practice The World Bank The World Bank Acknowledgments The core team for this study included Sudeshna Ghosh Banerjee, Kabir Malik, Juliette Besnard, and John Nash. The team benefited from the background report prepared by Economic Consulting Associates (ECA) and Prorustica, a consulting consortium led by Andrew Tipping and Peter Robinson. The team wishes to thank expert consultants Douglas Barnes and Subodh Mathur, who provided valuable inputs at various stages of the study. The team is grateful to Olivier Dubois and Alessandro Flammini from the Food and Agriculture Organization (FAO) for their inputs and review. The team is appreciative of the overall guidance provided by Lucio Monari and Meike van Ginneken, man- agers in the World Bank’s Africa Energy Group. The team also wishes to thank peer reviewers Vivien Foster, Malcolm Cosgrove-Davies, Dana Rysankova, Holger Kray, Melissa Williams, and Katie Kennedy Freeman for their valuable advice and constructive inputs. The team thanks Norma Adams for editing the report. Finally, the team gratefully acknowledges the funding provided by the Africa Renewable Energy Access (AFREA) program and the Energy Sector Management Assistance Program (ESMAP). Abbreviations and Acronyms ABC Anchor Business Community AMADER Malian Agency for Development of Household Energy and Rural Electrification CAADP Comprehensive Africa Agriculture Development Programme CAGR compound annual growth rate CHP combined heat and power CSR corporate social responsibility CTC cutting, tearing, and curling DRC Democratic Republic of the Congo ECA Economic Consulting Associates EEPCO Ethiopian Electric Power Corporation EFB empty fruit bunch ESC Ethiopian Sugar Corporation EWURA Energy and Water Utilities Regulatory Authority FFB fresh fruit bunch FiT feed-in tariff GDP gross domestic product GP global practice GTAP Global Trade Analysis Project IPP independent power producer IRR internal rate of return KPLC Kenya Power and Lighting Company KTDA Kenya Tea Development Agency LCOE levelized cost of electricity MSMEs micro-, small-, and medium-sized enterprises NPV net present value ODA official development assistance PV photovoltaic REA Rural Energy Agency (Tanzania) RVE Rift Valley Energy SAGCOT Southern Agricultural Growth Corridor of Tanzania SDG Sustainable Development Goal SE4ALL Sustainable Energy for All SHS solar home system SMEs small- and medium-sized enterprises SSA Sub-Saharan Africa TANESCO Tanzania Electric Supply Company Limited ZESCO Zambia Electricity Supply Corporation Executive Summary Increasing access to modern electricity services in and commercial power demand could increase the Sub-Saharan Africa is one of the main develop- feasibility of extending the grid or creating opportu- ment challenges facing the world over the next two nities for independent power producers and mini-grid decades. Inclusion of the target to “ensure access to operators. Drawing on a suite of case studies, this affordable, reliable, sustainable, and modern energy study offers insights on what it would take to opera- for all” in the Sustainable Development Goals (goal 7) tionalize the opportunities and address the challenges has brought a sharper focus to accelerating electric- for power-agriculture integration in Africa. ity access in the historically underserved regions of the world—most notably Sub-Saharan Africa. Two out of every three people in Sub-Saharan Africa live What is the scale of opportunity without electricity, a reality that is inconsistent with of power demand from the modern world. The majority of this population agriculture? without access to electricity is rural and poor. Rural electrification efforts in the region have not achieved Historical performance of agriculture in Sub- sufficient progress in increasing electricity access as Saharan Africa has been wanting. The share of these areas are typically commercially unattractive, agriculture in GDP has declined from 20 percent in characterized by sparsely distributed customers 2000 to 14 percent in 2013.1 A very small percent- with low electricity consumption and ability to pay, age of Africa’s agricultural production undergoes and a high cost of service to extend the grid. Rural industrial processing.2 In high-income countries, enterprises and households thus must cope without processing adds about US$180 of value per ton of electricity, relying instead on expensive, poor quality agricultural produce, compared to only $40 in Sub- backups (e.g., diesel, kerosene or other oil-based Saharan Africa; this disparity is aligned with the small sources), thereby stunting productivity, limiting size of Sub-Saharan Africa’s agribusiness sector rela- development outcomes, and imposing harmful tive to on-farm agriculture. In addition, for more than environmental impacts. The rural economies are four decades, the region’s share in global agricultural overwhelmingly dependent on agriculture; in fact, export markets has been on the decline. agriculture and agribusiness comprise nearly half of There are reasons to believe that agriculture Africa’s gross domestic product (GDP). These enter- productivity could turn the tide. Trends in economic prises require electricity to grow to their potential, growth and urbanization fuel the demand for food, while the expansion of rural energy services needs as do continuing improvements in infrastructure and consumers with consistent power needs to serve as a the benefits of lower oil prices. The potential urban reliable revenue source. market for agricultural goods and commodities is Can agriculture and energy come together in projected to reach US$1 trillion by 2030. There are Sub-Saharan Africa to offer a double dividend with a number of underlying structural incentives to pro- benefits to enterprises, households, utilities, and mote agriculture. The region has 45 percent of the private-sector service providers? This is the central world’s total suitable land area for expanding sustain- question of this study. That is, can energy intensive able agricultural production. Past gains in commercial activities along the agriculture value chains pro- crops (e.g., cashews, tea, and sesame seeds) indicate vide significant revenues to the power utilities and that the region can increase its agricultural pro- increase the financial viability of rural electrification? ductivity. But seizing this opportunity will require Combining agricultural load with other household farmers and agribusinesses to ramp up production xiv Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure ES.1: Energy intensive activities Figure ES.2: Estimated power demand across agriculture value chains from agriculture in 2030 8,000 Post-harvest & 7,000 6,915 Secondary On-farm primary processing processing 6,000 5,000 MW • Irrigation • Milling, drying, • Packaging, chilling, etc. bottling, etc. 4,000 3,786 3,000 Rural Urban/peri-urban 2,084 2,000 978 1,000 and develop agriculture value chains to enhance processing, logistics, market infrastructure, and retail 0 networks. Irrigation Processing (milling) Electricity is an important enabler for the 2015 2030 agriculture sector to realize its growth potential, especially for power intensive value chains. The need for electricity is distributed across the life of the today, to about 9 GW. The estimated incremental crop—from mechanized irrigation to processing for demand between 2015 and 2030 is 4.2 GW (fig- final consumption (figure ES.1). The power demand ure ES.2). Irrigation would provide about 75 percent for irrigation primarily comes from (i) sourcing bulk of agriculture’s demand, with the rest coming from water from a water body (e.g., a dam or river) and agro-processing. The irrigation demand estimate (ii) distributing it over the cultivated area. Bulk water assumes full exploitation of economically viable, pumping is typically the major source of demand and potential areas for new or rehabilitated irrigation depends on the vertical and horizontal distances of development, totaling nearly 6.8 million ha. This the scheme from the water source. Demand from would be dominated by small-scale scheme devel- distribution systems varies by the types of irrigation opment in the Gulf of Guinea and rehabilitation of system, which range in scale from manual to surface existing schemes in the Sudano-Sahelian region. flooding and localized ones to center pivots. Post- The agro-processing demand estimate is based on harvest and primary processing (e.g., milling and the electricity requirement for a typical processing drying) and secondary processing (e.g., packaging activity (milling), and thus does not capture demand and bottling) represent a growth area. It is clear that from the potential development of other processing milling is likely to increase significantly owing to the activities or storage. expected demand growth for such grains as maize, For 13 major agriculture value chains, electricity wheat, and rice. Similarly, such staples as cassava demand could increase by 2 GW (from 3.9 GW in are expected to experience increased demand for 2013 to 6 GW in 2030). This represents nearly half processing due to their perishable nature and use of the 4.2 GW of potential increase in electricity as an industrial input in the manufacture of other demand from agriculture calculated for Sub-Saharan products (e.g., glue in the case of cassava). Creating Africa. The 13 products are maize, rice, cassava, opportunities to piggyback viable rural electrification wheat, oilseed (soybean), horticulture (pineapple), onto local agricultural development will depend on sugarcane, oil palm, dairy, poultry, tea, floriculture the scale and profitability of agricultural operations, (roses), and cotton (lint). These were selected on crops, terrain, types of processing activity, and other the basis of their nature and magnitude of power site-specific local conditions. use for irrigation and processing, growth potential, By 2030, the region’s electricity demand from and ability to serve as significant loads for electricity agriculture is estimated to double from its level systems. Of the value chains studied, the per-hectare Executive Summary xv electricity demand is largest for poultry, because hybrid mini-grid (diesel-solar PV) to serve productive the process is more intensive, using less land for a users (tables ES.1 and ES.2). much larger yield. Other value chains with significant Irrigation is typically the largest source of power per-hectare demand are floriculture (roses), tea, and demand, along with processing activities in specific sugarcane. Together, maize, rice, and cassava add to instances. Irrigation usually has a larger load require- about 83 percent of the total incremental demand in ment than agro-processing activities, especially agriculture processing to 2030. For the 13 commod- in cases of supply to a given area (e.g., Tanzania’s ities analyzed, commercial-scale irrigated farming is Mufindi Tea Estate). Irrigation development and elec- the largest source of electricity demand. Commercial trification can significantly help increase the viability irrigated agriculture, which is highly mechanized, has of rural electrification. Taken alone, the smaller loads the largest potential for developing large power loads of agro-processing activities (e.g., milling and extru- across a range of farm sizes. sion) may not be sufficient to justify rural electrifica- tion investments, except when they provide a viable source of electricity generation (e.g., sugar) or have What are the case study lessons a large and consistent load requirement (e.g., tea). on economic and financial If the volumes of produce can benefit from powered viability? irrigation, supplemented by economies of scale, the load from the production could be significantly larger. This study analyzes eight case studies—six actual Irrigation and processing are often linked. In and two simulated—in five countries of Sub-Saharan many instances, increase in yields from irrigation is Africa; these provide important lessons on the ben- an important prerequisite for the development of efits and risks of large power loads, supply options, large-scale processing activities (as seen in Zambia). and viability. In Tanzania, the first case study is the This cause-and-effect relationship between irriga- Sumbawanga Agriculture Cluster, a concept-stage tion and processing was also observed in the cluster project located in the country’s Southern Agricultural concept (e.g., SAGCOT in Tanzania). Increase in the Growth Corridor of Tanzania (SAGCOT). The scale of processing activity could lead to a significant second case in Tanzania is the successful Mwenga increase in the power demand. Mini-Hydro Mini-Grid Project, which supplies the Successful integration of agriculture and power Mufindi tea estate and surrounding households in system development requires physical and market the country’s Southern Highlands. In Zambia, the infrastructure, which facilitate market access for first case is a grid extension to the ongoing Mkushi inputs and produce. Viable rural electrification relies Farming Block Project, stretching over 176,000 ha on a healthy and profitable agriculture sector. Better of land in the country’s Central Province. The infrastructure and market access improve agriculture second case study in Zambia is the Mwomboshi revenues, spurring further expansion in produc- Irrigation Development and Support Project (IDSP), tion and associated electricity demand. In Zambia, which is developing integrated irrigation agriculture for example, the strategic location of the Mkushi based around a recently built water storage dam farming block along a major international highway on the Mwomboshi River. In Kenya, the first case (T2 Highway and Tazara Railway, which connects examines floriculture development by the Oserian Lusaka and the Copperbelt in Zambia to the port at Development Company Limited (ODCL), a pioneer Dar es Salaam) has improved its development viabil- in using heat from geothermal wells for internal ity. The location of the farming block allows access power generation and consumption. The second case to markets for both inputs and produce. In Tanzania, in Kenya focuses on the Kenya Tea Development the Sumbawanga agriculture cluster benefits from Agency (KTDA) mini-hydro mini-grids. The two access to shared infrastructure and services, including simulated case studies are in Ethiopia and Mali. market access. This helped increase the viability of The Ethiopia study centers on a sugar estate with the agriculture sector as a creditable customer for self-generated power from bagasse and the opportu- electricity suppliers. nity of selling the power surplus to the main grid. The The seasonality of power demand from the Mali study analyzes capacity expansion of an existing agriculture sector can be a significant constraint xvi Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table ES.1: Summary of Ongoing or Planned Cases of Power-Agriculture Integration Project Tanzania Tanzania Zambia Zambia Kenya Kenya Name Sumbawanga Mwenga Mkushi Mwomboshi Oserian Tea Agriculture Mini-Hydro Farming Block Irrigation Flowers and Development Cluster Mini-Grid Development Geothermal Agency and Support Power Holdings Project (IDSP) Mini-Hydro Mini-Grids Overview Expansion A 4 MW hydro Extending a Grid upgrade Expansion of Development of electricity mini-grid transmission and grid the estate’s of hydropower supply to connected to line into a extension geothermal plants support the the main grid. farming area to support generation powering tea development Main local with significant irrigation capacity and factories and of an anchor load agricultural development distribution staff housing agriculture is Mufindi potential. and household network and selling cluster and Tea Estates electrification. to power surplus power surrounding and Coffee the farm’s to the grid. households Limited; 1,300 operations and through main households distribution power grid connected in within the extension. surrounding estate. communities. Commodities Maize, Coffee, tea Wheat, Tobacco, Floriculture Tea sunflower, soybean, wheat, finger millet, tobacco, soya, poultry, maize, paddy, vegetables, sunflower, sorghum coffee horticulture (tomatoes, onions, bananas) Financial The project The financial From a purely Positive With a positive Evaluation Viability is marginally viability of financial point financial NPV, financial NPV, of a sample financially the project of view and as estimated at the planned project, North unviable as a depends a stand-alone US$1.1 million. expansion Mathioya, stand-alone critically on the project, grid project of shows that project. ability to sell extension to 0.4 MW and the project excess power Mkushi was electrification is financially to the main profitable for of 2,000 viable, with grid. Despite the farmers households a NPV of financial but not for the is financially US$3.3 viability, capital utility; sharing viable. million; subsidies were of capital costs revenues provided to was however accrue from keep local an appropriate the sale of electricity and successful power to the tariffs low. approach grid and cost to project savings by tea financing. factories. Executive Summary xvii Project Tanzania Tanzania Zambia Zambia Kenya Kenya Economic Economic Economic Thanks to Positive Positive The same Viability benefits would benefits are households’ economic economic project is be significant positive (US$9 energy cost NPV was benefits were evaluated as (US$134 million); they savings, estimated at estimated economically million), come from increased US$2 million at US$2.5 viable, with justifying households’ yields from for the power million; the a NPV of the project; energy cost irrigation on line extension, main economic US$10 million; they come savings, small-scale mainly from benefit is based direct and mainly from reduced farms, and greater on increased indirect rural households’ reliance on job creation; irrigated household electrification cost savings, diesel backup the project’s tomato electrification impacts small-scale for the tea economic and maize and thus include irrigation estate, and NPV was production. savings due to electrification benefits, and job creation positive lower energy of staff margin uplift from newly (US$46 consumption housing, from market electrified million). costs (e.g., reduced access. businesses. less use of connection kerosene costs for and no more surrounding payment for households, cell phone and charging development services and of stand-alone disposable home systems. batteries). About 30,000 households will benefit from electricity connections. to viability. Large seasonal differences in electricity clusters (e.g., Sumbawanga in Tanzania) can increase dependent agricultural activities will impact the cost the viability of rural electrification. Cluster devel- recovery of electricity supply investments. In such opment has load diversity by design and thus is less cases, it is important to consider ways to mitigate the risky than relying on a single anchor load. If there is a impact of a variable load. One option, especially in the private electricity supplier and private off-takers, any case of mini-grid or captive generation, is the ability to such risk will be priced into the supply contract, thus sell excess power to the grid (e.g., Mwenga mini-hydro increasing the price of electricity for all customers. in Tanzania and KTDA mini-hydro development in In such cases, diversified cluster development can Kenya).3 During the post-harvest season, an increase also help reduce the price of electricity. In some such in the post-harvest processing activity may comple- instances, the public sector can also help mitigate this ment electricity demand from irrigation. In addition, risk through a grid connection and a feed-in tariff irrigation itself may reduce seasonality in agricultural (FiT), subsidies to increase the customer base, or production and thus electricity demand by allowing guarantee/insurance instruments. for multi-cropping (e.g., Mkushi in Zambia). Large-scale development of irrigation-based agri- When considering agricultural anchor loads, culture and sugar estates with excess generation can it is more risky for the investment to depend on a justify a main grid connection on a purely financial single large customer since any negative shock to the basis. Requirements for this—not all of which are read- customer would negatively affect operating reve- ily available in Sub-Saharan Africa—include relatively nues of the electricity supplier. As such, agricultural clear and empty land with good quality soils, a reliable xviii Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table ES.2: Summary of Simulated Cases of Power-Agriculture Integration Ethiopia: Power Generation Mali: Mini-Grid Expansion Project from Sugar Estates for Productive Uses Overview Self-generation of power from bagasse and sale Capacity expansion of an existing hybrid of power surplus to the main grid. mini-grid (diesel-solar PV) to serve productive users. Commodity Sugar Agro-industrial activities Financial Viability From the utility’s perspective, extending the From the perspective of Yeelen Kura, the grid to the sugar estate is not financially viable current financial situation of the Koury since the net present value (NPV) is negative— mini-grid is fragile; however, the capacity because it does not benefit from sales to the expansion project is profitable thanks estate, which self-supplies; however, from to a higher payment rate, additional the standpoint of the sugar estate, it is highly revenues, and proportionally low capital profitable (US$139 million). expenditure and operating expense (NPV of €103,000). Economic Viability The economic NPV for the whole period is The economic NPV for the expansion positive (US$367 million), thus justifying project project is slightly negative (−€18,000) as development. no significant savings are expected from agro-industrial customers (currently using individual diesel generators); however, it could become economically viable if other economic, environmental, and social benefits are considered (e.g., reduction in CO2 emissions, reduced reliance on imported fuels, and reduced exposure to price fluctuations). supply of sufficient water, and high quality physical and distance. The Sumbawanga cluster (Tanzania) and the market infrastructure. Suitable commodities include Mkushi farming block (Zambia) cases show that grid those typically cultivated on large-scale farms: maize, extension is the more viable option. wheat, sugar, rice, soybeans, and barley. Despite the advantages of the main grid, mini- The main grid has certain fundamental advan- grids may still offer the least cost solution to reach tages that may make it the most viable option, even unserved consumers, overcome grid unreliability, in cases where it is located at a distance. The multiple and leverage private-sector funds to accelerate rural generation sources connected to the main grid help electrification. Case studies in mini-hydro mini-grids mitigate the risk of power failure and enable the utility developed under the Mwenga (Tanzania) and KTDA to minimize costs by balancing supply profiles to match (Kenya) projects show how unreliable grid supplies demand. In contrast, a smaller isolated system based have led to the development of alternative generation on a single generation source may not be amenable to sources. However, the more typical case is establish- different load profiles and is at a greater risk of failure ing mini-grids in greenfield areas and access-deficit due to shutdowns of the sole generation facility. In countries setting up policies and regulations to create addition, due to economies of scale in generation and a level playing field and mitigate uncertainties for the ability to spread fixed costs over a wider set of private-sector, mini-grid operators. The two main ­ consumers, electricity from the main grid tends to be concerns are (i) the ability to be financially sound, cheaper than that from a smaller system. At the same either through charging cost recovery tariffs or time, the size of electricity load required to ensure via- receiving government subsidies and (ii) having regu- bility of grid extension increases with the capital costs lations that specify what happens when the large grid incurred for the extension, which, in turn, is related to reaches the mini-grid areas. Executive Summary xix A number of options exist to make projects finan- to bridge the gap between actual retail tariffs and the cially viable. First, to benefit from economies of scale, levels required for full cost recovery. the local generation capacity can be increased beyond the level of local demand, and surplus power can be sold to the grid. This option is particularly relevant in How can complementarities countries that have introduced FiT programs set above in power and agriculture be the utility’s avoided costs. Selling excess power makes harnessed? it possible to lower the per megawatt cost, but relies on the ability to sell excess generated power. For exam- To realize the full potential of agriculture-power ple, the capacity of Tanzania’s Mwenga mini-hydro integration in Sub-Saharan Africa, the region’s pol- mini-grid is greater than what the tea estate requires; icy makers and power companies must think about therefore, the surplus is sold to the utility and nearby demand creation. Governments should coordinate rural customers. Another option, as is done for the strategies in the power sector with complementary main grid extension projects in Zambia (i.e., Mkushi strategies on rural development and agricultural and Mwomboshi), is to require beneficiaries to partially extension. The experience of agriculture corridors, finance projects and share the development costs clusters, and growth poles should be analyzed and with major customers. Farmers partially contribute to applied on a wider scale. In addition, power compa- capital costs in exchange for receiving power. A further nies should coordinate with other related agencies option is load balancing across beneficiary categories, and institutions to maximize complementarities. which enables the spread of fixed costs, especially cap- Electricity can be prioritized in areas with large irriga- ital costs, across a larger pool of customers with diverse tion potential, combined with access to markets for peak-load profiles. agricultural goods. The sale of agricultural machinery, The role of subsidies to cover some costs should including irrigation pumps and small threshers, can be be highlighted. All of the distributed schemes have promoted as part of a package to encourage elec- received subsidy payments to decrease the level of tricity use in agriculture. In the process of developing cost recovery through retail tariffs. This contributes expansion plans, power companies should account for toward ensuring maximum capacity development, the electricity needs of, and benefits to, both small- increasing the project’s net present value (NPV), holder and commercial scale farmers. improving tariff affordability for customers, and Leveraging complementarities in rural devel- attracting private-sector participation. Subsidies are opment across sectors would likely result in higher particularly necessary for most privately developed, revenues for the utility companies and deliver small-scale projects under 5 MW. By subsidiz- greater economic benefits to rural areas. While ing household connections, which also tend to be power companies can prioritize regions with existing financially unviable, developers can be encouraged or potentially high levels of agricultural production, to expand their customer base to capture additional rural development or agricultural agencies can target subsidies, prioritizing smaller customers close to each areas that are able to take advantage of the many other rather than larger ones. productive use benefits of electricity. The utilities can National policy targets based on economic net create internal units responsible for encouraging the benefits, rather than financial viability, drive invest- productive and efficient use of electricity. Productive ments in rural electrification. For all the cases stud- use units can be responsible for promoting electric ied, the estimated economic viability was high. Power machinery in agriculture, from irrigation to harvest for agricultural use enables the development of pre- and post-harvest. Banks and other financial institu- viously unviable activities, which increases yields and tions should be incentivized to set up credit lines for lowers production costs. The benefits to households farmers and agricultural entrepreneurs to purchase and businesses include savings on energy expendi- agricultural machinery. Given the high expense of tures, better health, and improved educational out- using diesel powered engines for grain processing, comes. Wider benefits accrue from higher incomes campaigns by local government could be developed and improved quality of life. However, subsidies are to promote electricity as a substitute for diesel needed to make the schemes financially viable. All of engines among farmers in areas just gaining access the distributed schemes analyzed received subsidies to electricity. xx Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Coordinated planning encompassing geospa- Supporting the financial health of key sec- tial efforts and multi-agency inputs is necessary. tor institutions, central to the World Bank policy A geospatial map with information about future dialogue in the electricity sector, is important for developments of the national grid, as well as layered this agenda as well. The weak financial status of the data on agriculture and other rural infrastructure, is utilities prevents them from being able to develop important to understand where the load clusters are. financially viable projects without external support. These are the areas where feasibility studies of mini- Furthermore, their constrained cash flows increase grid developments could be carried out for potential the risk of non-payment for the power supplied by future tendering. Clarity in site identification and the private developers, which negatively impacts project regulatory environment is also useful for mini-grid costs and tariffs and, as a result, power affordability. developers and concessioners to allay fears on what If FiTs are not capped at the utility’s avoided costs, happens when the grid arrives. Such integrated maps, this situation could worsen, further deteriorating possibly housed in a national institution, can also the utility’s viability. From the perspective of power support more transparent decision making on infra- sector regulators, the extra cost and delays result- structure expansion and integrated rural development ing from inexperience in negotiating various supply approaches. arrangements may be a hindrance to developing Policy makers can support a stable regulatory private-sector power generation, distribution, and environment for electricity suppliers. To succeed, supply. projects must be implemented within a stable legal Finally, rapid changes over the last few years in environment that imposes requirements and provides small-scale generation and distribution technology, protection. Light-handed regulation of small-scale especially solar PV, have created opportunities to electricity systems is generally more favorable to test new models for viable rural electrification and developers and operators. For example, Tanzania’s power-agriculture integration. Recent techno- small power producer (SPP) framework allows private logical advancements and reduction in small-scale operators to function as power distributors and retail- generation costs have led to heightened interest in ers, charging fully cost-reflective tariffs. This type viable isolated mini-grid development models, such of regulation should tackle the economic barriers of as those based on shared solar PV systems and DC unaffordability and uneconomic supply. Regulation distribution lines. Compatible product development must also extend beyond the power sector to tackle (e.g., TVs, refrigerators, solar pumps, and grain mills) interactions with related sectors. Tanzania’s Mwenga is enabling increased productive use of electricity and Mini-Hydro Mini-Grid Project, one of the first proj- increased aggregate electricity demand from such ects of its kind, encountered significant delays when mini-grids to further improve their viability. While negotiating regulations over water rights, land access, there is limited experience of such mini-grids in oper- import laws, and building permits. Also, information ation (which thus explains why they are not reflected about future developments of the national grid and in our findings), this is a dynamic space with tremen- concession protection is crucial for dispelling devel- dous current interest and significant future potential opers’ reluctance and avoiding potential friction from to spur greater opportunities for power-agriculture tariff differences between customers. integration. endnotes 1. Authors’ calculation from the World Development Indicators (WDI) database. 2. Korwama (2011) estimates that 30 percent of agricultural produce in Sub-Saharan Africa is processed, compared to nearly 98 percent in some developed countries. 3. Apart from the mitigating impact of seasonal variation, the ability to sell excess power to the grid also helps to invest in large generation capacity and reduce costs due to economies of scale in generation. Agriculture and Power Nexus Chapter 1 A griculture predominates the livelihoods of the rural poor in Sub-Saharan Africa; thus, higher Box 1.1: Terminology growth in the agriculture sector, especially Clarification: Agriculture through increased productivity, is instru- and Agribusiness mental in reducing the incidence of extreme poverty in the region. Diao et al. (2012) estimate that the decline Agriculture refers to on-farm production. It in national poverty rates is up to four times higher for includes crops and livestock but not floriculture, agriculture-led growth, compared to growth led by nonag- ­ fisheries, or forestry. Although much agriculture ricultural sectors (e.g., 4.3 times higher for Kenya, 3.1 for in Africa is oriented to sustaining livelihoods, this Rwanda, 1.6 for Nigeria, and 1.3 for Ethiopia). Similarly, study focuses on commercial farming, recognizing ongoing research using the Global Trade Analysis Project that commercial farmers in Sub-Saharan Africa (GTAP) model of world trade finds that productivity are overwhelmingly small and medium in scale. growth in agriculture, compared to growth in other sec- Agribusiness denotes organized firms—from small- tors, is nearly three times as effective in reducing poverty. and medium-sized enterprises to multinational Agriculture and agribusiness comprise most income corporations—involved in input supply or down- generating activities in Sub-Saharan Africa’s largely stream transformation. It includes commercial rural economies (box 1.1), together accounting for nearly agriculture involving some transformation activities half of its gross domestic product (GDP) (figure 1.1). (even if they are basic). It includes smallholders and Agricultural production is the most important sector, microenterprises in food processing and retail to averaging 24 percent of the region’s GDP. Agribusiness the extent that they are market oriented. Indeed, input supply, processing, marketing, and retailing con- these producers and enterprises comprise the bulk tribute another 20 percent (World Bank 2013). Thus, of agribusiness activity in Africa today. transformation of the agriculture sector through improved productivity and incomes can simultaneously help achieve Source: World Bank 2013. both robust economic growth and poverty reduction. In other developing regions, agricultural transformation has resulted in declining numbers of the poor. Thus, for Sub-Saharan Africa, where poverty rates have remained constrains development of on-farm and off-farm eco- stubbornly high, utilizing agricultural transformation to nomic activities, as it does for other manufacturing and tackle poverty in rural areas—where more than 70 percent services firms. Rural electrification can raise productivity of the region’s poor live—is a critical part of any poverty and income when farmers switch from manual to electric- reduction strategy. ity powered inputs and small industries begin using electric For both agricultural and nonagricultural households, tools and machinery. Access to reliable electricity supply electricity is needed to raise living standards,1 as well as can increase productivity along the agriculture value chains enable broader economic development. Lack of access to and enable increased production and income generation reliable and affordable electricity in Sub-Saharan Africa for the farm sector and the rural economy as a whole. 1 2 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa The United Nations Sustainable Development Goals Past Performance: A Missed Opportunity (SDGs), adopted in September 2015, set a target for universal access to affordable, reliable, and modern energy Agricultural growth has typically lagged behind that of other services by 2030 (SDG 7). The acknowledgment of sectors in Sub-Saharan Africa. Vulnerability to weather modern energy access as a development goal builds on shocks, limited use of modern tools and inputs, low levels of the momentum created by the Sustainable Energy for processing, poor development of rural financial markets, and All (SE4ALL) initiative, which has galvanized the interna- market access barriers have all hindered agricultural growth tional community into action to achieve concrete energy and kept agricultural productivity and incomes low. Between related targets.2 Under SE4ALL, the three goals to be 2000 and 2013, the share of agriculture in GDP declined achieved by 2030 are: (i) universal access to modern by 6 percentage points (from 20 percent to 14 percent).3 energy services, (ii) doubling the share of renewables in Only a small percentage of the region’s agricultural the global energy mix, and (iii) doubling the growth rate of production undergoes industrial processing.4 For the energy efficiency. world’s high-income countries, processing adds about US$180 of value per ton of agricultural produce, com- pared to only $40 for Sub-Saharan Africa. This is related High Potential for Agricultural to the small size of the agribusiness sector compared to Transformation on-farm agriculture in Sub-Saharan Africa relative to other regions. For developing countries, including those in Historically, agriculture in Sub-Saharan Africa has Sub-Saharan Africa, the ratio of value added in agribusi- underperformed despite the region’s comparative ness to that of farming is typically 0.6. This ratio increases advantage stemming from abundant land and water to 2.0 for transforming countries (mainly in Asia), 3.3 resources. However, recent developments have created for urbanized countries (mostly in Latin America), and more favorable conditions for an agricultural trans- 13.0 for the United States, indicating significantly more formation. Today there is an expectation that well-­ value created in the downstream agribusiness sector than informed policies and investments can put agriculture on-farm production for countries outside Africa. These on a higher growth path to achieve its vast potential comparisons reflect the positive correlation between the and raise rural welfare. relative importance of agribusiness and economic growth: both per capita GDP growth (figure 1.1a) and human development indices (da Silva et al. 2009). Figure 1.1: Historical performance in agriculture a. Ratio of food processing to agricultural value added b. Market share of global exports 9.0 Food processing value added/agriculture added 0.6 8.0 % of world agricultural exports HUN 7.0 6.0 ARG 0.4 ROM BRA MEX 5.0 4.0 ZWE ECU IRN MYS BOL ZAF 3.0 PER 0.2 SVK SEN PHI MAR TUR 2.0 MWI THA IDN 1.0 EGY NPL BGD IND 0.0 0 UGA 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 0 2,000 4,000 6,000 8,000 GDP per capita, constant 2000 US$   Brazil Thailand SSA Sources: World Bank 2008, 2013. Note: In figure 1.1a, three-letter codes represent the following countries: ARG = Argentina, BGD = Bangladesh, BOL = Bolivia, BRA = Brazil, ECU = Ecuador, EGY = Egypt, HUN = Hungary, IDN = Indonesia, IND = India, IRN = Iran, MAR = Morocco, MEX = Mexico, MWI = Malawi, MYS = Malaysia, NPL = Nepal, PER = Peru, PHI = Philippines, ROM = Romania, SEN = Senegal, SVK = Slovak Republic, THA = Thailand, TUR = Turkey, UGA = Uganda, ZAF = South Africa, ZWE = Zimbabwe. Agriculture and Power Nexus 3 For more than four decades, Sub-Saharan Africa’s An Improving Outlook share in global agricultural export markets has been on the decline. By the early 1990s, the region’s share had The high yield gap between Sub-Saharan Africa and other fallen to about 2 percent, 5–6 percentage points lower regions underscores the large potential for Africa to catch than in the 1960s. Meanwhile, other important agricul- up with the productivity frontier (World Bank 2013). The tural exporters, including Brazil and Thailand, have seized increasing prominence of the agriculture sector among market share despite having a tiny fraction of Africa’s land policy makers, the private sector, and the development area, especially in the case of Thailand (figure 1.1b). community has been driven, in part, by the recognition of African imports of agricultural products have sky- decades of prior neglect of the sector by governments and rocketed due to the gap between regional demand and donors, as well as the urgent need to mobilize small-scale supply. From the 1990s to the 2000s, the balance of farmers to increase food production in order to avoid food trade in food staples for Europe and Central Asia, South security challenges in the near term. Asia, and East Asia and the Pacific moved from deficit Over the past decade, African governments have (i.e., imports exceeding exports) to surplus; however, for demonstrated a renewed and growing commitment toward Sub-Saharan Africa, this gap greatly expanded. While food agriculture. The improving policy environment, led by trade deficits are expected in regions without a compar- the Comprehensive Africa Agriculture Development ative advantage in food production, such as the Middle Programme (CAADP) (box 1.2), high investor interest, and East and North Africa, they are symptomatic of a missed technological advances that ease implementation of neces- opportunity in Sub-Saharan Africa, which is endowed with sary reforms, particularly in land administration, have cre- abundant natural resources for efficient production. ated excellent conditions for an agricultural transformation.5 The outlook for agricultural development in Sub- Saharan Africa is improving.6 Economic growth and Investment Funding Challenges urbanization have fueled an increase in food demand in Investment funding for the agriculture sector, especially Sub-Saharan Africa. In addition, continued improvements primary production, is limited by perceived high risks and in infrastructure and the benefits of lower oil prices have low returns. Poor infrastructure on farms and along the resulted in increased domestic food production. Although supply chains, low access to credit and product markets, recent declines in agricultural prices may dampen price and other regulatory hurdles have kept returns from incentives for agriculturalists, they may further increase agricultural investments in Sub-Saharan Africa below food demand and thus induce farmers to grow food and potential. Over the past decade, the increased inflows of other agricultural commodities for the market. commercial finance, especially foreign direct investment (FDI), have been vastly inadequate. Official development assistance (ODA) has helped, in part, to fill the gap. In Major Approaches to Agricultural 2003–12, ODA for agricultural projects in Sub-Saharan Development Africa rose 121 percent (from US$1.1 billion to $2.5 bil- lion). Over the same period, the share of aid allocated to There are two major approaches to agricultural devel- the agriculture sector in Sub-Saharan Africa grew from opment in Sub-Saharan Africa. The first is a cluster 37.4 percent to 40.3 percent, the highest share increase approach, which focuses on particular areas with a high for the period (Development Initiatives 2015). level of infrastructure access and development poten- The high costs of connecting agricultural land to back- tial. This generally involves support for large farms and bone infrastructure in Sub-Saharan Africa cannot be easily commercialized agriculture as growth poles. The second absorbed by most medium-sized farming businesses, let approach is smallholder agriculture, which centers on alone small-scale farms. But without these “last-mile” infra- support for smallholder farmers to increase their produc- structure investments, the region’s farmers cannot increase tivity and access to markets. These two approaches differ their productivity. Furthermore, without access to con- in their implications for electricity supply in rural areas. cessional funding, the establishment costs of an outgrower program, especially those involving provision of infrastruc- Cluster Approach ture services to small-scale farmer organizations, can be prohibitive, explaining why so few of the nucleus farm and Over the last 20 years, one rural development trend in outgrower models have been successfully established. multiple countries across Africa has focused on integrated 4 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 1.2: Africa’s Vision for Agriculture: CAADP Goals The Comprehensive Africa Agriculture Development Programme (CAADP), initiated in 2003, strives to improve country frameworks to support agricultural development. The CAADP’s initial 2015 target, extended through 2025, envisions that the continent should achieve the following goals: ºº Attain food security in terms of both availability and affordability and ensure access of the poor to adequate food and nutrition; ºº Improve the productivity of agriculture to attain an average annual growth rate of 6 percent, with particular attention to small-scale farmers, especially focusing on women; ºº Have dynamic agricultural markets among nations and between regions; ºº Integrate farmers into the market economy, including better access to markets, with Africa to become a net exporter of agricultural products; ºº Attain more equitable wealth distribution; ºº Become a strategic player in agricultural science and technology development; and ºº Practice environmentally sound production methods, featuring a culture of sustainable management of the natural resource base (including biological resources for food and agriculture) to avoid their degradation. Source: CAADP 2012. infrastructure and social development for specific areas. are likely to induce agricultural intensification. Both This cluster or corridor development approach has signifi- large-scale and smallholder agriculture will benefit from cant implications not only for the development of agri- increased productivity induced by spillovers, greater culture, but also for how electrification and other types of connectivity, and reduced transaction costs. The ability to institutions develop plans to serve such areas (annex A). serve wider markets for their goods and services will create Clusters are geographic concentrations of intercon- greater incentives to innovate. nected companies, including intermediate goods suppliers, The cluster approach brings together agricultural service and infrastructure providers, and associated insti- research stations, nucleus large farms and ranches, com- tutions in a particular product space or sector. Clusters mercially focused farmer associations, irrigated block farm- benefit from geographical agglomeration economies that ing operations, processing and storage facilities, transport may result from the proximity between intermediate and and logistics hubs, and improved “last-mile” infrastructure final goods suppliers, labor market pooling, and knowledge to farms and local communities. When occurring in the spillovers (Marshall 1890; Krugman 1991). Despite falling same geographical area, these investments result in strong transportation and communication costs, clusters con- synergies for agricultural growth, helping create the condi- tinue to be relevant today due to the underlying benefits tions for a competitive and low-cost industry. of increased firm productivity, innovation, and formation The essential elements of a cluster approach include of new businesses (Porter 1990). Transportation growth the following: corridors, a closely related concept, places the significant ºº Having a long-term strategy for agricultural develop- economies of scale of infrastructure development at the ment, recognizing that transformation occurs over an center of the benefits from spatial agglomeration. extended period (e.g., 10–20 years); In the case of agriculture, clusters can affect develop- ºº Understanding and leveraging vertical and horizon- ment in several ways. Improved access to infrastructure tal linkages between farms and other businesses to can lead to increased productivity of farms and companies maximize value addition; within a concentrated economic area. As opposed to ºº Commissioning robust analysis of the constraints on remote rural areas, these clusters of economic activity commercial agriculture and recommending how these benefit from joint access to necessary infrastructure can be addressed; services, linkages to upstream and downstream activi- ºº Establishing an independent public-private part- ties, and connectivity to markets. Better connectivity to nership organization to help coordinate and target markets and access to infrastructure, including electricity, Agriculture and Power Nexus 5 agricultural development programs and public invest- can be an important source of competitiveness in their ments; and own right. An additional benefit of smallholder led agri- ºº Leveraging government and development partner cultural growth is the much higher level of ­ second-round resources to catalyze socially and environmentally demand effects that occur when income gains are realized optimal private investment. by smallholder households, as opposed to large commer- cial farms.” Electricity is one of the fundamental requirements for Hazel et al. (2007) make the case for development cluster or corridor development. Investments in electricity of the smallholder sector, pinpointing the importance infrastructure must adequately account for long-term of infrastructure development to support it. “The case demand growth due to increased demand from large for smallholder development as one of the main ways to farmers, small farmers, farm service businesses, and other reduce poverty remains compelling. The policy agenda, tertiary development in such growth areas. Accounting however, has changed. The challenge is to improve the for medium- to long-term demand growth will allow bene- workings of markets for outputs, inputs, and financial fits to accrue from economies of scale and thus can lower services to overcome market failures.” The point is that costs to end consumers. numerous factors can support smallholder agriculture, including the coordinated efforts of farms, the private Smallholder Agriculture sector, nongovernmental organizations (NGOs), and government. Support can take the form of agricultural Most agriculture in Sub-Saharan Africa today involves research, agricultural extension, and infrastructure devel- smallholder farms, usually characterized by landholdings opment (e.g., roads and provision of electricity). of less than 2 ha, with a subsistence orientation. While the Given the “competing barriers” to agricultural large farm, agribusiness model has an important role to development, the provision of electricity infrastruc- play in promoting agricultural growth in Africa, small- ture, by itself, is unlikely to make an appreciable differ- holder agriculture is key to revitalizing the rural economy ence. Electricity investments must be coordinated with and tackling poverty. interventions targeting agricultural development (e.g., The question is what role should smallholder or family improving agricultural inputs and technology adoption; farms play, in contrast with large farms, in striving for pro- agricultural extension services; research on smallholder ductivity transformation in Sub-Saharan Africa. In agri- farming practices; and other infrastructure, including cultural economies, which describes most of Sub-Saharan roads, markets, and water supply). The combination Africa, smallholder agriculture comprises the majority of of these inputs will increase the growth of agricultural employment and production. With rising demand for sta- production and have a multiplier impact on the rural ple food crops and high-value commodities resulting from economy. rapid urbanization in the region, an increase in smallholder In short, it is not the role of electricity institutions productivity can arise from easing constraints on access to to promote agriculture; rather, their role is to support credit, infrastructure, and markets. Targeting the develop- agriculture in conjunction with other programs. This may ment of smallholder agriculture is also an effective way to seem a daunting task from a policy perspective, given reduce rural poverty. Thus, smallholders in Sub-Saharan that, in most governments, electricity, agriculture, rural Africa have a critical role to play as a source of agriculture development, and water institutions reside in isolated competitiveness. The World Bank (2009) finds that “con- “silos.” However, in countries with successful rural electri- trary to expectations, few obvious scale economies were fication programs, electricity companies have often found found in the production systems analyzed for the CCAA ways to deal with such silos, mainly through outreach and study. Compared with those of large commercial farms, coordination (Barnes 2007). For example, in Tunisia, the family farms and emerging commercial farms were typi- main electricity company (STEG) had regular meetings cally found to have lower shipment values at the farm level with rural development agencies and coordinated expan- and/or final distribution point (shipment values reflect sion plans to provide electricity in communities that were production and delivery costs). Large commercial farms receiving other development inputs. can play an important strategic role by contributing to Coordinated planning of rural electrification would the achievement of the critical mass of product needed to require a change in the way the electricity compa- attract local and international buyers, but the value chain nies operate, taking into account expected growth in analysis shows that investments in smallholder agriculture ­ energy-intensive agricultural activities and development 6 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa programs in the pipeline. To do this, electricity companies held its own for some cash crops (e.g., cocoa, rubber, need to develop an effective information sharing mech- fruits and vegetables, and tobacco) and has even gained anism with relevant agriculture sector stakeholders. This market share for others (e.g., cashew, tea, and sesame could involve reaching out to relevant agricultural agen- seed), showing some evidence of its productive potential. cies; promoting productive uses of electricity; and under- Third, Sub-Saharan Africa is poised for demographic standing future growth and development trends, especially transition and wealth creation, reflecting the growing with regard to smallholder agriculture. Electricity access aspirations of its people. According to the United Nations, for agriculture and rural businesses could effectively be between 2013 and 2050, the region’s population will more promoted as part of an overall strategy to support small than double, from about 900 million to 2.1 billion (United farmers through a variety of activities (e.g., development Nations 2013). While one-third of its population is already of farm cooperatives to purchase and market local farm living in urban areas, this proportion should increase to goods; machine rental; and agricultural extension, includ- 50 percent by 2035. Globally, urban food markets are set ing advice on irrigation practices, seeds, and fertilizers). to increase fourfold, exceeding US$400 billion by 2030 (World Bank 2013). For Africa’s 11 biggest economies, the middle class, defined as those earning at least US$450 per Agricultural Growth to Raise Rural month, tripled between 2000 and 2014 (from fewer than Welfare: Reasons for Optimism 5 million people to 15 million). Over the next 15 years, these numbers may rise by a further 25 million (Standard There are four main reasons to believe that agriculture in Bank Research 2014). Sub-Saharan Africa is poised for growth that can con- Sub-Saharan Africa’s rapid population growth, tribute significantly to raising rural welfare. First, relative accompanied by robust economic growth, is creating to much of the rest of the world, the region’s land and a huge regional urban market for agricultural goods. A water—the major natural inputs necessary for growing recent World Bank study on agribusiness predicts that crops and raising livestock—are underutilized, creating a the market for agricultural goods and commodities could huge potential for agricultural growth (figure 1.2). Of the reach US$1 trillion by 2030 (figure 1.3). It states that world’s total land area suitable for sustainable production “the majority of the increase in food consumption will expansion—that is, non-protected, non-forested land with occur in cities. Based on the United Nations’ projections low population density—Sub-Saharan Africa has the larg- of urbanization and assuming that the per capita value of est share by far, accounting for about 45 percent.7 In the food consumption is 25 percent higher in urban areas than case of Latin America, which accounts for only 28 percent rural areas, the urban market is set to expand fourfold in of land suited for production, 73 percent of that amount is 20 years” (World Bank 2013). This expansion in regional located within six hours’ travel time to the nearest market, demand will create an enormous opportunity for African compared to just 47 percent in Sub-Saharan Africa—a agriculture and agribusiness. result of the subcontinent’s generally poor state of infra- Fourth, agriculture is critical for managing the urban structure (Sebastian 2014). Sub-Saharan Africa also has transition that Africa will undergo. To date, this process significant untapped water resources. Only 2–3 percent has been driven to a large extent by populations being of the region’s renewable water resources are being pushed out of rural areas, rather than cities attracting a utilized, compared to 5 percent worldwide (World Bank workforce by acting as growth poles. It would be a more 2013). Its irrigation intensity, one of the lowest in the positive process were it driven by improving economic world, represents only 5 percent of total cultivated area, opportunities in the cities that would gradually pull in rural compared to 37 percent for South Asia and 14 percent in residents, rather than declining conditions and periodic Latin America (World Bank 2008). Despite an absolute disasters in rural areas that push residents out. The latter abundance of water resources, lack of irrigation develop- situation often leads to conflict and waves of migration ment and storage capacity has limited the availability of that cities find difficult to absorb, typically leading to water in certain basins, resulting in water stress. Also, the expanded slums. The objective of a transition strategy—of uncertainties related to climate change raise concerns which electrification is a key element—is thus to enhance about future water availability (box 1.3). living conditions and economic opportunities in rural areas. Second, despite Africa’s overall decline in the share In this context, agriculture and agribusiness can play of agricultural exports, a recent disaggregated view tells a critical role in jump-starting the economic transfor- a more nuanced story. Since the early 1990s, Africa has mation through development of agro-based industries in Agriculture and Power Nexus 7 Figure 1.2: Land and water resources potential in Sub-Saharan Africa a. Land potential, by world region b. African countries with largest available land resources Uncultivated arable land, Million ha million ha Sudan 202 Sub-Saharan Africa DRC Madagascar 123 Mozambique Latin America Chad Zambia 52 East Europe and Angola Central Asia Tanzania CAR 14 Ethiopia East and South Asia Cameroon Kenya 54 Rest of the world Mali 112 m ha Others 0 50 100 150 200 250 0 10 20 30 40 50 Land available Area less than 6 hours to market    Cultivated Available c. Aquifer productivity in Africa Aquifer productivity Very High: >20 l/s High: 5–20 l/s Moderate: 1–5 l/s Low-Moderate: 0.5–1 l/s Low: 0.1–0.5 l/s Very Low: <0.1 l/s Sources: World Bank and Schaffnit-Chatterjee 2014 (figure 1.2 a, b); British Geological Survey (figure 1.2c) (http:// www.bgs.ac.uk/research/ groundwater/international/africanGroundwater/maps.html). 8 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 1.3: Making Africa’s Power and Water Infrastructure Climate Resilient Uncertainty over water availability for productive Figure B1.3.1: Changes in irrigation uses is a critical issue facing Sub-Saharan Africa’s revenues from climate change, 2015–50 infrastructure investments, especially long-lived (present value) infrastructure (e.g., irrigation schemes, dams, and $24.8 billion power stations). Variations in annual rainfall and 90 gain monthly rainfall distribution, along with tempera- ture changes due to drier or wetter climates, could Di erence from reference case (%) 10 $1.8 billion $0.2 billion put power and water infrastructure at risk, affecting gain $0.3 billion gain gain $3.9 billion gain $0 billion gain operation and cost over their life span. Beyond 0 impacting the technical performance of infrastruc- $42.1 billion $2.4 billion ture, uncertainty about drier or wetter futures could –10 $13.2 billion loss loss loss significantly modify its financial viability, incurring $0.8 billion $7 billion losses or gains. In a drier scenario, for example, –20 loss loss shortfalls in irrigated production could raise demand $0.9 billion loss for food imports, and thus increase food prices (fig- –30 Volta Eastern Zambezi Nile Niger Senegal ure B1.3.1). Nile Equatorial Lakes Cervigni et al. (2015) highlight significant dispari- Basin ties across Africa’s seven main river basins: Congo, Maximum relative gain due to climate change/best scenario Niger, Nile, Orange, Senegal, Volta, and Zambezi. Maximum relative reduction due to climate change/best scenario The study estimates that, in dry scenarios, loss in irrigation revenue could range between 5 and Note: The bars reflect, for each basin, the range of economic 20 percent for most basins. For wet estimates, outcomes across all climate futures; that is, the highest increase (light blue bars) and highest decrease (dark blue bars) of irrigation revenue gains could reach 90 percent for the Volta revenues (discounted at 3 percent), relative to the no-climate- basin, but would be vastly less (1–4 percent) for the change reference case. The outlier bar corresponding to the Volta other areas. Under the driest scenarios, unmet irri- basin has been trimmed to avoid distorting the scale of the chart gation demand could drop by more than 25 percent and skewing the values for the other basins. Estimates reflect the in the Zambezi basin. The magnitude of impact will range, but not the distribution, of economic outcomes across all depend on the willingness and ability of decision climate futures. Each basin’s results reflect the best and worst makers to integrate climate projections and their scenarios for that basin alone, rather than the best and worst uncertainty into the planning and design of power scenarios across all basins. The Congo and Orange basins are and water infrastructure. excluded because the effects on irrigation are minimal. Africa’s need to tap its irrigation potential represents a window of opportunity to make power and water infra- structure climate resilient. Although such a paradigm shift will take time, practical steps to integrate climate resil- ience can be undertaken now. For example, Cervigni et al. (2015) recommend defining and promoting technical standards for integrating climate change into project planning and design and launching training programs target- ing relevant stakeholders. Source: Cervigni et al. 2015. Agriculture and Power Nexus 9 Figure 1.3: Projected value of food a vibrant agricultural sector. Investments in agricultural markets in Sub-Saharan Africa productivity can spur the development of downstream agribusiness; in turn, agribusiness investments stimulate 1,000 agricultural growth through the provision of new markets and development of a vibrant input supply sector. Micro-, 800 small-, and medium-sized enterprises (MSMEs) comprise the bulk of Sub-Saharan Africa’s agriculture-related value chains. In West Africa, for example, three-fourths of $ Billion 600 agriculture-related firms are micro or small enterprises (Staatz 2011). Taking advantage of this opportunity requires that 400 both farmers and agribusinesses ramp up production, while becoming more competitive; otherwise, the balloon- 200 ing demand will be filled by imports. This requires devel- oping agriculture value chains and agribusiness to enhance processing, logistics, market infrastructure, and retail 0 networks, all of which require electricity. 2010 2030 However, electricity remains a critical constraint to Urban Rural the development of the agro-industrial sector. According Source: World Bank 2013. to data from WBG enterprise surveys, the majority of firms in many countries of Sub-Saharan Africa identify lack of electricity access as a major obstacle (figure 1.4a). Figure 1.4: Electricity as a constraint to food-sector development in Sub-Saharan Africa a. By country b. Comparison with other sectors Percent of firms identifying electricity as a major Electricity considered as a constraint to invest in constraint to develop the food sector Sub-Saharan Africa (data collected from 2006 to 2014— (Enterprise Surveys—World Bank Group) Enterprise Surveys—World Bank Group) Total average Burundi Congo, Dem. Rep. Focus on the Guinea food sector Guinea-Bissau Ghana Tanzania Senegal Uganda All sectors Mali Angola Zimbabwe Rwanda 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Zambia   Nigeria Mauritius Mauritania Namibia Kenya Mozambique Madagascar Swaziland 0.0 20.0 40.0 60.0 80.0 100.0 Source: WBG 2015 (http://www.enterprisesurveys.org/). 10 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa In fact, the fraction of firms in the food sector that competitive for regional and global markets, and consider electricity a constraint to investment is signifi- ultimately increasing the profitability of agricultural cantly higher than the average fraction in all other sectors activities. (approximately 29 percent, compared to less than 15 per- cent) (figure 1.4b). Successful commercial agriculture is typically charac- Rural Electrification Has terized by the following elements: Lagged Behind ºº Ample suitable land, with benign climate conditions A majority of Africans—nearly 600 million people—live and reliable water availability. without electricity; instead, they rely on kerosene or ºº Private-sector participation in sector development, dry-cell batteries as coping mechanisms. The latest with higher skills levels and access to international estimates peg Sub-Saharan Africa’s electrification rate at capital and markets, with strong government support 35 percent overall, with 69 percent in urban areas and just (e.g., through a favorable policy and regulatory envi- 15 percent in rural areas (figure 1.5a). Viewed from space, ronment and publicly funded research and develop- the picture of Africa’s nightlights, showing large sections ment and infrastructure). of perpetual darkness, is a stark contrast to the rest of the ºº Affordable and reliable access to supporting infra- developing world, and the evolving disparity is enormous structure, in the form of reliable electricity supply, (figure 1.5b). transport links to markets, and irrigation in drier Historically, the region’s population growth has climates (often powered by grid-based electricity). outpaced the rate of expanding electricity access, and the ºº Clusters of large-, medium-, and small-scale gap in rural areas is enormous. Amid a population increase commercial farming, processing, and services firms of 202 million, only 59 million people have received concentrated in discrete geographical areas. Taken electricity. If business as usual continues, by 2030, together, the result is a reduction in costs of produc- Sub-Saharan Africa will be the world’s only region with tion through economies of scale, making prices more an increase in the number of people without electricity Figure 1.5: Electrification rate, by developing region a. Millions of people with and without access, 2012 b. Evolution of access (%), 1990–2012 2,000 100 90 80 1,500 23 363 70 0 60 555 1,000 754 50 977 40 22 589 30 61 500 271 807 20 477 659 14 10 0 237 0 94 147 278 272 36 0 0 46 74 54 109 87 0 1990 2000 2010 2012 Oceania CCA NA WA LAC SEA SSA DEV EA SA People with access (rural) People with access (urban) EA LA NA Oceania Population without access     SA SEA SSA WA Source: IEA and World Bank 2015. Note: CCA = Caucasus and Central Asia, EA = East Asia, LA = Latin America, NA = North Africa, SA = South Asia, SEA = Southeast Asia, SSA = Sub-Saharan Africa, and WA = West Africa. Agriculture and Power Nexus 11 access. Furthermore, the urban/rural disparity in elec- agricultural value added and incomes. Generally, the most tricity access is set to widen as most expansion is likely to dramatic changes in agricultural development due to rural occur in densely populated urban areas (IEA and World electrification have resulted from increased irrigation. Bank 2015). With greater access to electricity, it is more cost-effective The biggest challenge to rural electrification in the for farmers to irrigate their fields since electric pumps Sub-Saharan Africa region is the lack of commercial via- require low maintenance and are more efficient than die- bility of expanding connections. Low population density, sel alternatives. Irrigation also allows farmers to produce coupled with the limited purchasing power of most rural multiple crops in a single year and improve the productiv- consumers, implies that, in many cases, investment in ity of existing farms. These advantages lead to higher crop rural grid extension is cost-prohibitive. This problem is yields and incomes.8 compounded by the poor financial health of the region’s This relationship has most often been documented in distribution utilities, owing to a combination of factors India, which historically has emphasized the use of irriga- (e.g., low consumer base, historical mismanagement, tion pumps and new agricultural technologies to improve inadequate tariffs, high generation costs, and high rates agricultural productivity (Barnes, Peskin, and Fitzgerald of technical and nontechnical losses). The high cost of 2003). While efforts to improve rural development supply, coupled with low tariffs, puts an inordinate strain through electrification have been relatively successful on sector finances. in some countries, the question is whether this experi- This situation, in turn, traps the sector in a self-­ ence is applicable to Africa, with its low levels of existing reinforcing cycle of low investments in expansion and irrigation. improvement, resulting in an expensive, poor quality The productive impact of rural electrification depends electricity supply, circling back to low investments. Thus, heavily on several enabling factors: government policy, many of the region’s countries are stuck in a cycle of infrastructure, and complementary development pro- low generation capacity, excess demand, and inadequate grams. Electrification is an important enabler for the mobilization of private-sector investment. Breaking this development of rural businesses (e.g., small commercial negative cycle requires a multipronged approach custom- shops, grain mills, sawmills, and brickworks); however, ized to the financial, economic, and political realities of it cannot produce an explosion of economic activity in particular countries. Least-cost grid expansion, wherever the absence of roads and access to finance and markets. viable, should be creatively complemented by a decen- If these complementary conditions are inadequate, the tralized off-grid strategy based on distributed generation growth of rural economies, especially agriculture, will in the form of mini-grids, micro-grids, or stand-alone likely remain lethargic and may, in turn, adversely impact systems. the viability of the rural electrification program.9 One potential solution to address the region’s rural electrification challenge is having an anchor load, defined Agriculture as an Anchor Load as large consumers that offer power utilities a consistent for Rural Electrification and substantial source of revenue, which offsets a portion of the fixed costs of electricity supply to rural households. In recent years, African governments, donors, and the Anchor loads help ease the constraint posed by the low private sector have been reviewing the success stories demand profile of rural customers. Guaranteed demand of such countries as Brazil and Thailand in an attempt to from anchor-load customers ensures the power producer replicate or adapt agribusiness and rural electrification or utility a certain level of revenue, and may help to defray development models that take individual country charac- the fixed costs of rural electrification through demand teristics into consideration. In the case of India, the most aggregation (along with household and commercial notable example, rural electrification was strongly linked demand in neighboring communities of the anchor load). to the promotion of high-yield crop varieties and the In short, an anchor load helps overcome the problem spread of irrigated agriculture, facilitated by electric water of low demand, which constrains the viability of rural pumps with subsidized or free electricity. Here it was clear electrification. that the financial viability and reliability of rural electrifi- In some developing countries, the Anchor Business cation were linked to promoting productive uses. Community (ABC) model is being piloted, using cell- The financial viability of agricultural anchor loads rests phone towers and mining companies as anchor loads.10 In on the ability to use electricity to generate an increase in this context, the supply options range from self-supply 12 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa by the agribusiness to intermediate arrangements with along the various agriculture value chains, aggregated with an independent power producer (IPP) to grid extension. ­ commercial/household electricity demand, can potentially A recent study that analyzed the integration options make it feasible to extend the grid or create opportuni- between power and mining established a typology of ties for small IPPs and mini-grid operators. In addition power sourcing options for mines (Banerjee, Romo, and to demand aggregation, supplying both household and McMahon et al. 2015). agro-processing demand may create a balanced daily load Agriculture can potentially fit into this category of profile, helping to disperse capital and fixed operating anchor load to sustain small-scale supply arrangements costs over a larger set of consumers. with commercial establishments (including irrigation) and In addition to providing anchor loads, agricultural households in rural areas. In this way, electricity demand production can provide fuel for off-grid solutions in rural along the agriculture value chains, as well as commercial/ areas (annex B). Agricultural by-products can serve as household electricity demand, can create opportuni- cheap sources of locally available fuel for biomass electric- ties for the IPPs and mini-grid operators. In addition ity generation; they can be derived from various types of to demand aggregation, supplying both households and processing (e.g., cotton, groundnut, soybean, wheat, and agro-processing may create load balance; the demand of other cereals), but the most common ones are rice husks households and agro-processing peak at different times and sugarcane waste (i.e., field waste and bagasse). of the day, which can help to disperse capital and mainte- Such opportunities are now being commercially har- nance costs over a larger set of consumers. nessed in various countries and regions of the world. For The development of anchor loads can benefit both example, India has created a business model to serve rural centralized and decentralized approaches to rural elec- households using husk power, whereby agricultural residue trification. In the case of grid extension, promoting the (e.g., rice husks, mustard stems, corn cobs, and certain development of relatively large anchor customers in off- grasses) is cost-effectively converted into electricity. In grid areas could tip the balance in terms of the economic this study, the scope of agriculture’s role is limited to that viability of extending the grid to connect to the anchor of an anchor load in rural areas of the Sub-Saharan Africa load and bringing the grid closer to communities without region. electricity access. In current-day industrialized econo- mies, such anchor customers as mills and factories were an integral part of the electrification experience. In Sub- Study Purpose and Methodology Saharan Africa too, national grid expansion plans tend to prioritize district commercial centers and areas with Rural electrification is at a crossroads in Sub-Saharan factories or other large commercial customers. Beyond Africa; for many countries, the challenge is overwhelming, demand from the anchor customer, grid extension can be but opportunities are also emerging. It is up to govern- made viable through the potential to sell electricity back ments, the private sector, and international communities to the grid (in cases where there is an in-house generation in the region to decide how these opportunities will be facility). harnessed for the benefit of Africans living in the dark. Grid extension may not be viable if anchor customers Recently, the WBG’s Energy and Extractives Global are not large enough or are located in relatively remote Practice in the Africa Region commissioned a series of areas. In such cases, smaller isolated grid systems or mini- studies to explore potential solutions to the challenge grids can be used to save on costs associated with trans- of bringing power to Africa. This study, which follows on mission infrastructure. Mini-grids can be developed by the recent initiatives of Banerjee, Romo, and McMahon aggregating demand from the anchor load and surround- et al. (2015), Hussain et al. (forthcoming), and Hosier ing communities, with electricity generation and distribu- et al. (forthcoming), is designed as a joint effort between tion undertaken through a context-specific combination the Energy and Extractives, Agriculture, and Trade and of a small, in-house power producer and anchor business Competitiveness Global Practices. It also complements or public utility. the ongoing analytical work of the Latin America and For both on- and off-grid access solutions, the Caribbean region on energizing agriculture. presence of an anchor-load customer greatly improves This study’s overall aim is twofold: (i) to identify the financial viability. In principle, activities along agri- potential synergies between agriculture value chains culture value chains require electricity and thus might and rural electrification expansion and (ii) to examine serve this role. The electricity consumption of activities the challenges in harnessing this potential. Its specific Agriculture and Power Nexus 13 objectives are to (i) conduct an evidence-based analy- been developed, as well as those in progress or proposed. sis of the extent of the potential of power-agriculture The cases covered a range of commodities (e.g., fruits, integration for specific case studies on agriculture value floriculture, maize, sugar, tea, vegetables, and wheat). chains; (ii) assess alternative supply arrangements (busi- Since agriculture is a dispersed activity with varied ness models) for providing electricity to the combined scales of production, results of this analysis need to be power demand of agriculture and local commercial and considered with the following caveats. First, although residential demand; (iii) analyze barriers and institutional the study provides an estimate of power demand from mechanisms that will create the enabling conditions for agriculture in 2030, it was unable to capture the location private-sector participation in this space; and (iv) iden- of this demand, the extent to which it can be met by tify operationally relevant opportunities for piloting this simply increasing the generation capacity of national grids concept. (i.e., the grids already extend to production and process- This work builds on two background studies prepared ing areas), and whether alternative power sources (e.g., by the consulting consortium of Economic Consulting isolated electricity mini-grids) are the most viable supply Associates (ECA) and Prorustica in 2014–15, which options. Second, the study was unable to capture the nec- involved field visits and stakeholder discussions in the essary financial viability of power supply with reference to countries covered. The first study analyzed the landscape the price that the agricultural activities could afford to pay for rural electrification centered on agricultural activities, for power. while the second examined a set of eight case studies on The rest of this report is organized as follows. powered agribusiness activities from across Sub-Saharan Chapter 2 presents the context of power needs from Africa (Ethiopia, Kenya, Mali, Tanzania, and Zambia). agriculture, while Chapter 3 reports on the detailed anal- The primary focus of the landscape study was on power ysis of power needs by selected value chains. Chapter 4 consumption of agricultural activities within value chains, discusses power supply arrangements for a suite of case identifying where sufficient demand from the activity studies in three countries, encompassing technical, makes it possible to provide an economic or socioeco- economic, and financial analysis. Chapter 5 reviews the nomic rationale for an electrification project that may potential for harnessing power-agriculture synergies then be extended to support surrounding communities. and provides alternative integration scenarios using two The case studies comprised both national grid-connected simulated case studies. Finally, Chapter 6 summarizes the activities and those powered by distributed generation study’s key findings and recommends actions required to systems. They included power schemes that had already promote power-­ agriculture integration. endnotes 1. Households that connect to the electricity grid benefit immediately from better household lighting. With brighter light in the home, children spend more hours studying, adults have more flexible hours for completing chores and reading books, and home-based businesses remain open longer in the evenings, producing more items for sale. Once rural families connect to the grid, television sets, fans, and an array of other household appliances gradually become more affordable (Barnes 2014). 2. The SE4ALL initiative was launched by the United Nations (UN) in 2011. It is co-chaired by the UN Secretary General and World Bank Group (WBG) President; SE4ALL helped place energy access explicitly on the global development agenda, thus filling the gap left by the Millennium Development Goals (MDGs), which did not include any energy access goals. 3. Authors’ calculation from the World Development Indicators (WDI) database. 4. Korwama (2011) estimates that 30 percent of Sub-Saharan Africa’s agricultural produce is processed, compared to nearly 98 per- cent in some developed countries. 5. Focusing on the enabling environment, WBG (2016) measures regulations that impact firms in the agribusiness sector. It collects and reports data on 18 indicators for 40 countries across the world; the indicators capture aspects related to production of inputs and market enablers to help policy makers better understand barriers to growth and transaction costs imposed by the regulatory environment. 14 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa 6. Africa’s economy has been expanding at a relatively high rate. Following a very strong decade from the beginning of this century, growth in 2015 was more modest, at 3.7 percent (World Bank 2015). Growth rates over the next several years are projected at well above 4 percent. 7. About two-thirds of this area is spread over eight countries: Angola, Democratic Republic of the Congo (DRC), Madagascar, Mozambique, South Sudan, Sudan, Tanzania, and Zambia (World Bank 2013; Deininger and Byerlee 2011). 8. The impact of electricity will be lower in areas that use gravity-fed irrigation since the value added by electricity is likely to be rela- tively minor. The main impact will be realized by farmers using agricultural pump sets or other forms of mechanized irrigation. 9. A recent WBG study states that electricity access is critical to promoting a more commercialized agriculture sector in the devel- oping world; it emphasizes the importance of rural electrification as an enabling condition for agribusiness development, and discusses indicators on electricity access, reliability, and affordability (WBG 2015). 10. In the ABC model, anchor customers are the main off-takers for the generated power. Business refers to small local businesses and shops; community refers to households, farming needs (including irrigation), and local institutions. Power Needs of Agriculture Chapter 2 A gricultural transformation in Sub-Saharan to higher value urban and export markets. An increase Africa implies a shift away from household in an irrigated area to reach its estimated potential and subsistence farming toward a more market-­ improving existing irrigation practices will require electric- oriented farming sector that is effectively ity for water pumping. The mechanization of basic milling able to supply demand across the world. Achieving this or grinding that is largely done manually will require elec- transformation involves increased use of modern farm- tricity to run machines. Storage of high-value perishables ing inputs, greater value addition through post-harvest awaiting transport to demand centers will require electric- processing, and access to markets through transportation ity for chilling; and such processing activities as pulping, and storage. drying, heating, and packaging will also demand electricity. Electricity is a key input required to create greater This chapter explores the synergy between agricul- value added in the agriculture sector through enabling tural growth and rural electrification and provides initial irrigation, processing, and storage. Growth in agricul- estimates of power demand from agriculture in 2030. tural incomes is directly dependent on farmers’ ability The value generated by agricultural activities that demand to increase their yields through irrigation, processing of electricity can help tip the scales of commercial viability produce to retain a greater proportion of the value added of rural electrification interventions. along the full supply chain, and proper storage of produce to prevent spoilage. A growing agriculture sector will thus produce greater demand for electricity along its value Power Needs across the chain, from both on- and off-farm activities. Agricultural Agriculture Value Chain transformation, through increasing rural electricity demand, can thus go hand-in-hand with an expansion in Electricity input is vital for the adoption of modern rural electricity access. productivity enhancing technologies and thereby the A structural shift in agricultural markets is set to integration of small-scale farming into high-value and induce demand for electricity from agriculture. With export-oriented value chains. The implications for elec- growing domestic and export markets for agricultural tricity demand from such a shift in the agriculture sector products, the need for increased agricultural productivity of Sub-Saharan Africa will be determined by the extent to will necessitate greater on- and off-farm mechanization which modern techniques are adopted at each stage along of agricultural and agribusiness practices. In addition, eco- the value chain and the scale of each activity. In addition nomic growth is set to create markets for new products to electricity requirements, the potential of various crops and higher value commodities for urban markets and as to gain from irrigation and processing activities can vary intermediate inputs for various industries, especially in the widely. food sector. Depending on crop characteristics and target mar- Electricity demand from agriculture stems from the kets, value chains differ in post-harvest processing and various processes along the agriculture value chain—from preservation requirements. This creates differing on- and on-farm irrigation and off-farm grain milling to larger sec- off-farm demand for electricity for each value chain. ondary processing (e.g., pulping and packaging) that caters In order to examine the nature of electricity use along 15 16 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 2.1: Power needs across Powered irrigation systems are prevalent in Southern agriculture value chains and East Africa, and are emerging in West Africa. To a large extent, West Africa and the Sudano-Sahelian region utilize small-scale irrigation systems, which tend to be Post-harvest & gravity fed. Secondary On-farm primary processing Like other powered activities in agriculture, the processing electricity requirements of powered irrigation equipment depend on system scale, form of irrigation, and specific geographic conditions—the latter factor making it difficult • Irrigation • Milling, drying, • Packaging, to develop accurate estimates of electricity use for irri- chilling, etc. bottling, etc. gation. The two primary power demands for irrigation are (i) sourcing bulk water from some water body, such as a Rural Urban/peri-urban dam or river and (ii) distributing it over the cultivated area. Irrigation systems commonly used in Sub-Saharan Africa range in scale from manual systems to surface agriculture value chains, the sources of growth in future flooding and localized systems to center pivots. Manual electricity demand can be divided into three sources, as systems, including simple buckets to support small-scale follows (figure 2.1 and annex C): farmers, require no power. Surface flooding and localized systems (e.g., stationary drip schemes and pressurized ºº The potential for expanded irrigation, which is the pri- systems, such as sprinklers1) require power to source the mary on-farm source of electricity demand. bulk water that cannot be accessed by gravity only. Center ºº The potential growth in post-harvest and primary pivots may require power for bulk water supply, as well processing activities from both new and existing as for pressurizing water for the system and possibly for production; activities include cleaning/drying, milling, system mechanics (e.g., motors to rotate the pivot span). cassava processing (chipping), chilling and cold stor- In all four cases, power demand is related to system age, meat processing, and oil extraction. scale, but will vary per unit of area covered. In each case, ºº The potential growth in secondary processing activ- pumping bulk water comprises the major demand and will ities that cater mainly to urban markets and provide depend on the vertical and horizontal distances of the intermediate inputs to other production processes; scheme from the water source (table 2.1). activities include thermal treating, canning, bottling, For irrigation systems that use gravity to redistribute and packaging. water, power may only be required for bulk water pump- These several activities are presented in decreasing ing into storage (if needed). The most efficient pumping order of rural presence. Virtually all irrigation occurs systems do this to meet infield demand, running nearly in rural areas, and post-harvest and primary processing continually. But some systems may design their capacity usually occur shortly after the rural harvest, depending with larger pumps so as to require pumping for fewer on scale. Secondary processing is more likely to take hours within a day. This design is inefficient from the view- place near trading hubs and demand centers in urban or point of electricity supply, as it would require a greater peri-urban areas, although, under appropriate conditions, peak generation load. some smaller-scale operations can be viable in rural areas. Benefits from irrigation come from increased yields The prevalence of irrigation potential in rural areas and and reduced weather-related risks. Enhanced irrigation the benefits across value chains imply that irrigation is the practices may thus result in large benefits from increased largest potential source of power demand from agriculture crop yields, leading to higher farm revenues. Giordiano in Sub-Saharan Africa. et al. (2012) find that Sub-Saharan Africa has considerable area under small-scale irrigation or improved agricultural water management. The study estimates that investments Irrigation Potential in dry-season irrigation for rice could potentially increase The irrigation intensity in Sub-Saharan Africa is the lowest yields by 70–300 percent. The same study estimates that in the world; only 6 percent of the region’s cultivated land investment in relatively low-cost motorized pumps, ben- is irrigated, compared to 44 percent in Asia (FAO 2005). efiting 185 million across the Sub-Saharan Africa region, Irrigation intensity and technique vary across the region. could generate net revenues of up to US$22 billion a year. Power Needs of Agriculture 17 Table 2.1: Power Demand for Irrigation, by System Type Cultivation Estimated Power System Methods Power Demand/Unit Typical Area Type Supported Crops Supported Components (kW/ha)a Coverageb Surface flooding Small- and large- Rice, sugarcane, Possibly bulk water, 0.5–0.9 600 m2– (furrow and paddy scale commercial. tomatoes, infield pumping. 20,000 ha systems) citrus. Micro irrigation Small-scale Floriculture, Possibly bulk water, 0.5–0.9 600 m2– (drip and trickle) and intensive horticulture, infield pumping. 20 ha schemes commercial. seedling propagation, citrus, vegetables, potatoes. Micro jet irrigation Some small-scale, Floriculture, Possibly bulk water, 0.5–0.9 5–50 ha mostly large-scale horticulture, citrus, infield pumping. commercial. macadamia, some tree crops. Portable impact Small- and large- Floriculture, Possibly bulk water, 0.5–0.9 600 m2– sprinkler systems scale commercial horticulture, grain infield pumping. 20,000 ha (drag-line and (broad-scale). crops, tobacco, hand-move) bananas, sugarcane, potatoes. Center Small- and large- Wheat, barley, Possibly bulk water, 0.7–2.2 9–150 ha pivot scale commercial soya, maize, infield pumping. (65 ha per pivot is (broad-scale). groundnuts, typical on farms of sorghum, paprika, 50–5,500 ha) tobacco, sugarcane, rice. Source: ECA and Prorustica (2015). Note: The categories provided in this table are general as no two schemes are identical. a. Assumes an average distance of 300 m from the water source to the irrigation scheme. b. Indicates the system scale commonly seen in Sub-Saharan Africa. Irrigation offers distinct seasonal advantages for crop and farming practices. Despite this, multi-cropping, along production as it can help overcome rainfall variability and with the nearly constant need of water supply for efficient even temperature extremes by maintaining adequate cropping (especially under drip irrigation), does reduce levels of soil moisture year round. In the summer, the seasonal variation to a certain extent. primary advantages are greater reliability of water supply Africa’s grossly underutilized agricultural potential (i.e., reducing the impact if rainfall is less than expected) should be tapped by significantly growing the area under and the ability to plant crops early without waiting for cultivation to cover most economically viable areas. You rains. In the winter, when rains are not expected, irrigation et al. (2009) developed estimates of potential increase is indispensable for cropping, allowing for the production in irrigable area in the region using detailed topographical of wheat and other winter crops and more crop cycles per data and economic parameters (figure 2.2). The study year for rice. Therefore, annual use of irrigation allows found that both large- and small-scale irrigation schemes year-round cropping. can be economically developed in Africa, with economic The extent of irrigation and the associated electricity internal rates of return (IRRs) exceeding 12 percent.2 is likely to be characterized by some amount of seasonal- Investments in irrigation over this cut-off could poten- ity. The magnitude of the seasonal variation in irrigation tially increase irrigated areas by 7.7 million ha, with 5.8 depends on crop choice, weather variations, and irrigation million ha coming from small-scale schemes. 18 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Countries with the greatest potential for large-scale All of these countries have more than 100,000 ha of investment are Ethiopia, Mali, Mozambique, Nigeria, potential, based on existing or projected development of Sudan, Tanzania, Zambia, and Zimbabwe (You et al. 2009). mainly multipurpose water-storage reservoirs. Except for Southern Africa, small-scale irrigation projects in Sub- Saharan Africa are generally estimated to have higher IRR Figure 2.2: Potential new or than large-scale irrigation. This implies that economically rehabilitated irrigable land in Sub- viable, small-scale irrigation projects could increase in Saharan Africa area under irrigation to a greater extent than large-scale 2,000 projects (table 2.2).3 By far, the greatest potential is found in Nigeria, which accounts for more than 2.5 million ha or nearly half of suitable hectares. Such countries as Cameroon, Chad, Irrigation potential (’000 ha) 1,500 Ethiopia, Mali, Niger, South Africa, Sudan, Tanzania, Togo, and Uganda each has at least 100,000 ha of potential. To begin to tap this potential, the CAADP Program 1,000 for Investment in Agricultural Water targets region-wide expansion of the irrigated area by 3 million ha, approxi- mately doubling the current area by 2030 (World Bank 500 2013). In some areas, this expansion could be carried out even more quickly: the World Bank’s proposed Sahel Irrigation Initiative has a goal of “doubling the irrigated 0 areas in Sahel in five years through improved public n rn rn ea policies and increased private-sector involvement.” Much l lia tra he te in he n Gu s ut Ea Ce a of this irrigation would be gravity fed, but some of it, So -S of no lf especially small-scale irrigation, would require pumping da Gu Su for transport and/or extraction. And there is an additional Large-scale Small-scale Rehabilitation synergy: the development of hydroelectric power sources Source: You et al. 2009. can often be combined with irrigation projects. Table 2.2: Potential Investment Needs for Large-Scale, Dam-Based and Complementary Small-Scale Irrigation Schemes in Sub-Saharan Africa Large-scale Irrigation Small-scale Irrigation Increase in Investment Average Increase in Average Irrigated Area Cost IRR Irrigated Area Investment Cost IRR Region (million ha) (million US$)a (%) (million ha) (million US$)a (%) Sudano-Sahelian 0.26 508 14 1.26 4,391 33 East 0.25 482 18 1.08 3,873 28 Gulf of Guinea 0.61 1,188 18 2.61 8,233 22 Central 0.00 4 12 0.30 881 29 Southern 0.23 458 16 0.19 413 13 Indian Ocean Islands 0.00 0.00 n.a. 0.00 0.00 n.a. Total 1.35 2,640 17 5.44 17,790 26 Source: You 2008. Notes: The average value for IRR was weighted by the increase in irrigated area. Benin, Chad, and Madagascar have no profitable, large-scale irrigation; n.a. = not available. a. These estimates are one-time investment costs rather than annualized figures. Power Needs of Agriculture 19 Primary and Secondary Processing Aggregate Electricity Demand from Irrigation and Processing Electricity is a vital input in value-added processing activities, such as post-harvest cleaning and drying to By 2030, we estimate that electricity demand from remove moisture and prevent spoilage (e.g., for cereals agriculture could double from today’s level, reaching and legumes), milling (e.g., of maize, rice, and cassava), about 9 GW. This is a simplified estimate as the varied and crushing. Specific processing activities for high-value nature of product value chains and associated irrigation, agricultural products also rely on electricity inputs (e.g., processing, and storage activities makes it impossible wet-processed coffee using machinery for pulping). to develop a comprehensive, region-wide estimate. The Furthermore, electricity can improve storage of pro- demand emerges from considering the potential increase duce through cold chains, thereby reducing income loss in irrigation and post-harvest activities. Assumptions from spoilage and increasing the ability to specialize in about increased development of irrigation and processing high-value perishable products (e.g., dairy, meats, fruits, potential, unit electricity use, and accompanying growth and vegetables). It is estimated that about 30 percent in crop yields underlie this estimation. Growth in agricul- of agricultural produce is wasted due to spoilage. Cold tural production catering to domestic and export demand storage and drying can reduce this figure substantially. and accompanying movement up the agriculture value Electric fans for air precooling, ice-making machines chain are expected to increase electricity demand from and hydro-coolers can improve cooling efficiency in cold irrigation and post-harvest processing. storage rooms. By 2030, about 3.1 GW in additional electricity Though difficult to estimate accurately due to the demand is expected from the development of irrigation dispersed potential, primary and secondary processing potential across Sub-Saharan Africa (figure 2.3). Given represent a significant growth area in Sub-Saharan Africa. the region’s significant underutilized water resources, The expected demand growth for grain milling is likely along with the ubiquitous benefits from irrigation across to increase significantly (e.g., maize in Nigeria, wheat in most value chains, it is expected that irrigation will Zambia, and rice in Tanzania). Similarly, increased demand account for a significant portion of electricity demand for processing of cassava—a widely produced and con- from the agriculture sector.5 The estimated demand sumed staple in many countries (e.g., Angola, Democratic from irrigation is based on fully exploiting potential areas Republic of the Congo, Mozambique, Nigeria, and for new or rehabilitated irrigable areas, totalling nearly Uganda)—is expected due to its perishable nature and use 6.8 million ha.6 This area is dominated by small-scale as an industrial input in the manufacturing of glue. scheme development in the Gulf of Guinea (with more Additional primary and post-harvest processing (if than 1.5 million ha in Nigeria alone) and rehabilitation of developed to full potential), together with the activities existing schemes in the Sudano-Sahelian region (with over discussed above, could significantly change the rural elec- 1 million ha in Sudan) (table 2.4).7 tricity markets. Table 2.3 summarizes the various activities Figure 2.3 shows that about an additional 1.1 GW is that can serve as anchor loads for rural electrification, expected from the development of the region’s agro-pro- along with the value chains they are part of and examples cessing potential. Power demand from the development of countries where they are present and likely to grow. of agricultural processing activity is based on increased The creation of opportunities for viable rural electri- growth in both primary crop production and the propor- fication on the back of local agricultural development will tion of crops that are processed. Currently, the percent- depend on various site-specific factors, including the scale age of crop production processed through electrified value and profitability of agricultural operations, crop, terrain, chains is quite low (conservatively estimated at 10 per- type of processing activity, and other local conditions. cent). By 2030, this percentage is expected to grow to Rural electrification opportunities will be best served 15 percent as a result of the increased participation of by agro-processing activities that generate electric- small-scale farmers in formal value chains. ity demand close to rural population centers, generate Given the varied nature of processing activities by enough income to cover electricity supply costs, are type, scale, location, and technology, the estimate is based sufficiently large in relation to household demand,4 and on the electricity requirement of a typical processing have relatively low seasonal variation. 20 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 2.3: Key Power-Intensive Agribusiness Activities Value SSA Countries/ Scale of Growth Potential Chains Regions Where Power Demand/ of Value Chain and Activity Supported Activity Occurs Supply Activity New large- Maize, rice, wheat, Most countries Single areas can demand Many areas that can be supported scale irrigation oilseed, sugarcane, tea, > 15 MW of capacity. likely to require farms of > 250 ha; floriculture crop choice depends on market prices. Substitute Maize, rice, cassava, Most countries In unconnected rural Many towns in agricultural areas power for oilseed towns, demand unlikely will have this demand; risky as diesel in small- to exceed 500 kW for anchor load for electrification. scale milling the whole town. New large- Maize, rice, wheat, Most countries Demand can be Widespread opportunity. Reliant scale milling oilseed, sugarcane, oil > 1 MW from a single on base supply from commercial palm, tea, cotton mill. estates; crop choice depends on market prices. Milking and Dairy Few countries > 800 kW peak demand. Small markets in SSA; climatic cold storage conditions not ideal for dairy farming. Cold storage Floriculture, export Ethiopia, Kenya, and 10 MWh/ha per year. Continued demand for floriculture vegetables Uganda (floriculture in Europe, leading to agribusiness and export vegetables); growth in select countries; Rwanda and Tanzania challenges with horticulture (export vegetables) through demand for high quality, competitive retail markets driving down margins and tariff restrictions in European markets. Biomass-fueled Rice, oil palm Many countries (rice); Can provide > 10 MW of Beyond Africa, export market for generation West Africa and East power (ha/ton). rice is challenging and unreliable and Southern Africa for palm oil. Water intensity restricts locations; depends on reliable supply of biomass from commercial estates. Bagasse-fueled Sugarcane Eastern and Southern Can provide > 10 MW of Large market for crop, but generation Africa (South Africa) power (70 kWh/MT of price-dependent. Water intensity sugarcane, or restricts locations; depends on 243 kWh/MT of reliable supply of bagasse from bagasse). commercial estates. Source: ECA and Prorustica (2015). Power Needs of Agriculture 21 Figure 2.3: Estimated electricity demand activity (milling) and thus does not capture the electricity (MW) from agriculture for Sub-Saharan demand from the potential development of other process- Africa in 2030 ing activities or storage. In 2012, the Food and Agriculture Organization of the United Nations (FAO) estimated crop production at about 852 million metric tons (MT). Assuming a growth rate of 2.4 percent annually (Alexandratos and Bruinsma 2084 2012), crop production would reach 1.3 billion MT by 2030 (table 2.5). The power demand from crop produc- tion is estimated by assuming that processed crops will consume, on average, the amount of power needed for an 978 average wheat mill in Zambia—some will have greater con- sumption and others less. This “average mill” is assumed 6915 to handle 8 MT per hour, operating year round at 16 hours per day and 6 days a week. This would result in approxi- 3786 mately 40,000 MT per year and have a power capacity demand of 400 kW. The total estimated electricity demand from agricul- 2015 2030 ture is indicative of the scale of the opportunity for rural Irrigation Processing (milling) electrification to benefit from agricultural growth poten- Source: ECA and Prorustica (2015). tial. The overall magnitude of electricity demand provides Table 2.4: Method for Calculating Power Demand from Irrigation Prominent Countries Estimated Estimated Estimated with Irrigable Area Proportions/ Power Use Power Use Category (thousand ha) Power Use (kW/ha) (MW) Large-scale Ethiopia (191) Much of East and Southern Africa 1.2 kW/ha for 1,285 Nigeria (609) requires bulk-water pumping, West area requiring Sudan (238) Africa less so; 50% requires bulk- bulk water, Zimbabwe (142) water pumping and 50% just infield 0.7 kW/ha Total = 1,352 equipment. otherwise Small-scale Cameroon (170) Most schemes are very basic in riparian 0.7 kW/ha for 1,051 Chad (231), Mali (219) areas; 40% requires power, and 60% is area using power, Nigeria (1,538) entirely gravity fed with no power. 0 otherwise Tanzania (196) Uganda (445) Total = 3,754 Rehabilitation Somalia (135) Most rehabilitation consists of gravity 0.7 kW/ha for 793 Sudan (1,064) fed, colonial-era schemes; 10% is large- area using power, Total = 1,688 scale with bulk water, 30% large-scale 0 otherwise without, 20% small-scale with power, 40% small-scale with no power. Sources: You et al. (2009); ECA and Prorustica (2015). 22 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 2.5: Power Demand for Crop Processing Primary Processed Number of Crop Production Crops Processed Production 400-kW Total Power Year (million MT) (%) (MT) Mills Required Demand (GW) 2012 852 10 85.2 2,129 0.851 2030 1,306 15 196.0 4,893 1.960 Sources: FAO; Alexandratos and Bruinsma (2012); ECA and Prorustica (2015). a sense of the investment in generation capacity that will agriculture sector. The latter informs the likely viability of be required to meet agricultural needs and the addition accounting for agricultural growth in rural electrification to rural electricity demand that is expected, owing to the strategy and planning. endnotes 1. Some sprinklers are pressurized, while others are solely gravity operated. 2. Conditional on having initial investment costs at best-practice levels and if market access, complementary inputs, extension of credit, and a supportive policy and institutional environment are in place. 3. The higher IRR for small-scale irrigation is due to the existence of large amounts high-potential rainfed cultivation located far from large-scale developments that could be profitably converted into small-scale irrigation (You et al. 2009). 4. Although even a relatively small agricultural load can potentially help to push aggregate demand in a given area over the threshold of economic and financial viability. 5. In the context of climate change, the future availability of water will depend critically on improvements in water management practices and planning (box 1.3). World Bank (2016a) predicts that, under business as usual, water management in Southern and East Africa will not experience negative effects on GDP, while other parts of Sub-Saharan Africa could experience about a 6 percent fall in GDP in 2050. 6. You et al. (2009) classifies areas based on their anticipated IRR on irrigation investment. The numbers reported here are based on an anticipated 12 percent return, which is a typical benchmark for such projects. 7. You et al. (2009) was published before the independence of South Sudan and thus classifies the whole of Sudan together. Power Needs in Selected Value Chains Chapter 3 A gricultural production in Sub-Saharan Africa The need for post-harvest electricity input varies, is fairly diversified, and no single cereal crop depending on the nature of the crop, the type of value predominates across the region. In terms chain (or targeted market) and local conditions. A case in of production quantity, maize is the most point is Kenya’s dairy sector: 86 percent of the country’s important, followed by sorghum, millet, and rice; the milk supply is driven by small-scale farmers and small- and importance of each crop varies by individual countries. medium-sized enterprises (SMEs), with milk being sold In West and Central Africa today, cereals comprise less to small-scale vendors. Parallel to this, larger dairy farms than 20 percent of agricultural value added (compared with either integrated dairy herds and/or formal links to to 35 percent for Asia prior to the Green Revolution), dairy farmer cooperatives provide pasteurized milk and with the remainder coming from other staples (especially processed dairy products via cool chains for sale to higher roots and tubers), horticulture, export crops, and livestock income urban consumers through supermarkets (World (Schaffnit-Chatterjee 2014). Bank 2013). Owing, in part, to diversity in agricultural production, This chapter examines potential electricity use along agriculture value chains also vary widely across the region 13 selected value chains. Electricity demand from on-farm and even within countries. Value chains vary by length, activities and rural processing presents an opportunity for technologies utilized, value added, and markets served.1 the development of anchor loads to spur rural electrifica- Many value chains operate in both informal and formal tion. The source of electricity may vary on a case-by-case markets, with the former catering to low-income, domes- basis, and opportunities for biomass based generation tic consumers and the latter catering to higher income for particular value chains (e.g., oil palm and sugar) are urban and export markets (World Bank 2013). highlighted. In addition, bottom-up estimates of potential The value chains for the region’s bulk commodities future electricity demand from the selected value chains (e.g., maize and rice) are primarily informal, in contrast to are presented. more market-oriented, semi-processed and consumption ready products. As a commodity moves along the value chain to the ultimate market and consumer, hygiene and Selection of Value Chains quality standards become more stringent. Such commod- ities as sugar, tea, and oil palm are processed virtually at The value chains selected for this study help illustrate the the point of primary production, while other commodities nature of electricity demand from the rural agriculture (e.g., fruits, vegetables, and livestock products) must be and agribusiness sectors, along with the power-demand processed within a relatively short period before they profile. These value chains represent both high growth deteriorate. Still others have parallel value chains; that potential and the ability to create electricity demand is, for the same commodity, some value chains focus on for irrigation and/or processing in rural areas (table 3.1). lower end consumers in domestic markets, while others The potential for agricultural electricity demand extends are more formal, with strong processing and stringent well beyond the value chains discussed here and is often quality control. driven by site- and country-specific factors that create 23 24 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 3.1: Analysis of Commodity Value Chains, by Scale and Region/Country Commodity Scale (if applicable)a Region/Country Maize Small and large East and Southern Africa Rice Small and large Tanzania (primarily) Cassava Small West, Central, East, and Southern Africa Wheat Large Southern Africa Oilseed Small (primarily) East and Southern Africa Horticulture (pineapple) Small and large West, Central, and Southern Africa Sugarcane Small and large East and Southern Africa Oil palm Small and large West and Central Africa Dairy Small and large Kenya Poultry Large East and Southern Africa Tea Large East and Southern Africa Floriculture (roses) Large East Africa Cotton Small West, East, and Southern Africa Source: FAOSTAT (http://faostat3.fao.org). a. Farming systems are defined in terms of labor type and not merely scale. Large-scale commercial farming is defined by family labor that is predominantly managerial, with full-time labor hired for specific tasks and production catering to market supply. opportunities along other crop and processing activities. large existing potential on the extensive (area expansion) The case studies presented in chapter 4 analyze examples and intensive (yield growth) margins. of such opportunities. According to future production estimates, cassava The commodity value chains shown in table 3.1 were and maize—primary staple food crops in the region—will selected according to the following criteria. Starting with remain dominant over the period until 2030. Sugarcane, a the top 20 commodities by production value for 2012 well-established industry with conducive growth con- (from FAOSTAT), the list was modified to assure the inclu- ditions, is also expected to remain dominant across the sion of (i) key export commodities (e.g., tea, cotton, and region for the foreseeable future. In addition, recent high horticulture); (ii) value chains based on assessed electricity growth rates of cotton, pineapple, and rice suggest that use; (iii) commodities with large production volume and these commodities will likely gain greater regional impor- importance for local food markets with potential for future tance in the coming decades. growth in processing requirements (e.g., cassava and maize); Cassava. In terms of production quantity, cassava is (iv) commodities that figure in the top ones by value for Sub-Saharan Africa’s most important crop, accounting for many countries in the region (e.g., tea and soybean), which more than half of global production. Nigeria is the leading may not appear on a region-wide list; (v) commodities global producer, followed by the Democratic Republic of with large irrigation schemes (e.g., irrigated wheat); and the Congo (DRC), Angola, Ghana, and Malawi.2 Cassava (vi) value chains with the potential to supply fuel for elec- is experiencing growing demand as a staple food crop and tricity generation (e.g., oil palm and sugarcane). an intermediate input into various other commercial value Table 3.2 shows the estimated production volume chains (e.g., starch and livestock feed). The crop is still for the selected commodities in 2030, along with their mainly grown under small-scale farming conditions with estimated average annual growth rates between 2013 limited use of irrigation. Owing to its drought tolerance and 2030. Future projections are calculated using the and ability to grow in relatively poor soils, production is historical growth rate (between 2009 and 2013) for fairly widespread in rural areas across the region. Further each commodity (FAOSTAT) and applying a concavity development to make the crop’s value chain more market parameter to project a declining growth rate over time. oriented can have large effects on the livelihoods of small The assumed growth rates are qualitatively more conser- farmers. Growth in cassava production depends critically vative than those assumed by Alexandratos and Bruinsma on improved processing and drying of roots to reduce bulk (2012), who predict mostly convex growth rates, owing to and prevent deterioration. Power Needs in Selected Value Chains25 Table 3.2: Comparison of Historical and Projected Commodity Growth Rates and Estimated Production Assumed Average Growth Rate, Annual Growth, Estimated Production Projected Production 2009–13 2013–30 in 2013 in 2030 Commodity (%) (%) (million MT) (million MT) Cassava 6.4 2.8 157.7 252.7 Maize 5.8 2.5 65 101.2 Sugarcane 1.7 0.8 73.9 84.6 Rice (paddy) 5.9 2.6 22.6 35.5 Wheat 5.1 2.3 7.1 10.6 Pineapple 9.5 4.2 4.4 9 Dairy 1.6 0.7 3.2 3.6 Poultry 1.5 0.6 2.7 3 Cotton (lint) 8.1 3.5 1.3 2.5 Oil palm −0.7 −0.3a 2.4 2.2 Tea 5.4 2.4 0.7 1 Oilseed (soybean) 2.6 1.2 0.5 0.6 Sources: FAOSTAT and World Bank estimates. a. The oil palm industry is now considered less attractive; some developments are proving unsustainable and are being converted to other uses. Maize. Due to its tolerance of diverse climates, South Africa and Mozambique lead in terms of area under maize is one of the world’s most widely grown crops. In cultivation (table 3.3). Eighty percent of the world’s sugar 2013, total global production was estimated at more than is produced from sugarcane, while the other 20 percent 1 billion metric tons (MT). In Sub-Saharan Africa, maize is from sugar beet (FAO 2009). The most common pro- is one of the most prevalent cereals, with more than duction model is contracting commercial and small-scale 65 million MT produced in 2013 (table 3.2). However, the outgrowers to supply the sugar refineries. region’s average yield of 1.4 MT per ha is low compared Rice (paddy). Sub-Saharan Africa has witnessed to the global average of 5 MT per ha, and 11.6 MT per ha rapid growth in rice production, driven mainly by urban- in the United States (Iowa) (2009 figures, FAO). A few ization. The compound annual growth rate (CAGR) of countries are dominant in maize production, but their domestic production has averaged about 6 percent, with market share is less pronounced. Maize’s utilization is wide more than 22 million MT reached in 2013. According to ranging; it serves as a leading food staple and important the Africa Rice Center’s analysis, the region’s rice yields feed crop, as well as an input in the processing of food, have increased in real terms by an average of 108 kg chemicals, and fuels (ethanol).3 In East and Southern per ha annually, comparable to the Green Revolution’s Africa, maize is principally a food staple, accounting for growth rates in Asia (Seck et al. 2013). Despite such rapid 30−50 percent of low-income household expenditure.4 growth, rice imports have also increased significantly; As such, growth in production is expected to increase, in 2012, 12 million MT were imported. The region has propelled by growing regional demand. considerable potential for production growth through Sugarcane. According to the FAO, sugarcane is the increasing the area under cultivation and increasing yields. world’s largest crop in terms of production quantity, with Wheat. Among all cereals, wheat is the most highly 1.83 billion MT produced in 2012. Brazil is its largest traded. As of 2013, it was the world’s third most widely pro- producer, followed by India. Sub-Saharan Africa accounts duced cereal (behind maize and rice), at a total of 713 mil- for roughly 4–5 percent of global production, with about lion MT.5 In Sub-Saharan Africa, Ethiopia and South Africa 74 million MT produced in 2013. The region’s largest are the main wheat producers. Generally, production has producers are South Africa, followed by Sudan and Kenya; not kept pace with the region’s growing demand for wheat; 26 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 3.3: Countries in Sub-Saharan Africa with Similar Commodity Production and Processing Systems Commodity Countries Maize Kenya, Malawi, Mozambique, Tanzania, Zambia, and Zimbabwe; also Burkina Faso, Ghana, Mali, and Nigeria (but not at such large commercial volumes) Rice Madagascar and Tanzania Small-scale cassava Angola, DRC, Mozambique, Nigeria, Tanzania, and Zambia Irrigated wheat Zambia and Zimbabwe Rainfed wheat Ethiopia and Kenya Commercial soya Zambia and Zimbabwe Sugarcane Ethiopia, Kenya, Malawi, Mozambique, South Africa, Swaziland, Tanzania, and Zimbabwe Oil palm Cameroon, Côte d’Ivoire, and Ghana Dairy Kenya, Ethiopia, Rwanda, South Sudan, and Uganda Poultry Kenya, Malawi, Zambia, and Zimbabwe Tea Kenya, Malawi, Rwanda, and Uganda Floriculture (roses) Ethiopia, Kenya, Tanzania, Uganda, Zambia, and Zimbabwe Cotton Benin, Burkina Faso, Côte d’Ivoire, Mali, Mozambique, Tanzania, Uganda, Zambia, and Zimbabwe Source: ECA and Prorustica (2015). thus, wheat imports have been on the rise. Among the a minimum, suggesting that dairy storage and processing region’s handful of countries that are fully self-sufficient centers are located in the vicinity of dairy farms. in wheat production, Zambia is noteworthy; that coun- Poultry. Population growth, changing diets resulting try’s annual production, mainly commercial in scale, totals from urbanization, and income growth are the major 300,000 MT (table 3.3).6 Many parts of East, Southern, drivers of Sub-Saharan Africa’s ongoing demand for and Central Africa are suitable for wheat production. poultry. During 2000–11, poultry (meat) production Pineapple. In Africa, horticulture, in the form of trop- across the African continent grew by 5 percent per year, ical fruit production, caters mainly to own consumption reaching 4.62 million MT in 2011. Major producers are and domestic markets; in some countries, it also caters to in Northern Africa: Egypt, Algeria, Morocco, Libya, and Europe and other export markets (e.g., canned fruits and Tunisia. In Sub-Saharan Africa, 2013 production totaled pulp). After banana, pineapple is Sub-Saharan Africa’s 2.75 million MT, with South Africa and Nigeria as lead most important tropical fruit. Nigeria is the region’s producers. These two countries are also the region’s major largest pineapple producer. Kenya, the second largest, egg producers; and hatcheries are usually large-scale com- ranks among the world’s top five exporters of pineapple; mercial operations. Modern poultry complexes are usually canned pineapple, exported mainly to Europe, is its largest integrated with chicken farms to reduce the costs associ- manufactured export. ated with the transport of live animals. Contract farmers Dairy. The robust growth in dairy production reported receive chicks from the hatchery, ideally housing them in many parts of Sub-Saharan Africa today is being driven in climate-controlled chicken houses. Broiler processing by economic growth and urbanization. Traditionally, milk operations are typically located on-site at poultry farms. has been produced for own consumption or local con- Cotton (lint). Cotton is one of Africa’s main cash sumption by farmers; however, growing urban demand crops among small-scale farmers. In 2013, Sub-Saharan is increasing the need for cold supply chains to maintain Africa produced 1.3 MT of cotton (lint) (table 3.2). The product quality. According to the FAO, the region’s dairy region’s major producers are Burkina Faso, Mali, Côte production totaled 3.2 million MT in 2013. Along with d’Ivoire, Benin, and Zimbabwe. In West Africa, Burkina this demand growth is the demand created for process- Faso and Mali each produce about 400,000 MT per year. ing milk-derivative products (e.g., cheese, butter, and In East and Southern Africa, Zimbabwe is the lead pro- evaporated milk). Transport of raw milk, which is prone ducer, with an annual output of 200,000–300,000 MT to spoilage, is generally uneconomical; thus, it is kept to in seed cotton (table 3.3). Power Needs in Selected Value Chains27 Oil palm. The source of palm oil, one of the world’s high-value agricultural activities, generating revenues of leading edible vegetable oils, oil palm constitutes 60 per- $100,000–200,000 per ha.9 cent of the global trade in vegetable oils (World Bank 2011a). Oil palm fruit yields two distinct types of oils: (i) palm oil, which is edible, used mainly in the form of Electricity Demand vegetable oil and (ii) palm kernel oil, which is extracted and Farming Scale from the seed kernel, used as an input to process other foods (e.g., biscuits and margarine), manufacture house- Electricity demand along the value chain is likely to vary hold products (e.g., soap, shampoo, and cosmetics), and by scale or type of farming operations (e.g., commercial produce biodiesel fuel. Southeast Asia (mainly Malaysia versus small-scale) due to differences in farming processes and Indonesia) produces 85 percent of the world’s palm (e.g., irrigation) and the extent and nature of post-harvest oil. In Sub-Saharan Africa, West Africa is the main processing (box 3.1). While farming in Sub-Saharan Africa producer. Nigeria is the largest producer; however, Côte is predominantly in the form of smallholder agriculture, a d’Ivoire, DRC, Ghana, Guinea, and Uganda are also estab- significant portion of the future potential rests on increas- lishing major operations. While commercial-scale farmers ing yields on such farms by employing more modern account for most production, small-scale farmers also inputs and connecting them to higher value markets and find oil palm an attractive crop since it is relatively high value chains (i.e., employing large-scale operations). yielding and requires limited labor inputs. It is useful to compare electricity needs across these Tea. Tea is one of Sub-Saharan Africa’s most impor- types of agricultural arrangements. The implication for tant export commodities, especially for East Africa. Kenya overall magnitude depends on the evolving proportions is the world’s largest exporter of black tea. In 2011, it of commercial and small-scale farming techniques in the produced 378,000 MT, about two-thirds of Sub-Saharan Africa’s output. Uganda and Malawi are the region’s next two largest producers, while Tanzania and Rwanda are experiencing steady growth in production (table 3.3).7 Box 3.1: Farm Type Definitions Tea-growing usually occurs on large plantations, with processing located either on-site or nearby. Defining farming systems in terms of labor can be Oilseed (soybean). Although Sub-Saharan Africa’s useful, given that the definitions do not depend on soybean production is fairly small by global standards, production scale or crop type. Accordingly, three contributing only 1 percent of global production, the types of farm systems are distinguished here: region’s production is growing faster than the world aver- age (ACET 2013). South Africa has the highest growth in Family farms. These small-scale farms are char- percentage terms, while Nigeria has the largest absolute acterized by the predominant use of family labor, growth.8 Soybean is grown mainly on small farms, while lack of permanent workers, and presence of sea- commercial soybean farming is prevalent in South Africa, sonal labor hired during peak production times. Zambia, and Zimbabwe. Soybean is sold for both human Small investor farms. The owners/family mem- consumption and as an animal feedstock. bers are involved primarily in management and Floriculture (roses). The introduction of rose supervisory roles, while the bulk of labor input is cultivation in Sub-Saharan Africa began in Kenya provided by hired farm workers; this group is less about three decades ago. To this day, Kenya remains defined in Africa, but most, if not all, of their well-­ the region’s main producer and exporter of roses; that crops are produced for market. country also has the highest area under rose cultiva- tion, followed by Ethiopia and Uganda. Rose production Large-scale commercial farms. Family labor for in Ethiopia has been growing rapidly, and the country these farms is exclusively or predominantly mana- is fast establishing itself as a major exporter, to some gerial. A permanent hired staff of full-time work- extent capturing market share from Kenya. Most pro- ers, specialized to a certain degree (e.g., drivers), duction is for export markets, especially Europe, which produces primarily for market. generates more than US$1 billion in export revenues for the region (International Trade Center 2014). On Source: Poulton et al. (2008). a per hectare basis, rose production is one of the most 28 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa region. For example, greater proportional growth in the operating as outgrowers for commercial estates; thus, the adoption of commercial-scale farming, which depends scale of power demand cannot be viewed independent of more heavily on power input, will induce higher overall larger commercial estates.13 The figures include dairy with electricity demand by the agriculture sector. zero values to highlight that informal dairy value chains do Examining typical electricity use for irrigation and not utilize power in Sub-Saharan Africa. processing shows that, for most of the value chains Given the economies of scale in generation capacity, analyzed, irrigation constitutes a large proportion of the commercial agricultural activities are likely to be more potential electricity demand. As small-scale farming financially viable anchor loads to support affordable rural largely relies on rainfed or gravity irrigation, electricity electricity supply to rural Sub-Saharan Africa. However, demand from commercial-scale irrigated agriculture is due to recent technological improvements, accompa- an order of magnitude greater than from smallholder nied by the creation of enabling regulatory conditions, agriculture. Figure 3.1 compares typical rates of power electricity provision in the form of mini-, micro-, and even usage for large-scale irrigated and small-scale rainfed (or pico-grids has dampened the scale economies in electric- gravity fed) irrigation for selected value chains. For the ity generation and distribution investments. Increasingly, most widely grown crops in Sub-Saharan Africa, including advances in renewable energy technologies, such as solar maize, rice, and cassava, irrigation accounts for the highest photovoltaics (PV), are allowing viable electricity infra- potential electricity load.10 structure investments catering to smallholder agriculture As shown, potential peak power loads for small-scale and rural households. Even for more conventional tech- informal production are quite small relative to loads from nologies, ubiquitous small-scale, informal agriculture can commercially irrigated production on a per unit basis enhance the viability of rural electrification on the margin. (figure 3.1b), although this is partly offset by the predom- As discussed earlier, given the diversity of conditions inance of smallholder agriculture across the region, repre- across agricultural areas, site-specific opportunities still senting over 80 percent of the cultivated area (Livingston, exist if cost-effective technologies (e.g., biomass, solar, or Schonberger, and Delaney 2011). small hydro), which may not exhibit strong economies of Though irrigation accounts for a major part of the scale in installed capacity, can be utilized. potential on a per unit basis, post-harvest processing can play a significant role in supporting rural electrification, especially in the case of some commodity value chains. Electricity Demand in the Selected Adding electricity demand for processing to that for Value Chains irrigation, commercially oriented value chains such as sug- arcane, tea, floriculture, and dairy have the overall highest The development of power profiles for each commodity, potential electricity demand (figure 3.1a). Tea is easily the region, and farm type utilized a range of information most power-intensive commodity, with demand ema- sources. Value chains were analyzed in terms of their nating primarily from processing (figure 3.1c).11 Activities nature and magnitude of power use for irrigation and with potentially large loads from processing (sugarcane, processing, growth potential, and ability to serve as an tea, and floriculture) are developed and operated mainly anchor load. by large single entities or organized groups of small-scale To enable comparison, the power profiles presented farmers (see case study 6, chapter 4).12 In such cases, the below are for (arbitrary) standardized farm sizes of power load and potential power supply are usually part of 300 ha, based on the unit electricity demand presented in the planning process; examining options and incentives for table 3.4. The 300 ha benchmark was chosen to rep- rural electrification can be integrated into the planning resent the cultivated area that might constitute a typical stage itself. project site.14 However, in Sub-Saharan Africa most agricultural pro- Maize. For the maize value chain, the input of rural duction occurs in small-scale, informal value chains. The electricity is primarily for irrigation (largely restricted to potential power demand from small-scale agriculture is large-scale farming) and milling (figure 3.2a). The gain in much less than from commercial agriculture. Lower yields value from electricity use comes from the higher yields mean that a larger area is required to produce sufficient resulting from irrigation and the saving of labor and higher production volume for processing facilities. Figures 3.1b productivity resulting from electricity powered (versus and 3.1d exclude small-scale sugarcane, oil palm, and tea manual) milling. The estimated electricity demand from since these typically occur only with small-scale farmers these two activities is about 1.17 kW per ha for large-scale Power Needs in Selected Value Chains29 Figure 3.1: Potential peak capacity and energy demand for large- and small-scale systems a. Peak capacity: Large-scale irrigated production b. Peak capacity: Small-scale rainfed production 2.5 0.12 2.0 0.09 1.5 kW/ha kW/ha 0.06 1.0 0.03 0.5 0.0 0.00 O ne e ce Ca t Pi eeds ga s alm Co a e iry y O va ric n n e va ce ds at a ple Te aiz ur ir aiz o to he ssa Ri ssa rca t Ri Da he Da ee ult t t il p M M ap W Co ils ils W Ca ne O Su Flo Irrigation Processing    Irrigation Processing c. Energy demand: Large-scale irrigated production d. Energy demand: Small-scale rainfed production 1,000 120 750 90 kWh/MT kWh/MT 500 60 250 30 0 0 O ne e ce Ca t Pi eeds ga s alm Co a e ir y y O va Flo tton n e va ce ds t a ple Te aiz ur ir aiz to a he ssa Ri ssa rca Ri Da he Da ee ult t il p M M ap W Co ils ils W Ca ric ne O Su Irrigation Processing    Irrigation Processing Note: Unit electricity demands are constructed from various sources and field observations by ECA and Prorustica. Figure 3.1a does not plot poultry as it is a significant outlier and not feasible to depict on the same scale. Figure 3.1c omits floriculture due to the incomparability of yield data. Figures 3.1b and 3.1d are restricted to those commodities with significant production on smallholder farms (thus omitting such cash crops as tea, sugarcane, floriculture, and horticulture). 30 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2a: Electricity input in the maize value chain Irrigation Drying Crushing/milling • Electricity for pumping water • Mostly solar energy • Milling in town centers as and drip irrigation • Potential for use in certain scale is required green-rated heat from CHP generator production and about 0.77 kW per ha for small-scale and milling is 1.04 kW per ha for large-scale, irrigated irrigated production, suggesting that 300 ha of cultivated production and 0.03 kW per ha for small-scale (paddy) maize will require about 250–350 kW of installed power production with no irrigation. Thus, for a cultivated area generation capacity. of 300 ha, the power demand is in a range of 9–315 kW, Rice. For rice, irrigation and milling are the primary depending on farming type. Additionally, rice husk bio- sources of rural electricity demand (figure 3.2b). Because mass provides a readily available and cost-effective fuel rice can be grown under a variety of irrigated or rainfed source to generate electricity to supply mills and poten- water regimes, electricity demand for irrigation varies tially the neighboring community.15 by type of cultivation. The value gain from electricity Cassava. For cassava, the electricity demand ranges from use is from the higher yields resulting from irrigation (an 0.02 kW per ha to 0.56 kW per ha, depending on whether increase of up to 4 MT per ha) and the value added from the land is under irrigation (figure 3.2c). For a 300 ha culti- milling. The estimated electricity demand from irrigation vated area, the power demand would be about 160 kW. Figure 3.2b: Electricity input in the rice value chain Irrigation Drying/dehusking Milling/polishing • Electricity for pumping water • Mostly manual methods and • Milling in town centers as and drip irrigation solar energy certain scale is required • Potential for use in heating from CHP generator Figure 3.2c: Electricity input in the cassava value chain Irrigation Drying/peeling/chipping Grating/milling • Electricity for pumping water • Mostly manual methods and • Milling in town centers as and drip irrigation solar energy certain scale is required • Potential for use in heating from CHP generator Processing chain: High-quality cassava flour Washing Peeling Washing Grating Pressing Flash drying Milling Power Needs in Selected Value Chains31 Wheat. For winter wheat production, powered demand for electricity. Irrigation for other horticultural activities include irrigation; on-farm drying, cleaning, and crops (e.g., beans, peas, and potatoes) is fairly limited and conveying in and out of silos; and milling (figure 3.2d). usually small in scale. Owing to perishability, electricity is The value added from electricity use is through the higher needed for cooling and to power a cold chain from farm yields from irrigation (an increase of about 4 MT per to market, although this is usually provided in the form of ha) and electric milling and processing. The total power mobile refrigeration units (reefers). The value added from demand from irrigation and post-harvest processing is electricity use in the pineapple value chain includes higher estimated at 1.1 kW per ha for large-scale production and yields resulting from irrigation, increased product value 0.52 kW per ha for small-scale production. For a 300 ha resulting from juicing and canning, and reduced wastage cultivated area, power demand would be in a range of due to cold storage (figure 3.2f).16 The electricity demand 150–230 kW, depending on the farming type. from irrigation is estimated at 0.75 kW per ha for com- Oilseed (soybean). For soybean, the value added from mercial production, implying that 225 kW of power would electricity use occurs through the higher yields made be needed for 300 ha cultivated area. In addition, the possible by irrigation and increase in value from processing by-products of post-harvest processing can potentially (figure 3.2e). The total electricity demand resulting from provide biomass for electricity and heat generation, which irrigation and milling is estimated at 1.04 kW per ha for can significantly reduce power costs.17 large-scale production and 0.64 kW per ha for small-­ scale Sugarcane. Sugarcane yields are highly responsive to production. These figures suggest power demand in a irrigation; thus, water pumping for irrigation is an impor- range of 200–300 kW for a 300 ha cultivated area. tant source of electricity demand in the sugarcane value Horticulture (pineapple). Along the pineapple value chain. In addition, sugar mills constitute considerable chain, juicing and canning activities comprise the main processing demand for electricity (figure 3.2g). The value Figure 3.2d: Electricity input in the wheat value chain Irrigation Drying/cleaning Grinding/milling Figure 3.2e: Electricity input in the soybean value chain Pressing/expelling/ Irrigation Shelling/dehusking Grinding/milling extruding • Electricity for pumping water and • Cleaning/washing, drying, and • Includes heating and grinding • Extraction, using oil expellers drip irrigation storage usually done prior to • Oil would need further processing shelling • By-product is cake used as • Solar power generally used for poultry feed drying Figure 3.2f: Electricity input in the pineapple value chain Irrigation Cutting/juicing Treating/packaging • Electricity for pumping water • Electric machines used for • Thermal treatment and cooling and drip irrigation slicing and juice extraction • Packing and canning and concentration 32 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2g: Electricity input in the sugarcane value chain Irrigation Milling Refining • Electricity for pumping water • Milling: washing, chopping, • Further refining of raw sugar and drip irrigation shredding, and crushing to produced from milling extract cane juice • Usually located near urban • Subsequent clarification, markets concentration, and crystallization to produce mill-white • Biomass by-product used for electricity and heat generation gains from electricity use are derived from the higher used to power factories, with the surplus power exported yields from electricity powered irrigation and the price to the national grid. For both the South Africa sugar differential between raw cane and partially processed industry and Uganda’s Kinyara sugar manufacturer, the sugar. The increased yields from irrigation could reach power output is approximately 30 kWh per MT of crushed 50 MT per ha and even up to 150–200 MT per ha if the sugarcane. latest drip irrigation methods are utilized. On top of the Oil palm. The processing of oil palm usually occurs on value added, maintaining processing activities close to the or nearby the farm due to its bulky nature and ability to farm helps to reduce transport costs. The combined power produce biomass used to generate the heat and electric- demand of irrigation and refining is estimated at 1.81 kW ity required for oil extraction and processing. Oil palm per ha for large-scale production and 1 kW per ha for irrigation is largely rainfed. The main sources of electricity small-scale production. These figures imply that a 300 ha demand are oil processing and extraction from the fresh cultivated area will demand 300–550 kW of power, fruit bunches (FFBs) (figure 3.2h). Though uncommon, depending on the scale of production and related farming drip irrigation can raise yields by 6 MT of FFB per ha. The practices. value gained from using electricity is through processing The biomass residue (bagasse) from sugarcane and reduced transport costs. For milling, the estimated processing has a high potential to generate electricity. electricity demand is 0.02 kW per ha, suggesting a Refineries often produce their own electricity and sell the 6 kW power requirement for a 300 ha cultivated area. excess to the grid. Bagasse generated electricity could Substantial amounts of solid palm oil waste are available become important for the rural populations of sugarcane from the palm oil mills, which are energy self-sufficient; producing nations. For example, in Ethiopia, the Wonchi, that is, they produce their own energy to operate and Metehera, and Finchaa sugar factories produce approx- use the surplus generated to supply estates, sell to the imately 300,000 tons of sugar each year, powering an grid, and possibly sell to villages and towns in the area installed electricity capacity of 62 MW. The electricity is (box 3.2).18 Figure 3.2h: Electricity input in the oil palm value chain Irrigation Oil extraction Refining • Electricity for pumping • Sterilization, stripping, • Refining of extracted crude oil water and drip irrigation digesting, and pressing used • Not necessarily nearby oil palm • Electricity-powered to extract oil extracted from plantations irrigation uncommon in the FFBs Sub-Saharan Africa Power Needs in Selected Value Chains33 processing milk-based products (e.g., butter, cheese, Box 3.2: Palm Oil and Power and evaporated milk). The value gain from electricity use Integration in Uganda results from reduced spoilage due to cold storage,19 the ability to access urban markets, and the value added from One example of an integrated palm oil/power processing milk products. For large-scale operations, the setup is Uganda’s Bugala Power Station, a 1.5 MW estimated power demand is about 0.61 kW per ha. Animal biodiesel-fired thermal power plant located on manure from dairy farms may also be used to generate Bugala Island on Lake Victoria. The power station electricity. is integrated with the palm oil processing plant Poultry. Hatcheries are usually relatively large-scale owned by Bidco Oil Refineries Ltd., which also commercial operations that require electricity input for owns a 6,500 ha palm oil plantation on Bugala a host of processes, including egg incubation and clean- Island. The oil-processing factory generates heat ing. For poultry (meat) production, processing plants use through biomass incineration, used to supply electricity to power conveyor belts, cooling and heating, superheated steam to help extract oil and also and cutting (figure 3.2j). The value added from electricity turn turbines and create electricity in the process. use results from reduced spoilage, increased egg-laying The electricity is used inside the factory, with any productivity, higher labor productivity, value addition from excess sold to neighboring towns. processing, and ability to supply higher value urban mar- kets. The estimated energy demand for ­ commercial-scale broilers (meat) and layers (eggs) is 75 kW per ha each. Dairy. Dairy production systems can potentially A typical 1–2 ha operation would generate a demand create significant electricity demand in rural areas where of about 150 kW (300 kW if the two operations are there are commercial milk producers or cooperatives. co-located). The main source of rural electricity demand from dairy Tea. For the tea value chain, electricity demand is production is cold storage, and machines for electric- from irrigation and processing activities. Irrigation is ity powered milking are also becoming more prevalent mainly rainfed since most tea is grown in areas with abun- (figure 3.2i). Another potential source is machinery for dant rainfall. Even so, there is a considerable potential Figure 3.2i: Electricity input in the dairy value chain Milking Cold storage Pasteurization • Power-driven milking • Individual solar chillers might • Requires heating machines usually used for be an option for smaller- • Centrifuging and dehydration medium- and large-scale scale dairy farmers may be required for other systems derivative products (e.g., cream and dry milk powder) Figure 3.2j: Electricity input in the poultry value chain Egg or meat Incubation Temperature control processing • Temperature-controlled • In Sub-Saharan Africa, • Electricity is generally used to egg incubators temperature controlled, power refrigeration, conveyor poultry layer houses usually belts, lighting, air conditioning, require cooling rather than pumps, compressed air, and heating (apart from egg other mechanical drives incubation) 34 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.2k: Electricity input in the tea value chain Irrigation Shredding/rolling Fermenting/drying • Most tea estates are rain-fed, • Weathering required prior to • Electricity is generally used to but some use supplementary shredding power refrigeration, conveyor irrigation • Cutting, tearing, and curling belts, lighting, air conditioning, (CTC) uses electricity pumps, compressed air, and other mechanical drives Figure 3.2l: Electricity input in the floriculture (roses) value chain Irrigation Cooling • Accounts for about half of • Uses about 35 percent of the the energy consumed electricity consumed • The remainder is used for general facility needs, lighting, and other purposes value gain from irrigation (i.e., increased yields of up Floriculture (roses). In Sub-Saharan Africa, roses to 8 times from sprinkler irrigation and up to 16 times are cultivated mainly in large-scale greenhouses, and from drip irrigation) (figure 3.2k). Thus, the value gain most power demand is from irrigation and cold storage from electricity use results from both increased yields in (figure 3.2l). Electricity is usually sourced through diesel response to irrigation and the value addition from process- generation sets. All farms have on-site cold storage, and ing (including reduced transport and spoilage costs). In growing is done in temperature controlled environments. Sub-Saharan Africa, there is considerable potential for tea For large-scale production, power demand is estimated at producers to gain from increasing yields and moving fur- 2.37 kW per ha, with irrigation accounting for nearly half ther up the processing value chain. In Kenya, 88 percent of energy consumption; thus, a 300 ha cultivated area of tea production is exported raw in bulk; but in Rwanda can be expected to have about 700 kW of power demand. and Uganda, processing is rising. Electricity demand from Cotton (lint). For cotton (lint) production, electricity tea cultivation and processing is estimated at 1.91 kW powered irrigation is not prevalent. Rather, electric power per ha for large-scale plantations and 0.51 kW per ha for is used mainly for seed crushing and ginning (figure 3.2m). small-scale, rainfed facilities. For a 300 ha cultivated area, Due to perishability, cotton ginning must be done soon power demand is in a range of 150–575 kW, depending after harvest. Gins are usually located near reliable power on the scale of cultivation and associated farming and sources in rural and peri-urban towns. Moving ginning post-harvest practices. closer to farms would save on transport costs and possible Figure 3.2m: Electricity input in the cotton (lint) value chain Textile Irrigation Ginning Oil extraction manufacturing • Mostly does not use • The process of separating • Oil presses, expellers used to • Ginned cotton, spun into electricity powered irrigation cotton fibers from the seeds extract oil from seeds yarn, enters various textile value chains Power Needs in Selected Value Chains35 Table 3.4: Power Demand for Standard 300 ha Cultivated Area Per Unit Total Electricity Capacity Electricity Capacity Required for 300 ha   (kW/ha) for Irrigation and Processing Cultivated Area (kW) Agricultural Commodity Small-scale Large-scale Small-scale Large-scale Maize 0.77 1.17 230 350 Rice 0.03 1.04 9 312 Wheat 0.52 1.10 156 330 Cassavaa 0.56 168 Oilseed (soybean) 0.64 1.04 192 312 Horticulture (pineapple)b 0.75 225 Sugarcane 1.00 1.81 300 543 Oil palmb 0.02 6 Teac 0.51 1.91 153 573 Cotton (lint)b 0.03 0.03 9 9 Floriculture (roses)b 2.37 711 Poultryb 75.00 22,500 Dairyb 0.61 183 Note: Choice of the 300 ha benchmark reflects the amount of cultivated area that may constitute a typical project site. For example, this would amount to 300 households, each having 1 ha of landholdings. While this benchmark is somewhat arbitrary (i.e., project sites are likely to have a variety of crops under cultivation), it can be used to construct back-of-the-envelope estimates on electricity demand from the value chains presented. a. Cassava is small-scale only. b. Horticulture (pineapple), oil palm, cotton (lint), floriculture (roses), poultry, and dairy do not use electricity for small-scale operations or are only large-scale operations. c. Small-scale tea cultivation uses rainfed irrigation. spoilage. Cottonseed crushing is done to produce cotton- is considerable. For small-scale production, potential elec- seed oil (used in some instances as a biofuel for vehicles) tricity demand ranges from 9 kW for rice or cotton (lint) and livestock feed. The power demand from cotton to 300 kW for sugarcane. For large-scale production, it cultivation and processing is estimated at 0.03 kW per ha ranges from 6 kW for oil palm to 711 kW for floriculture for both large- and small-scale farming production. This (roses); poultry is an outlier, at 22.5 MW. These estimates implies that a 300 ha cultivated area will have about 9 kW are useful for considering whether the economics of these in power demand. values chains make them viable anchor loads for rural For each of the 13 selected value chains, table 3.4 electrification. summarizes the estimated electricity demand for a Using the forecasted production for the 13 value 300 ha cultivated area and the per-hectare electricity chains presented in table 3.2, along with the constructed demand estimates from irrigation and processing. The unit unit electricity demand for each commodity, a bottom-up estimates show that per-hectare electricity demand is estimate of the total increase in demand for electricity largest for poultry by far, followed by floriculture, tea, and stemming from the selected value chains can be con- sugarcane. The potential per-hectare demand for poultry structed. The calculations show that electricity demand (meat) is considerably higher because the process is much could increase by 2 GW (from 3.9 GW in 2013 to 6 GW more intensive, using less land for a much larger yield. in 2030). This figure represents nearly half of the total The higher per-hectare demand estimates for large-scale potential increase in electricity demand from agriculture production mainly reflects the use of commercial-scale calculated for Sub-Saharan Africa in chapter 2 (4.2 GW). irrigation and the power input required to process large To the extent that the value chains selected represent yields. The range of values for the 300 ha cultivated area the best potential of the agriculture and agribusiness 36 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 3.3: Potential power demand in sectors in Sub-Saharan Africa, the estimated electric- 2030 from processing for small-scale ity demand provides a good indication of the possible agriculture, by selected value chains electricity-agriculture synergies (figure 3.3). The required underlying assumption is the percentage of irrigated Tea, 0.7% and processed production. Clearly, even by 2030, not Dairy, 5.6% Oilseed, 0.5% Pineapple, 0.3% Poultry, 0.3% all production is likely to be cultivated on irrigated land Wheat, 5.9% or processed using electricity driven machinery. With little detailed data available on irrigation and processing proportions by value chain, this study makes conserva- tive assumptions for each of the value chains considered: Cassava, 19.7% only 15 percent of the land is assumed to be irrigated and 15 percent of crops are assumed to be processed.20 Rice (paddy), 26.4% Maize, 37.2% Sugarcane, 3.3% Note: The underlying calculations assume concave production growth until 2030, based on historical average growth rates (2009–13), and 15 percent of the crop being irrigated and processed—no estimate available for floriculture endnotes 1. Of course, all of these factors are correlated. A value chain catering to export markets would likely add more value to the primary product through many production and processing steps and use of greater modern inputs. 2. FAOSTAT 2013 (http://faostat3.fao.org). 3. FAOSTAT 2014 (http://faostat3.fao.org). 4. International Institute of Tropical Agriculture (IITA) (http://www.iia.org/maize). 5. FAOSTAT 2013 (http://faostat3.fao.org). 6. In Zambia, an abundance of water and access to cheap grid electricity have played a significant role in the adoption of large-scale irrigated farming systems. 7. Tea and coffee are Rwanda’s most important exports (e.g., tea exports in 2013 totalled US$55 million); see FAOSTAT 2014 (http://faostat3.fao.org). 8. Production growth in Nigeria is driven by poultry-sector demand. 9. Estimates of ECA and Prorustica (2015). 10. For further analysis of commercial irrigated agriculture’s potential, see case studies 1 and 3 (chapter 4). 11. The load from processing rainfed tea is just 0.6 kW per ha. 12. Floriculture may not demand a large load in absolute terms as estates are seldom larger than 50 ha (requiring less than 120 kW for production). Exceptions may be additional power requirements for staff housing (see case study 5, chapter 4). Power Needs in Selected Value Chains37 13. Data for horticulture (pineapple) is missing and therefore not included. 14. A complementary analysis is the ongoing work in Latin America and the Caribbean on energizing agriculture; the study estimates energy demand for processing for selected value chains, and proposes energy efficiency options and associated costs (World Bank 2016b). 15. In India, this model has had some success through husk power systems. 16. Data on the electricity requirements of post-harvest activities (juicing, cooling, and canning) were unavailable. 17. An example is Del Monte’s biogas plant in Kenya, which is based on pineapple residue. 18. The produced biomass consists of empty fruit bunches (EFBs), palm kernel shells, fibers, and possibly solids from decanters; in most cases, this biomass is used to boil water and generate (super-heated) steam. 19. According to the FAO, economic losses for the dairy sector in Kenya, Tanzania, and Uganda total up to US$56 million per year. 20. The assumption for the irrigated proportion of a crop is in the ballpark of the CAADP target of doubling the land under irrigation by 2030; considering that about 6 percent of cultivated area is currently irrigated (FAO 2005), irrigated production has dispropor- tionately greater yield, and the selected value chains are the best performing crops in the region. Lessons from Ongoing Power-Agriculture Integration Projects Chapter 4 T his chapter presents a suite of case studies on Another important consideration is the trade-off power-agriculture integration in several coun- between affordability and cost recovery in setting elec- tries of Sub-Saharan Africa. All three countries tricity tariffs. While different regulatory environments covered—Tanzania, Zambia, and Kenya—show a afford different levels of flexibility in tariff setting for high potential for on-farm and agro-processing activities individual schemes, it is instructive to assess the tariff level to contribute toward regional and, in some cases, national that can optimally balance the cost recovery objective and power-sector development. These cases offer indicative affordability, in particular with respect to the anchor cus- analysis of specific project areas in terms of their potential tomer. The case studies aim to answer two key questions: and viability for furthering rural electrification.1 The objec- (i) Up to what price is power affordable for agriculture tive is to provide a point of reference for the potential of activities? and (ii) Below what price is power uneconomic power-agriculture integration and to highlight some of the to supply? important issues to consider in trying to promote such an Each case study is organized into four sec- integration. Each case study project asks (i) whether the tions: (i) power demand (agriculture and residential/­ investment in expanding rural electrification is economi- commercial), (ii) power supply options and commercial cally viable and (ii) under what conditions private-sector arrangements, (iii) financial viability, and (iv) economic participation in electricity supply is feasible. viability. Annex D presents the maps corresponding to the A standard cost-benefit analysis reveals that most case study project areas. of the projects analyzed are economically viable and are thus worth undertaking by governments.2 The social and economic benefits generated as a result of rural electrifi- Case Study 1. Tanzania: Sumbawanga cation often outweigh the costs incurred and may justify Agriculture Cluster well-designed subsidies to improve the financial viability of the project. Indeed, if the economic value of the grid The Sumbawanga agriculture cluster is located in the extension exceeds the economic costs (due to positive Southern Agricultural Growth Corridor of Tanzania externalities), an otherwise financially unviable project (SAGCOT), on the country’s western border (map D.1). can be undertaken with subsidy financing to cover the SAGCOT focuses on the coordinated development of shortfall. small and commercial agriculture, physical and market In many cases, private-sector participation is desir- infrastructure along the transport corridor that runs from able for developing and operating electricity supply as it Dar es Salaam through to (and immediately across) the can improve supply efficiency and reduce the financial Zambian border at Tunduma.3 Small-scale farmers are and capacity burden on public-sector providers. Thus, integrated into commercial value chains as outgrowers when analyzing various supply options, it is instructive and benefit from the agglomeration economies that lower to consider their commercial viability in order to under- costs of access to shared infrastructure and inputs (e.g., stand whether private-sector participation is viable and electricity, roads, markets, labor, and extension services) the amount of subsidy that may be required to attract (table 4.1). private-sector operators and developers. 38 Lessons from Ongoing Power-Agriculture Integration Projects 39 Table 4.1: Sumbawanga Agriculture Cluster at a Glance Project overview Expansion of electricity supply to support the development of an agriculture cluster and surrounding households through main power grid extension. Commodities Maize, sunflower, finger millet, paddy, and sorghum. Description Powered irrigation and residential demand are the main drivers of increased power demand. Grid extension is a viable option given that the grid extension passes through the Sumbawanga cluster to connect other load centers beyond it. Forecasted size of the load and limited local generation potential make grid extension the most feasible option. Powered irrigation is an important concentrated source of electricity demand. In its absence, greater dispersion of electricity demand over a wider area may reduce viability; thus, a greater cultivated area will be required to have large enough demand from processing. Financial As a stand-alone project, it is marginally financially unviable. A relatively small increase in electricity viability demand from agriculture or residential consumers would increase the financial viability of the grid extension. Economic Economic benefits would be significant (US$134 million) and justify the project. The benefits come viability mainly from household cost savings, small-scale irrigation, and increased commercial sale of produce. Still at a concept stage at the time of this writing, a few farmers use petrol and diesel-powered pumps which the Sumbawanga agriculture cluster aims to integrate are inefficient in water use and costly to run. To date, small-scale and commercial farming, along with process- there has been little penetration by solar pumps. ing and storage facilities, transport, and logistics hubs, With demographic and agricultural growth, forecasted and improved ‘last mile’ infrastructure to farms and local demand for electricity is expected to far exceed the cur- communities over an area of 27,000 km². The cluster rently available capacity. To meet this future demand, the has strong natural characteristics for agricultural devel- Government of Tanzania, through the Tanzania Electric opment, including proximity to Lake Tanganyika, good Supply Company Limited (TANESCO), intends to extend quality soils, and high rainfall. However, owing mainly to a 220 kV line from Tunduma (on the Zambian border) to its geographical isolation, the area lacks both physical Sumbawanga (and beyond through Mpanda to Kigoma on infrastructure (e.g., good roads, rail access, and power) Lake Tanganyika). and market infrastructure (e.g., integrated production and processing, traders, finance, and input suppliers). Power Demand Access to reliable and affordable electricity is critical to realize the cluster’s potential. Currently, the The annual power demand in the Sumbawanga region has Sumbawanga area benefits from a power capacity of the potential to increase to an estimated 60–70 MW by 10.6 MW serving a population of just over 1 million people 2030. Irrigation and residential demand are the expected (table 4.2).4 Where it is available, farmers and agribusi- main drivers of load growth, with commercial and pro- nesses purchase power from TANESCO (including from cessing loads playing a relatively less significant role its mini-grids). There is very little powered irrigation, but (figure 4.1). Agricultural demand. The majority of growth in electricity demand from agriculture will come from devel- Table 4.2: Sumbawanga Geographic oping the region’s irrigation potential, roughly estimated and Demographic Features at 50,000 ha.5 Assuming 35,000 ha of this amount is Feature Value dedicated to small-scale agriculture implies a total energy Estimated population (2012) 1,000,000 demand of roughly 25.5 MW by 2030 from both bulk Population growth rate (%) 4.0 water pumping and in-field irrigation. Newly irrigated land, higher quality inputs, crops switching, and knowledge Electricity connection rate (% of households) 7.0 sharing are expected to increase yields from 461,000 MT Sources: SAGCOT; ECA and Prorustica (2015). to 1.09 million MT by 2030 (table 4.3). 40 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.1: Estimated peak load and energy demand, by sector Power Capacity Demand (MW) Energy Demand (MWh/year) Source of Demand 2012 2030 2012 2030 Irrigation 0.0 25.5 0 48,450 Processing 0.4 4.4 2,000 22,000 Residential 3.9 26.7 26,232 174,327 Commercial 0.2 2.6 85 1,056 Total 4.5 59.2 28,317 245,833 a. Peak load b. Energy demand 80 300 60 200 GWh MW 40 100 20 0 0 2015 2017 2019 2021 2023 2025 2027 2029 2015 2017 2019 2021 2023 2025 2027 2029 Irrigation—small-scale Irrigation—large-scale Irrigation—small-scale Irrigation—large-scale Processing Residential Processing Residential Non-residential    Non-residential Source: ECA and Prorustica (2015). Table 4.3: Total Power Demand from Agriculture by 2030 Agriculture Activity Power Capacity Demand (MW) Hours of Operation/Year Energy Demand (MWh/year) Irrigation 25.5a 1,900 48,450 Processing 4.4b 5,000 22,000 Total 29.9 6,900 70,450 Sources: SAGCOT; JICA; Rukwa District Council; WREM International; ECA and Prorustica (2015). a. Based on a potential area of 50,000 ha under irrigation and an estimated power demand for irrigation of 0.65kW/ha (0.3kW/ha for small-scale farms and 1kW/ha for commercial farms). b. Based on a processed production of 472,500 MT and an estimated 11 mills required (400 kW). Lessons from Ongoing Power-Agriculture Integration Projects 41 Figure 4.2: Estimated volume of crops that may utilize electricity for processing 500 Volume of production (’000 MT) 400 300 200 100 0 2015 2017 2019 2021 2023 2025 2027 2029 Rainfed volume processed Irrigated volume processed Source: ECA and Prorustica (2015). The power demand for post-harvest processing Together, this implies an estimated power demand of will depend on the crops produced and the volume about 4.4 MW by 2030 (figure 4.2). of production. Electricity demand is expected for Residential/commercial demand. Based on the post-harvest processing of crops (e.g., milling and oil regional population growth rate of 4 percent, Rukwa’s extrusion) such as maize, paddy rice, beans, millet, population is expected to reach 2 million by 2030, repre- sorghum and sunflower.6 Greater electricity supply senting 400,000 households.7 Considering the house- and better access to markets for farmers would boost holds’ annual consumption and anticipating that their the electrification rate of agro-processing activities. demand and consumption will likely evolve over time with An estimated 40 percent of the current crop yield the adoption of additional electric appliances, residen- and an assumed 75 percent of the increased yield due tial consumers will be the main driver of energy demand to irrigation expansion will be processed by 2030. (table 4.4). Table 4.4: Residential and Commercial Data to Calculate Power Demand Residential 2012 2030 Population 1,000,000 2,025,817 Population growth 0.04 People per household 5 5 No. of households 200,000 405,163 Household connection rate 7% 20% Households connected 14,000 81,033 Per household peak consumption (kW) 0.28 0.33 Per household energy consumption (kWh/month/HH)a 156 179 Total peak (MW) 3.9 26.7 Total energy consumption (MWh) 26,232 174,327 Commercial  No. of customers 6 75 Consumption peak (kW) 34 34 Consumption energy (kWh) 14,085 14,085 Total peak (MW) 0.2 2.6 Total energy consumption (MWh) 85 1,056 Sources: SAGCOT; ECA and Prorustica (2015). a. Assumes a daily demand of 5.13 kWh per household. 42 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.3: Comparative cost of power supply options in Sumbawanga 100 90 80 80 80 Cost of power (US¢) 60 Cost to serve 40 Tari charged 25 25 20 11 10 11 0 Diesel mini-grid Hybrid mini-grid Solar mini-grid Main grid connection Source: ECA and Prorustica (2015). Commercial demand from current loads averages The least-cost method is thus estimated to be an 85 kWh per month across six TANESCO customers, with extension of the national grid. This would allow for more a peak load of 0.21 MW. Should per-customer demand efficient generation capacity sizing for demand on the levels remain as observed when electricity was supplied in system at more competitive costs. In deciding how much other areas of comparable size (e.g., Morogoro, Iringa, and transmission capacity to invest in, it is more feasible Mwanza), the number of customers would increase to 75; to install adequate capacity to meet future projected thus, annual power consumption would rise to 1,056 MWh demand rather than upgrade capacity in response to by 2030, and power-capacity demand would reach increase in demand. The subsection below describes a 2.6 MW (table 4.4). scenario where sufficient capacity is directly incorporated into a project’s initial design. Power Supply Options and Commercial Arrangements Financial Viability: Extension of Main Grid from Mbeya to Sumbawanga The analysis considered various options for additional and Rukwa power capacity to meet projected demand. Localized generation potential from diesel, solar, hybrid, hydro, and The financial viability of grid extension is estimated from bagasse/biomass was considered, along with the option the perspective of TANESCO. To supply activities in to extend the national grid. Preliminary analysis showed Sumbawanga, both grid extension and generation capacity insufficient potential for hydro- and biomass-based gen- expansion are required. However, generation capacity eration, so these options were ruled out. expansion is on a national least-cost basis; the focus here The option to expand mini-grid capacity, based on is on the viability of the transmission and distribution diesel, solar or a hybrid of the two, was also found unviable network development (table 4.5). for the region. The cost of a diesel-based mini-grid is The costs associated with provision of grid electricity estimated at US¢90 per kWh, which is much higher than to Sumbawanga consist of the cost of electricity gener- the cost of extending the national grid.8 Even if hybrid ation and transmission and distribution costs (expansion solutions enable the lowering of generation costs (i.e., at and operation). The corresponding revenues would US¢80 per kWh), they are still much more costly than be those of electricity sales at the national tariff level grid extension. Finally, solar mini-grids are not adapted to (table 4.6). the load profiles of agro-processing and irrigation activi- ties, which would imply expensive investments in storage and backups (figure 4.3). Lessons from Ongoing Power-Agriculture Integration Projects 43 Table 4.5: Estimated Capital and Operating Costs for Transmission and Distribution Expansion Operating Expense Grid Extension Cost Total Cost Assumption AC Losses Assumptions Distance (km) (thousand US$/km) (million US$) (%) (%) 11 kV 200 15 3.3 3 4.6 33 kV 200 35 7.7 3 4.6 220 kV 350 138 53.1 3 4.6 Subtotal (million $) 64.1 Present value (million $) 61.2 18.3 3.8 Total (million $) 83.4 Sources: Ministry of Energy and Minerals (MEM), Power System Master Plan; ECA and Prorustica (2015). Table 4.6: Estimated Power Consumption At the assumed 10 percent average cost of capital, the and Transmission and Distribution project is marginally financially unviable as a stand-alone Tariff Requirement project (table 4.7). However, TANESCO’s ability to attract financing on more favorable terms or greater reve- Variable Value nues from electricity demand, would improve the project’s Cost (million US$) 83.36 financial viability. On the other hand, a larger proportion Estimated consumption (MWh) 1.2 million of consumers paying lower lifeline tariffs, lower electricity Transmission and distribution, tariff requirement demand, and/or higher costs would further reduce the (US¢/kWh) 6.9 financial viability of the investment in grid extension. Source: ECA and Prorustica (2015). Economic Viability Table 4.7: Financial Present Value Analysis of the project’s economic viability adds social of Grid Extension net benefits to the financial net benefits accruing to the developer (TANESCO). Thus, the economic analysis Value includes benefits accruing to newly connected house- Variable (million US$) holds, benefits from improvement in agricultural yields, Revenue, based on TANESCO tariff 167.34 market access, and jobs creation (table 4.8). Transmission costs (83.36) The economic analysis shows that the economic bene- Generation costs (89.83) fits significantly outweigh the associated costs. In fact, the Effective project shortfall (5.85) benefits accruing to the households alone are sufficient to Internal rate of return (%) 12 justify the investment in grid extension. Sources: ECA and Prorustica (2015); World Bank. Note: Assumes a consumption of 1.2 million MWh over 20 years. The generation cost is based on cost for the upcoming Kiwira coal plant, at US¢ 7.5 per kWh (TANESCO 2012 Power System Master Plan Update, May 2013). The coal plant near Mbeya is expected to be completed by 2020. The average retail tariff is about US¢ 14 per kWh. 44 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.8: Economic Costs and Benefits of Sumbawanga Grid Extension Present Value of Cost/Benefit Economic Cost/Benefit Beneficiaries (number) (million US$) Net financial costs (5.85) Household cost savingsa 52,671 households by 2030 42.00 Small-scale irrigation 35,000 farmers (1 ha each) 34.50 Margin uplift from market access All small-scale farmers 26.80 Import substitution Tanzania broadly 8.52 No. of jobs created by electrifying the agriculture field 3,750 24.00 No. of jobs created by electrifying the town 550 4.20 Economic net present value 134.14 Source: ECA and Prorustica (2015). a. These are the additional households that are assumed to be connected from the grid extension project—over and above the baseline (w/o project). Additional household benefits may include better health outcomes from reduced fuel use, better educational outcomes for school going children, women’s time savings, and better nutrition. Case Study 2. Tanzania: Mwenga surrounding rural communities. The project was the first Mini-Hydro Mini-Grid green-field development under the Small Power Purchase Agreement (SPPA) scheme. The SPPA was signed with The 4 MW Mwenga mini-hydro mini-grid project is TANESCO in 2009, and the plant was commissioned in located in Tanzania’s Southern Highlands, close to the 2012 (table 4.9). RVE owns and operates the distribution Mufindi Tea and Coffee Company (MTC) (map D.2). network connecting roughly 20 villages and relies on a The project is operated by the Rift Valley Energy (RVE), mobile phone based pre-paid vending system for electric- a 100 percent subsidiary of the Rift Valley Corporation, ity billing. which also owns MTC. The project came about as a result Notwithstanding its long and complex development of MTC’s need to supplement electricity from the main process, Mwenga is considered Tanzania’s most success- grid to ensure access to a reliable source of uninter- ful private mini-grid development project. For the tea rupted power. Cofinanced by the European Union (EU) factory, the mini-grid is an opportunity to switch from and the Rural Energy Agency (REA), the project was grid-based power to a more reliable supply produced by developed as an independent power producer (IPP) to renewables. Although the project was initially designed to supply power to the main grid, local tea industry, and supply only the MTC, having power lines extending from Table 4.9: Mwenga Mini-Hydro Mini-Grid at a Glance Project overview A 4 MW hydro mini-grid connected to the main grid. Main local anchor load is the Mufindi Tea Estates and Coffee Limited; 2,600 households connected in the surrounding communities. Commodities Coffee, tea. Lessons learned The tea estate is the main anchor load of the grid connected mini-grid. Given the seasonality in tea processing operations, the peak load demand more than doubles during the summer season. This impacts the choice of power supply arrangement. Excess supply was sold to the grid, which helps mitigate the impact of seasonality. While residential consumers are numerous, their power demand is not high enough, at least initially, to mitigate the impact of a seasonal anchor load. Financial The project’s financial viability depends critically on the ability to sell excess power to the main grid. viability Despite financial viability, capital subsidies were provided for the project to keep local electricity tariffs low. Economic Economic benefits are positive (US$9 million) and come from households’ energy cost savings, reduced viability reliance on diesel backup for the tea estate, and job creation from new electrified businesses. Lessons from Ongoing Power-Agriculture Integration Projects 45 the hydro plant through nearby villages facilitated the All excess power from the mini-grid (about 80 per- connection of 2,600 households, as well as other com- cent of generated power) is sold to TANESCO, in accor- munity facilities. Beyond enhancing electricity access, the dance with its SPPA and feed-in tariff (FiT) arrangement; project has replaced the use of diesel and kerosene with these have been instrumental in guaranteeing offtake and sustainable hydropower among neighboring communities. have helped justify development of a scheme of its size, thus benefiting from economies of scale. Selling power Power Demand only to local consumers would not have justified the proj- ect in terms of its scale or commercial viability. Demand for power from the Mwenga mini-grid comes from the main grid (TANESCO), commercial and com- Power Supply Options, Commercial munity users, agriculture, and residential customers. As Arrangements, and Financial Analysis local demand is expected to grow, the sales to the grid are expected to decline. Local demand growth is expected to Proximity to the Mwenga River enabled the tea plant be led by the informal and semi-formal agriculture and to access a renewable source of power, with sufficient forestry sectors, highlighting the significant economic volume and head to develop a 4 MW run-of-the-river, development potential of the project. mini-hydro plant. The project is owned and operated by Agricultural demand. In terms of power for agri- MTC’s sister company and both are held by the RVC culture, MTC mainly requires electricity for processing. parent company. Specifically, electricity is used to power large motors, The project was developed as a private-public part- fans, and vibrating sieves (used to cut to length the leaves, nership and partly funded through public funds, including and wither, dry, sort, and grade the tea). The tea factory’s elements of grant and concessional loans from the EU peak load averages about 700 kW (with a summer peak and REA.10 The use of concessional funds was necessary of 900 kW and a winter peak of 400 kW), with an annual to reduce the tariff burden on local electricity custom- power consumption of 2,880 MWh.9 ers. While the electricity regulator allowed RVE to set Community and commercial demand. In addition cost-reflective tariffs, as per Tanzania’s SPP framework, to supplying agro-processing activities, the Mwenga fairness and affordability concerns led to the tariff being mini-grid project specifically targets facilities such as set in line with the tariff on the main grid. The regulator schools and clinics, as well as small commercial businesses, has allowed recent adjustments in the tariff, which is thereby improving electricity access for productive uses. currently TZS 100 per kWh up to 75 per kWh (equivalent According to RVE, annual power consumption for com- to US¢6.25 per kWh under the pre-devaluation exchange munity and commercial users is estimated at 2,988 MWh. rate). However, since 80 percent of the generated power Residential demand. Residential customers comprise is sold to TANESCO under the SPPA and FiT, the viabil- the majority of the customer base; however, most resi- ity of Mwenga’s hydro plant is not relying on the profit- dential customers have very low demand and pay lifeline ability of selling electricity to local communities. tariffs. Annual demand from the 2,600 customers is estimated at just 936 MWh (table 4.10). Table 4.10: Estimated Power Demand from Mwenga Mini-Hydro Plant Forecast Total Monthly Usage Connections Connections Approved Tariff (all customers) Customer Group (current) (2030) (TZS/kWh) (MWh) Households 2,600 5,600 100 78 Commercial 374 557 205 114 Public/community services 468 668 205 135 Tea estate 1 1 Uncertain 240 TANESCO 1 1 189 1,922 Total monthly usage (MWh) 2,489 Source: RVE. 46 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.11: Economic Costs and Benefits of Mwenga Mini-Hydro Plant Present Value of Cost/Benefit Economic Cost/Benefit Benefits (million US$) Net financial costs 0.0 Development subsidies received by project (7.1) Household cost savings (no. of households)a 5,600 6.4 Tea company savings from reduced diesel backup requirement (hours/year)b 288 1.4 Jobs created by electrifying villages (no.)c 1,120 8.6 Economic NPV 9.3 Source: ECA and Prorustica (2015). a. Households are assumed to save $14 per month from access to electricity; b. diesel backup requirement is assumed to be 10% of the total power consumption; c. it is assumed that 65 percent of the businesses will each create 1.5 jobs. Each job created is valued at the average expected salary: $1500/year. Financial Analysis Case Study 3. Zambia: Mkushi Farming Block The financial analysis considers the Mwenga mini-hydro project from the perspective of the revenues and costs The Mkushi farming block project is located in Zambia’s incurred by the owner, RVE. However, information on Central Province (300 km northeast of Lusaka) and revenue, operating cost, and capital expenditures was stretches over 176,000 ha of land (map D.3). The Mkushi confidential and thus not available. Despite this limitation, farming block is one of Sub-Saharan Africa’s largest multi- discussions with the operator allow us to make certain farmer commercial farming areas outside South Africa. salient points: Mkushi produces the largest share of Zambia’s wheat ºº Tanzania’s SPP framework allows RVE to charge a (40 percent) and soybean (21 percent), and is its sixth tariff that should ensure full cost-recovery, including largest maize producer. Other export crops grown in the a return on capital, even if all capital is at commercial area include tobacco, soya, vegetables, and coffee (Chu rates, and adjusted for any subsidies received. 2013). Mkushi experiences distinct dry winter seasons ºº In practice, social concerns implied that the tariff (May to October) and wet summer seasons (November to was set equal to the main grid. Thus, in order to April). Irrigation is thus critical for growing winter crops, accommodate this lower tariff, subsidies for capital especially wheat (table 4.12). expenditure were sought to reduce the effective cost Electrification of the Mkushi farming block occurred recovery, such that it aligned with the tariff. over time, given the evolving demand and difficulty of raising the necessary capital. Mkushi was first connected Given that RVE, a private-sector company, continues to the grid in 1996 through a 33 kV line. This effort was to operate the facility, one can assume that the project at financed by the government and a group of 20 farmers least breaks even financially. who contributed US$10,000 per km (50 percent of the total cost), which was the policy of the Zambia Electricity Economic Analysis Supply Corporation (ZESCO) at the time. However, unreliable power supply due to inadequate feeder capacity Economic net present value (NPV) is estimated at about meant that farmers had to continue to use backup diesel US$9 million, based on a 10 percent discount rate over generators for irrigation. A subsequent grid expansion the assumed project life till 2030 (table 4.11). Benefits was undertaken in 2000, followed by a third in 2005 accrue from household energy cost saving, reduced reli- to connect all farmers and many households in the area. ance on diesel backup for the tea estate, and job creation Expansion of the national grid into the area has enabled from newly electrified businesses. the area under irrigation to expand to about 18,000 ha Lessons from Ongoing Power-Agriculture Integration Projects 47 Table 4.12: Mkushi Farming Block at a Glance Project Overview Extending a transmission line into a farming area with significant agricultural potential. Commodities Wheat, soybean, tobacco, soya, vegetables, coffee. Description Irrigation counts for more than 90 percent of total power demand. Given their interest in the project, farmers accepted to contribute to capital costs. The grid extension enables a significant increase in household connection rates (from 2 percent in 1995 to 7 percent in 2014). However, more than 30,000 households remain unconnected to the main grid. Financial From a purely financial perspective and as a stand-alone project, grid extension to Mkushi was not Viability profitable for the utility. However, in order to expand access to new farmers coming into the area, sharing of capital costs was an appropriate and successful approach to project financing. Economic Thanks to household energy cost savings, increased yields from irrigation on small-scale farms, and job Viability creation, the project’s economic NPV was positive (US$46 million). Figure 4.4: Total peak load in Mkushi, 1995–2014 25.0 Irrigation 20.0 Milling Peak load (MW) Residential 15.0 Commercial 10.0 5.0 0.0 1995 2000 2005 2010 2014 Source: ECA and Prorustica (2015). and led to the subsequent development of milling Agricultural demand. Among agriculture activities, activities. irrigation has been the main driver of power demand, Out of 150 commercial farms hosted on the farm- with milling accounting for only a small share of total ing block in 2014, 80 farms have developed irrigation agricultural power demand. Power demand for irrigation schemes to enable wheat production in winter and to grew from 0.5 MW to 18 MW between 1995 and 2014, supplement summer crops. The availability of water and with a yearly consumption of 34,200 MWh in 2014 the connection to the national grid, supported by ZESCO (figure 4.5).11 In addition to development of irrigation and the Zambia National Farmers Union, were central to schemes, two mills were installed in the area following development of these irrigation schemes and processing arrival of the grid. Power demand for milling was esti- facilities. mated at 800 kW,12 for a consumption of 4,000 MWh (table 4.13). Power Demand Residential/commercial demand. Between 1995 and 2014, household connection rates grew from 2 percent to Between 1995 and 2014, overall peak load in Mkushi 7 percent, with the corresponding number of connected (from agriculture, residential, and commercial consump- households increasing from 362 to 2,516 (table 4.14). tion) increased from 0.6 MW to 20.1 MW. Over that Over the same period, power demand from residential period, irrigation accounted for more than 89 percent of and commercial customers increased from 0.13 MW to total power demand (figure 4.4). 1.32 MW, with households representing 67 percent. 48 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.5: Power demand from irrigation and milling in Mkushi, 1995–2014 20.0 Power demand 16.0 from irrigation Power demand (MW) Power demand 12.0 from milling 8.0 4.0 0.0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Source: ECA and Prorustica (2015). Table 4.13: Power Requirements for Irrigation and Milling in the Mkushi farm block Agricultural Activity Requirement 1995 2000 2005 2014 Irrigation Irrigated land area (ha) 500 7,000 10,000 18,000 Power demand (MW) 0.5 7 10 18 Power consumption (MWh) 950 13,300 19,000 34,200 Milling Power demand (MW) 0 0.4 0.8 0.8 Power consumption (MWh) 0 2,000 4,000 4,000 Sources: Ministry of Agriculture; ECA and Prorustica (2015). Table 4.14: Electrification Rates and Power Load of Households in Mkushi Consumer Type 1995 2000 2005 2014 Residential Households (no.) 18,092 21,488 25,521 34,782 Household connection rate (%) 2 3 4 7 Households connected (no.) 362 603 1,004 2,516 Power demand per household (kW) 0.24 0.26 0.29 0.35 Peak demand (MW) 0.09 0.16 0.29 0.88 Total consumption (MWh)a 222 408 751 2,247 Commercial Total demand (MW) 0.04 0.08 0.15 0.44 Total consumption (MWh)b 190 349 642 1,921 Sources: Ministry of Agriculture; Zambia Census 2010; ECA and Prorustica (2015). a. Assuming household energy consumption of 75 kWh/month. b. Assuming that commercial consumers operate 14 hour per day 6 days a week. Lessons from Ongoing Power-Agriculture Integration Projects 49 Total power demand in 2014 was 20.1 MW, with cor- specifying their peak demand load. They were required to responding annual energy demand of 42,368 MWh. Of cofinance up to 50 percent of the cost of the line exten- this amount, 18 MW came from irrigation, 0.8 MW from sion and pay for the transformers. processing, 0.88 MW from households, and 0.44 MW from commercial customers. Financial Analysis Power Supply Options and Commercial Given the cofinancing arrangement, the financial analysis Arrangements of extending the grid to the Mkushi farming block was analyzed from the perspective of both ZESCO and a Zambia has one of the lowest electricity tariffs in Sub- representative farmer newly settled in the area. From the Saharan Africa owing to fully depreciated hydropower utility’s standpoint, even after capital costs were partially dominating the generation mix. This implies considerable paid for by customers, the revenue generated from the benefits from reliable electricity supply to farmers who grid extension remained below the costs incurred. The previously relied on backup diesel generation. This, along financial NPV was estimated at US$8.9 million, mainly with the relative proximity of the main grid, ruled out a because of the very low electricity tariffs (table 4.15). mini-grid option. The farmer was required to invest in half of the line As described above, to extend the grid to Mkushi, extension for 20 km (US$10,000 per km), a transformer farmers were initially required to apply to ZESCO, ($50,000), and irrigation capital ($2,500 per ha). Table 4.15: Financial Analysis of Mkushi Farming Block from the Perspective of the Utility and a Representative Farmer Thousands of US$ Factor 1995 2000 2005 2014 Utility   Tariff revenue 14 161 244 424   Capital costs 1,300 10,000 5,045 0   Operating costs 39 339 342 342   Net benefits −1,325 −10,178 −5,142 82   Financial NPVa −8,89 Representative Farmer (500 ha of irrigated land)b  Wheat   Extra profit 60  Maize    Extra production because of irrigation (MT) 1,250   Extra profit 199    Total extra revenue from irrigation 259   Capital costs 1,500c    Electricity consumption from irrigation (MWh) 950    Cost of electricity 33   Net benefits −1,275 226 226 226   Financial NPV 523  IRR (%) 17 Note: The financial NPV is calculated over a 20-year project life starting from the initial investment (1995–2014). a. The estimated negative NPV is over 20 years. Given the magnitude of the stream of revenues relative to the costs, considering 30-year project life will not make the project financially viable from the utility’s perspective. b. Irrigated production of 500 ha of wheat in winter and 500 ha of maize in summer. c. For a 20km connection expansion. 50 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa However, after deducting the cost of electricity and yields and job creation (table 4.16). The economic NPV capital costs from the extra profit generated by irrigation, is estimated at about US$46 million, which justifies the the financial NPV for a representative farmer was positive 130-km grid extension (table 4.17). ($522,653), showing that the representative farmer The project faced various implementation barriers. benefited from increased yields, owing to supplementary Since it was not financially profitable for the utility, the summer irrigation, as well as irrigated winter cropping shortfall had to be covered by subsidies. Other issues that (table 4.15). had to be overcome included lack of access to capital for project financing, lack of coordination between farmers, Economic Analysis and insufficient grid capacity to provide reliable power supply. Moreover, ZESCO and farmers competed over From an economy-wide perspective, between 1995 and water availability and use; the utility wanted water for its 2014, the largest benefits from access to grid electricity hydropower plant, while the farmers wanted it to irrigate accrued from savings on electricity expenditure, dis- their lands. placement of imports due to increased wheat and maize Table 4.16: Net Social Benefits of Grid Extension, Mkushi Factor 1995 2000 2005 2014 Savings on Energy Consumption Electrification rate (%) 2 3 4 7 Households electrified (no.) 362 603 1,004 2,516 Savings from grid electrification per household ($/month) 10 Total savings on energy consumption (million $) 0.04 0.07 0.12 0.30 Import Savings Wheat  Irrigation area (ha) 500 7,000 10,000 18,000   Production (MT) 3,000 42,000 60,000 108,000  Import substitution value of wheat (million $)a 0.21 2.94 4.20 7.56 Maize   Production without irrigation (MT) 2,750 38,500 55,000 99,000   Production with large-scale irrigation (MT) 4,000 56,000 80,000 144,000   Benefit of locally grown production over imports (million $)a 0.11 1.51 2.15 3.87 Revenue from Job Creation Job creation from area under irrigation 143 2,008 2,868 5,163 Extra income from irrigation (million $) 0.22 3.01 4.30 7.74 Present Value of Social Benefits over the Period 1995–2014 (million $) 65.47 Note: Assumes a 10 percent discount rate over a 20-year project life. a. Import substitution is valued at the difference between farm gate price in Zambia and import price. Table 4.17: Economic Costs and Benefits of Grid Extension, Mkushi Factor Value (million US$) Financial NPV of utility −8.90 Present value of capital cost contributions from farmers –10.83 Present value of social benefits 65.47 Economic NPV 45.74 Source: ECA and Prorustica (2015). Lessons from Ongoing Power-Agriculture Integration Projects 51 Case Study 4. Zambia: Mwomboshi Power Demand Irrigation Development and Currently, Mwomboshi’s access to grid electricity is low. Support Project The northern bank of the river has no electricity supply. Among small-scale farmers who are not connected to The Mwomboshi Irrigation Development and Support electric power, only a small portion uses petrol or diesel Project (IDSP) is situated along the banks of the pumps for irrigation purposes. Along the southern bank, Mwomboshi river in Zambia’s Central Province (World electricity from the national grid is used to power staff Bank 2011b) (map D.4). The IDSP aims to support irriga- housing, crop irrigation, processing, and other small-load tion development in order to increase agricultural yields activities (e.g., offices, water pumping, and tea drying). and incomes in the area. The project also includes support Planning for sufficient capacity to consider future for complementary infrastructure, including roads and loads from expanded farming activities includes upgrading electricity. Irrigation will be developed from water storage the current 11 kV line to a 33 kV line with a 30 km grid (via construction of small- and medium-sized dams) and extension to the north side of the river, which would pro- transport to individual farms (table 4.18). An extension vide all farmers with electricity. By 2031, it is estimated of the grid to and within the site will be funded under that the aggregated peak load from agriculture, house- the project and handed over to the utility to operate holds, and commercial activities will reach 6.4 MW, repre- (ZESCO). senting an 18.5 percent average annual increase from the Direct beneficiaries of the IDSP are the area’s 3,700 2016 peak load (figure 4.7a). Driven by irrigation, power inhabitants, along with small-scale and commercial consumption is forecasted to reach up to 15,000 MWh farmers. Commercial farms are located along the south- by 2031 (figure 4.7b). ern bank of the river, while small-scale farming is mainly Agriculture demand (irrigation). In addition to the on the north side. The connection to electricity is critical 439 ha currently underirrigated in Mwomboshi, the IDSP to enable irrigation development, which creates greater plans to add an extra 3,200 ha, distributed between opportunities to increase incomes. small-scale and commercial-scale farms. This will allow for Covering 100,000 ha, on-farm irrigation develop- the release of bulk water supplied from a water storage ment can be categorized into four tiers: (1) small par- dam through pump stations for irrigation schemes. The cels of less than 1 ha each, which utilize flood irrigation project will become the area’s major power load, requir- systems; (2) individual farms with parcels in a range of ing 2 MW to supply the southern bank of the dam and 1–5 ha, which utilize spraying irrigation schemes; (3) plots 3.1 MW for the north side. Once the first pumps are larger than 60 ha each, cultivated by a community or installed, the power consumption of pumping stations commercial farm that uses modern irrigation systems is forecasted to rise from 872 MWh in 2016 to about (e.g., center pivots); and (4) large parcels cultivated by 10,000 MWh by 2031 (table 4.19). large-scale commercial farmers that are supplied water Agriculture demand (milling). Development of the through a bulk-water storage facility (figure 4.6). region’s wheat milling capacity will evolve along with the increasing yields expected from irrigation. Total energy Table 4.18: Mwomboshi Irrigation Development and Support Project at a Glance Project overview Grid upgrade and extension to support irrigation development and household electrification. Commodities Tobacco, wheat, poultry, maize, sunflower, horticulture (tomatoes, onions, bananas). Description Electrification is mainly driven by irrigation of small-scale and commercial farming, leading to crop diversification and increased yields. The project also targets near universal residential access in the area by 2031. Proximity of the existing grid and power needs meant grid extension was the only option considered viable. Financial Positive financial NPV estimated at US$1.1 million. viability Economic Positive economic NPV estimated at US$2.0 million for the power line extension, mainly from greater viability irrigated tomato and maize production. 52 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.6: Mwomboshi IDSP plot sites developed for small-scale farmers Bulk water infrastructure—Pump and mains pipes, may include dam/reservoir Tier 3—Professionally managed farm block under pivot irrigation growing marketed food and cash-crops, purchasing produce from emergent farmers, and providing support services. Tier 2—Emergent farmers growing food and horticultural crops under sprinkler or other irrigation for sale to and supervised by the Water source, e.g. river professional farmer (5 ha each). Tier 1—Smallholder gardens on land currently farmed can grow vegetables etc. for local and subsistence consumption under some basic form of irrigation, e.g. furrow (1 ha each). Source: World Bank 2011b. Figure 4.7: Mwomboshi peak load and power consumption forecast a. Peak load 7 Peak load from irrigation 6 Peak load from milling 5 Peak load residential 4 Peak load commercial MW 3 2 1 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 b. Power consumption 15,000 Power consumption 12,000 from irrigation Power consumption 9,000 from milling MWh Power consumption 6,000 residential Power consumption 3,000 commercial 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 Source: ECA and Prorustica (2015). Lessons from Ongoing Power-Agriculture Integration Projects 53 Table 4.19: Irrigation Power consumption over this period should rise from 78 MWh Requirements in Mwomboshi, Zambia to 1,137 MWh. Nonresidential demand, led by com- mercial activities, is assumed at half of residential power Irrigation Requirement 2016 2031 consumption. Its peak consumption is thus expected to Power demand (MW) 0.5 5.1 increase from 0.015 MW in 2016 to 0.22 MW by 2031 Power consumption (MWh) 872 9,757 (figure 4.8). Source: ECA and Prorustica (2015). Power Supply Options and Commercial Arrangements Table 4.20: Milling Power Requirements in Mwomboshi, Zambia Since the southern part of the area is already connected to the national grid, no other supply option has been Milling Requirement 2016 2031 considered for improving power availability. To do so, the Power demand (MW) 0 0.6 Ministry of Agriculture and Cooperatives and ZESCO will Power consumption (MWh) 0 3,000 sign a Memorandum of Understanding (MOU) framing Source: ECA and Prorustica (2015). responsibilities for the construction and maintenance of Note: Assumes a mill operates 5,000 hours per year (16 hours a the new power line. ZESCO will own the assets and be day, 6 days per week)/mill size: 200 kW. responsible for line maintenance after construction and will recover its operating costs through tariff revenues. demand from milling is expected to be significantly lower than that from irrigation (table 4.20). The first mill is Financial Analysis expected to be installed when total production from com- From ZESCO’s perspective, the grid upgrade project in mercial farmers and the marketed portion (80 percent) Mwomboshi is financially viable, with a positive NPV of of small-scale production reaches 20,000 MT. The plan US$1.1 million. Given the current average electricity tariff is to add an additional mill for every 20,000 MT of extra of US¢3.5 per kWh and the estimated level of demand, production. the utility’s revenues are calculated as the additional reve- Residential/commercial demand. The IDSP plans to nues received by the utility due to the project (table 4.21). increase household connections from 15 percent (2014) to 97 percent (2031). Based on a per-household power demand estimate, peak load would increase by 2 percent Economic Analysis a year as the household load evolves over time. Total The IDSP is estimated to generate positive net benefits residential peak load should therefore increase from with a NPV of US$2.0 million. The economic benefits are 0.03 MW in 2016 to 0.45 MW by 2031, while electricity driven largely by the increase in yields of irrigated tomato, Figure 4.8: Residential and commercial demand, electrification rate 2016–2031 1,500 100.0% Residential and non-residential demand 1,200 80.0% Electrification rate (%) (MWh/year) 900 60.0% 600 40.0% 300 20.0% 0 0.0% 2020 2021 2022 2023 2025 2026 2027 2028 2029 2030 2031 2016 2017 2018 2019 2024 Peak load residential Non-domestic power consumption Electrification rate Source: ECA and Prorustica (2015). 54 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.21: Financial Analysis, Mwomboshi Factor Assumption Electricity tariff (US¢/kWh) 3.5 Transmission tariff (US¢/kWh) 1.0 Transmission OpEx (% of CapEx) 3 Cost of capital (%) 10 Line expansion (km) 30 Cost of grid expansion ($/km) 30,000 Total cost of transformers ($) 175,000 Net Present Value (NPV) Calculations 2016–2031 Present value of revenues (million $) 2.4 Capital costs (million $) 1.1 Present value of operating costs (million $) 0.3 Financial NPV (million $) 1.1 IRR (%) 20 Source: ECA and Prorustica (2015). wheat, and maize production (table 4.22). Irrigation will is connected to the main grid and purchases electricity allow farmers to increase production through better yields from the utility, it can generate power at a lower cost. To and crop diversification. The electrification savings to increase output by 0.4 MW, a planned upgrade of the farmers from using diesel pumps and switching to elec- generation plant aims to provide power to both industrial trified irrigation schemes will be minor since only a small activities and some 2,000 households. number of farmers are currently using these irrigation solutions. As a result, the total present value of social ben- Power Demand efits for the entire project is estimated at US$34 million. However, as these benefits are the result of the whole Currently, ODCL’s power demand is 3.2 MW, with irrigation project in Mwomboshi (not only the electrifi- 13 MWh in annual consumption. Seventy percent of the cation component), the share of the cost of power line company’s total energy consumption is for industrial use— extension is used as a benchmark to allocate the share of mainly heating, ventilation, air conditioning (HVAC), benefits accruing to the electrification investments in the refrigeration, irrigation (pumping, drip irrigation, and project area. spraying), and lighting. Except for heating directly sup- plied by steam, many other industrial processes (e.g., ven- tilation, refrigeration, and irrigation) require electricity. Case Study 5. Kenya: Oserian Part of the power generated by ODCL is distributed Flowers and within the company’s estate to the community (e.g., Geothermal Power staff housing, schools, and clinics) and sister companies (e.g., tourism lodge). Currently, 2,000 households are The Oserian Development Company Limited (ODCL) connected to electricity through a mix of power from operates a 216 ha flower farm—including roses, carna- ODCL’s own power generation (95 percent) and utility tions, and statice—situated in Kenya’s Nakuru County power (5 percent). However, 2,000 other households (map D.5). The farm produces and exports 380 million within the estate remain without an electricity connec- stems annually, and employs 4,600 people (table 4.23). tion. ODCL is planning an increase in power generation ODCL is a pioneer business in its use of heat from by improving generation efficiency (via installation of geothermal wells for internal power generation and con- a partial condenser). The improvement in efficiency is sumption; its 50 ha Geothermal Rose Project is the larg- expected to increase generating capacity by 0.4 MW. The est of its kind. In addition to geothermal heat, a 3.2 MW expansion project seeks to supply these additional house- generator is dedicated to powering the farm’s operations holds for basic electricity uses (e.g., lighting and mobile and distribution within its estate. Although the company phone charging) and to power such facilities as schools Lessons from Ongoing Power-Agriculture Integration Projects 55 Table 4.22: Economic Costs and Benefits of the IDSP Project, Mwomboshi Benefit 2016 2019 2031 Revenue from job creationa Jobs resulting from the project — 313 313 Present value of increase in employees’ income ($ million) 3.4 Increase in profit revenue Small-scale (MT) Tomato production with project 5,000 57,833 57,833 Maize production with project 1,000 6,403 6,403 Wheat production with project — 3,602 3,602 Present value of profit of extra production ($ million) 20.5 Commercial (MT) Wheat production with project 2,634 12,240 12,240 Maize production with project 3,512 16,320 16,320 Present value of profit of extra production ($ million) 3.5 Savings from import substitution Present value of wheat and maize import substitution savings ($ million) 6.7 Savings from household electrification Electrification rate (%) 15 46 97 Electrified households without project 93 101 144 Electrified households with project 93 283 598 Present value of household electrification savings ($ million) 0.2 Total present value of economic benefits ($ million) 34.0 Financial NPV of utility ($ million) 1.1 Share of line upgrade project cost to total IDSP project cost (%)a 2.6 Net social benefits ($ million) 0.9 Economic NPV ($ million) 2.0 Source: ECA and Prorustica (2015). Note: Assumes that present values are over the 15-year period (2016–31). a. Because the project has multiple complementary investments, it is hard to disentangle the benefits accruing to the power line extension without a simplifying assumption; it is thus assumed that the accrual of benefits to electricity versus other investments is in the same proportion as the accrual of costs. Table 4.23: Oserian Flowers and Geothermal Power Project at a Glance Project Overview Expansion of the estate geothermal generating capacity and its distribution network to power the farm’s operations and distribution within the estate (staff housing, community facilities, and sister companies). Commodities Floriculture. Description ODCL’s captive power generates 95 percent of its requirements internally. Industrial use (heating, ventilation, irrigation, and lighting) represents 70 percent of the company’s total energy consumption. Since no power is exported to the grid or sold beyond the estate, ODCL has a license from the Energy Regulatory Commission for captive power generation and distribution. Financial With a positive financial NPV, the planned expansion project of 0.4 MW and electrification of 2,000 Viability households is financially viable. Economic Positive economic benefits estimated at US$2.5 million. The main economic benefit is based on Viability increased household electrification and, as a result, the savings are due to lower energy consumption costs (e.g., less use of kerosene and no more payment for cell-phone charging services and disposable batteries). 56 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 4.9: Power uses and sources at ODCL Energy use (in GWh/a) Energy sources (in GWh/a) Losses, 1.5 Communities, 1.2 KPLC, 0.6 Other uses, 1.4 Flower farm, 8.9    Production, 12.5 Source: ODCL. and a clinic. The limited increase in capacity implies that After this generation expansion, it is expected that the monthly household consumption may be constrained; plant will generate an additional 2,500 MWh per year. This however, households willing to upgrade may get individ- will include 600 MWh to offset electricity bought from ual connections through the state-owned utility, Kenya KPLC, another 600 MWh to supply the local population that Power and Lighting Company (KPLC) (figure 4.9). does not yet have access to power, and the remaining 1,300 MWh to cover ODCL industrial processes (figure 4.11). Power Supply Options and Commercial Arrangements Financial Viability ODCL’s captive power generates 95 percent of its The planned expansion project of 0.4 MW and electrification requirements. Power is generated from a farm-operated of 2,000 households is marginally financially viable, with plant, and steam is bought from the Kenya Electricity a positive financial NPV of US$3,742. The costs incurred Generating Company (KenGen) under a 15-year purchase for generation and distribution expansion and operation are agreement. Since no power is exported to the grid or sold slightly more than offset by the revenue from cost reduc- beyond the estate, ODCL has a license from the Energy tion in electricity purchased from KPLC. An investment of Regulatory Commission for captive power generation US$1.2 million is required for expansion of generation (partial and distribution. ODCL supplies power to staff workers condenser) and the distribution network (conductors, trans- within the estate using a mix of geothermal generation former, and switchgear). Also, operating cost is not expected and the main grid supply. Households consume low levels to increase as the expansion will not consume additional of energy and are not metered individually, and KPLC bills resources (e.g., the same volume of purchased steam). In ODCL rather than individual households. Over the years, fact, the increased output will lower the per-unit cost from ODCL has developed a skilled, in-house engineering team US¢6 per kWh to US¢5 per kWh. The operating cost will dedicated to geothermal power generation. therefore amount to $125,000 (table 4.24). To meet unmet power demand and offset electricity In comparison, the savings from the reduced pur- purchased from the utility, an investment of US$1 million chases from KPLC amount to $342,000. Staff house- is planned for expanding geothermal plant capacity up to holds are to be supplied electricity free of charge. 3.6 MW (figure 4.10). An additional $0.2 million will be Charging households cost-reflective tariffs would incur required to finance the distribution network extension. additional costs due to metering and billing. Considering ODCL is considering charging electricity customers a these costs in the analysis shows that, in order to break cost-reflective tariff, but this would require an additional even, a cost-recovery tariff of US¢8 per kWh would be $0.2 million investment in individual meters. required. Lessons from Ongoing Power-Agriculture Integration Projects 57 Figure 4.10: Output of ODCL’s power plants and expected increased output 3.0 2.5 2.0 1.5 1.0 0.5 0.0 s 00 rs 00 rs 00 rs rs 18 rs 22 rs 00 rs 00 rs 00 rs 00 rs 00 rs 00 rs 10 rs 110 rs s 14 rs 15 rs 16 rs 17 rs 19 rs 20 rs 21 rs 23 rs 24 rs hr hr h h h 0h h h h h h h h h h h h h h h h h h h h 00 07 02 03 00 00 08 09 00 00 00 00 00 00 00 00 01 04 05 06 00 00 00 00 00 12 13 OW 202 power plant OW 306 power plant Additional output Source: ODCL. Figure 4.11: Electricity output of capacity expansion project and intended uses 600 MWh to connected community (o setting KPLC tari of 0.18 $/kWh) 1,300 MWh for industrial Output of capacity use of farm expansion project 2,500 MWh/a 600 MWh for new connections to 2,000 households Source: ECA and Prorustica (2015). Table 4.24: Financial Analysis, ODCL Economic Analysis Item US$ Amount The expansion project constitutes a relatively small Revenues 342,000 portion of the estate’s electricity use; most electricity is Power generation Opex costs 125,000ª used for irrigation and refrigeration. The main economic Capex costs 1,200,000b benefit from the expansion project is thus from increased Margin –983,000 household electrification and, as a result, the savings due to lower energy consumption costs (e.g., less kerosene use Discount rate (%) 10 and no more payment for cell-phone charging services Financial NPV ($ amount) 3,742 and disposable batteries). An electricity connection is Source: ECA and Prorustica (2015). estimated to save households US$11 per month, implying a. Assumes a cost per kWh of $0.05. b. Assumes $1 million for distribution and $200,000 for metering. 58 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa $2.5 million in total net economic benefit (NPV) over the projects in the small-scale tea subsector aimed at reducing life of the project. No significant impact is expected in factory operating costs, improving power supply reliability, terms of job creation or commercial development. and diversifying tea farmers’ revenue sources. The power generated from these schemes will be used primarily in the tea factories, with the surplus sold to KPLC under a Case Study 6. Kenya Tea Development power purchase agreement (PPA). KTDA is in the process Agency Holdings: Mini- of setting up several small hydropower projects for its tea Hydro Mini-Grids factories. One hydropower plant has been operational in the Imenti tea factory since 2010; an additional 17 proj- This case study analyses the mini-hydro based tea factory ects are in the pipeline, ranging from 0.5 MW to 9 MW, electrification project of the Kenya Tea Development eight of which are at an advanced stage of development, Agency (KTDA). The agency is planning the implemen- with feasibility studies completed. tation of several small-scale (≤ 15 MW) run-of-the river hydropower projects at various locations in Kenya to Power Demand serve a number of tea factories under its management (map D.6). Considering the near-term pipeline, along with the KTDA is the single largest producer and exporter of operational Imenti plant, the total installed capacity is tea in Kenya. The company was created in 2000, sub- 24.4 MW. About 40 percent of power generated will be sequent to privatization of the Kenya Tea Development used primarily for the tea factories’ self-consumption, Authority. KTDA is the holding company of a number supplying mainly tea industrial processes. The remaining of subsidiaries owned by small-scale tea companies. The 58 percent of output will be sold to KPLC under a PPA agency currently manages 63 factories in Kenya’s small-­ and feed-in-tariff (FiT) scheme. Farmers will benefit from scale tea subsector. Currently, its network covers about the electricity supplied to the factories that they partially half a million small-scale farmers, with each tea factory own, but residential electricity connections will only be owned by 5,000–10,000 tea farmers (table 4.25). provided through KPLC, and not directly though KTDA. KTDA Power Company Limited, a subsidiary of Approximately 187,500 small-scale farmers, representing KTDA, is charged with consolidation, investment, and 25 tea factories, will benefit from these power projects management of energy initiatives undertaken by tea to run their farming activities. Currently, 70 percent of factories managed by KTDA. Notably, KTDA Power neighboring households (i.e., more than 130,000 farm- Company supports the development of hydropower ers) lack access to electricity. Table 4.25: Kenya Tea Development Agency Holdings: Mini-Hydro Mini-Grids at a Glance Project Overview Development of hydropower plants powering tea factories and staff housing, and selling surplus power to the grid. Commodities Tea. Description The operational power plant and eight projects have a total installed capacity of 24.4 MW. About 187,500 small-scale farmers, representing 25 tea factories, will benefit from these power projects to run their farming activities. Mini-hydro plants provide a more reliable power supply to tea factories at lower cost and avoid the need for backup generators. Financial Viability Evaluation of a sample project (North Mathioya) shows that the project is financially viable, with a NPV of US$3.3 million. Revenues accrue from the sale of power to the grid and cost savings by tea factories. Economic Viability The same sample project is evaluated as economically viable, with a NPV of US$10 million. Direct and indirect impacts on rural electrification include the following: electrification of staff housing, reduced connection costs for surrounding households, development of stand-alone home systems. About 30,000 households will benefit from electricity connections. Lessons from Ongoing Power-Agriculture Integration Projects 59 Power Supply Options Figure 4.12: KTDA’s North Mathioya hydropower project: financial benefits The KTDA tea factories have two feasible supply options and power sold for meeting their power requirements: (i) purchase from the main utility at the retail tariff or (ii) self-generate Revenues (total $3.3m) electricity through the planned hydropower projects. Grid-supplied electricity is often unreliable, with frequent outages and voltage fluctuations. The need for a reliable power supply for tea operations requires investment in backup diesel generation, which adds to the overall cost of electricity. Where feasible, a captive mini-hydro genera- tion plant, with the ability to sell excess power to the main Power sold grid, is an attractive option both financially and in terms of (total 28.4 GWh/a) increased reliability. In terms of commercial arrangements, KTDA Power Company leads the project development cycle (e.g., permitting acquisition, securing land, and raising capital) and forms special purpose vehicles (SPVs) in the form of regional power companies for each project (e.g., North Mathioya Power). The factory farmers served by the Revenues from PPA, 51% Power consumed, 36% mini-hydro plant are shareholders, and raise 35 percent of Revenues from tea Sale of power to KPLC the investment cost as equity from deductions of farm- farms, 49% under PPA, 64% ers’ tea revenues. Electricity to residential consumers in Source: ECA and Prorustica (2015). the area will be provided through KPLC and not directly through the project. of about $1 million per year, as well as pressure to reduce tariffs along with KPLC’s national rates. Given these Financial Analysis assumptions, subsidies for both capital expenditure and operating expenses would be required. The financial analysis focuses on the North Mathioya (5.6 MW) hydropower project from the perspective of Economic Analysis the SPV owners. Project revenues derive from the sale of electricity to the grid at the FiT.13 The remaining elec- Although KTDA power projects are not involved in the tricity sold to tea factories is valued at the avoided cost retail sale of electricity to neighboring communities, they of grid plus diesel backup electricity at US¢16 per kWh have several direct and indirect impacts on rural electri- (figure 4.12).14 fication. First, they provide electricity to staff housing, The costs include the capital and annual operating which represents an average of 60 households per factory. expenditures of the generation plant incurred by the Second, they may facilitate grid access for the surround- SPV, at US$22.5 million and $165,800, respectively. ing households by reducing connection costs. Third, these Comparing the present value of the stream of revenues areas will be targeted by a pilot project—led by the KTDA and costs, the project is estimated to be financially viable, subsidiary, Greenland Fedha (microfinance institution), with a NPV of $3.3 million (at 10 percent cost of capital) and the KTDA Foundation—which aims to finance solar and an IRR of 13 percent. home systems (SHSs) for farmers and support their grid Although the project does not include household or connections. community electrification, except for factory staff hous- The estimate of economic benefits is based on facil- ing, a simplified financial analysis shows that such activity itating households’ access to electricity connections. Tea would be financially unviable without subsidies. Despite factory activities remain unchanged, although they gain the relatively high margin between household retail rates access to a more reliable, cheaper source of power supply. (US¢20 per kWh) and the PPA rate (US¢9 per kWh), Approximately 30,000 households will benefit from distribution and retail would require an additional capital electricity connections, which will offset their expenditure expenditure of US$15 million and administrative expenses on traditional or more expensive forms of energy. 60 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa The project will facilitate grid connection by con- concept in Tanzania also shows this cause-and-effect necting the generation facility. Costs are estimated at relationship between irrigation and processing. Increase US$500 per grid connection, with a monthly electricity in the scale of processing activity can lead to a significant bill of $3 per household. Also, the above-mentioned increase in power demand. SHS scheme in place for farmers will further increase The seasonality of power demand from the agricul- connections,15 with an average household savings of $11 ture sector can significantly constrain a project’s viability. per month.16 Thus, development of the North Mathioya Large seasonal differences in electricity-dependent agri- hydropower project will provide households net economic cultural activity will impact the cost recovery of invest- benefits; the project’s NPV is $6.7 million, implying ments in electricity supply. In such cases, it is important $10 million in total economic NPV.17 to consider ways to mitigate the impact of a variable load. One option, especially for mini-grid or captive generation, is the ability to sell excess power to the grid, as in the Key Conclusions from cases of mini-hydro development in Tanzania (Mwenga) the Case Studies and Kenya (KTDA).18 Increased processing activities in the post-harvest season may complement electricity The six case studies discussed in this chapter offer varied demand from irrigation, and irrigation itself may reduce contexts for power-agriculture integration. Each is unique seasonality in agricultural production and thus electricity in terms of the type of anchor load and country setting; demand by allowing multi-cropping (e.g., in the case of thus, one must be cautious about generalizing from the les- Zambia’s Mkushi farming block). sons learned from any particular case. Keeping this in mind, Finally, when considering agricultural anchor loads, this section discusses key findings from the six case studies it is more risky for the investment to depend on a single in terms of large power loads, supply options, financial and large customer since any negative shock to the customer economic viability, and financing of development. would negatively affect operating revenues for the elec- tricity supplier. For this reason, agricultural clusters (e.g., Large Power Loads Sumbawanga in Tanzania) can be used to increase the viability of rural electrification. Clusters development, by The viability of providing electricity depends critically on design, has load diversity and thus involves less risk than the existence of a large and stable demand for electricity reliance on a single anchor load. While not included in the (or supply, especially if the grid is supply constrained). case studies discussed in this chapter, the presence of a In rural areas, it is likely that the largest single source of private electricity supplier and private off-takers will price power demand is either agriculture or an agriculture-­ any such risk into the supply contract, thus increasing the related commercial activity. Residential electricity could price of electricity for all customers. In such cases, diversi- also be a significant source of demand (e.g., in the case fied cluster development can also help reduce the price of of Tanzania’s Sumbawanga agriculture cluster); however, electricity. The public sector may also help mitigate this risk this demand is often relatively dispersed, which reduces its through a grid connection and FiT, subsidies to increase the viability. customer base, or various guarantee/insurance instruments. In rural agricultural areas, irrigation is often the single largest potential source of electricity demand, Supply Options as exemplified in Tanzania’s Sumbawanga agriculture cluster, Zambia’s Mkushi farming block and Mwomboshi’s Most of the grid extension projects are justified by irriga- IDSP. These projects also show that the loads for agro-­ tion development, with agro-processing as a supporting processing activities (e.g., milling and extrusion) are activity. These developments require cultivating suitable comparative smaller, suggesting that the latter activi- commodities (e.g., maize, wheat, rice, and sugar), typ- ties, taken alone, may not be sufficient to justify rural ically grown on large-scale commercial farms, enabling electrification investments. These several projects also large production volumes. Small-scale farmers can then highlight how irrigation and processing are often linked. be incorporated alongside; however, they also need other The Zambia cases show how increased yields from irriga- forms of support, including access to a reliable water tion are an important prerequisite for the development supply, good physical and market infrastructure, and clear of large-scale processing activities; the agriculture cluster land with good quality soils. Lessons from Ongoing Power-Agriculture Integration Projects 61 The case studies discussed indicate that the national if the regulation allows it. Given this difficulty, financial grid usually plays an important role in the viability of rural viability, in most cases, depends on the ability to sell electrification investments—either in the form of the bulk power and lower costs. The Oserian geothermal and main supply option for agricultural and rural electricity KTDA projects show that estate-type developments demand (e.g., the Sumbawanga cluster in Tanzania) or (floriculture and tea in these respective cases) can under- as the main off-taker of the locally generated electricity take financially viable electricity investments, benefiting from a small power producer (e.g., the Mwenga mini-­ from reduced electricity costs and selling excess electric- hydro mini-grid in Tanzania). Whether the grid is the most ity to the grid. Another example is the case of the IDSP in viable supply option depends on various factors, includ- Zambia, where a grid extension was financially viable from ing distance to the grid, size and stability of electricity the utility’s perspective, owing to proximity to the grid demand, grid reliability, and local resource potential for (i.e., lower costs) and complementary investments in a generation. large irrigation scheme that increased electricity demand. Supplying rural electricity demand though small power In contrast, grid extension to the Mkushi farming block, producers (SPPs) depends critically on local generation also in Zambia, was not financially viable for the utility, potential (e.g., for mini-hydro, geothermal, and bio- despite a capital cost-sharing arrangement with benefi- mass). Viable generation potential can be a cost-effective ciary farmers. option in cases where the grid is far away, unreliable, or The choice of optimal tariffs—such that costs are expensive. In the latter case, especially, SPPs may benefit recovered and electricity consumption is affordable to primarily from selling to the grid and supplying local agri- farmers, businesses, and other customers—depends on cultural activities and residential customers in the process the size of the financial surplus generated from electricity (e.g., Tanzania’s Mwenga mini-hydro mini-grid). consumption and the constraints on how to allocate it Companies specializing in the agriculture or agribus- across various suppliers and customers. Additional consid- iness sectors may be unwilling to enter into electricity erations, such as parity with the main grid tariff, are the generation and, especially, the distribution business. main determinants (e.g., Mwenga mini-hydro mini-grid in This would be a departure from their core activities and Tanzania). may not be financially attractive enough to change their If there is flexibility in setting tariffs, then the range business model. In this respect, a variety of arrangements of feasible tariffs would be determined by the difference are possible, depending on the context and capacity of the between the customer’s willingness to pay (WTP) and entities involved. For Kenya’s Oserian geothermal project the supplier’s willingness to accept (WTA).20 A custom- and KTDA’s mini-hydro project, the companies chose to er’s WTP will be determined by the monetary benefit develop and operate the generation plant and supply their from consuming a unit of electricity. For households, operations, preferring to sell power to the grid and leave this may be a reduction in spending on their current retail power supply to the utility. For Tanzania’s Mwenga energy supply options, which are usually more expensive mini-hydro project, by contrast, RVE manages the mini- and less reliable (e.g., kerosene lamps or batteries). For grid generation and distribution, including retail power agricultural consumers, it may be driven by a reduction sales.19 in backup energy supply and/or increased revenues from higher productivity. A supplier’s WTA will be determined Financial and Economic Viability by development and operating costs, often represented by the levelized cost of electricity (LCOE) (table 4.26). The case studies discussed show that a rural electrification Assuming the WTP is more than the WTA, an optimal project can be financially viable where there is a creditable tariff may be negotiated based on some surplus allocation large off-taker and access to concessional loans/grants rule. Otherwise, if the WTP is lower than the WTA, the for capital investments. All six projects were estimated to government must step in to provide subsidies to bridge generate economic benefits well in excess of associated the gap as long as the project remains economically viable. costs, thus implying that all were economically viable. For all six of the cases analyzed in this chapter, the Tautologically, financial viability rests on the ability economic viability was high. For projects that are not to charge cost-reflective tariffs. In the case of mini-grid financially viable, economic viability is an important cri- development, charging consumers a tariff that is much terion to determine whether subsidies should be provided higher than the grid tariff might be difficult to do, even and at what level. Even with financial viability, subsidies 62 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Table 4.26: Typical LCOE Values for Small-Scale Generation and Distribution Systems Generation System Power Plant Size Range Capital Expenditure LCOE Operating Time Technology (kW) (US$/kW) (US$/kWh) (hours/year) Diesel genset 5–300 500–1,500 0.3–0.6 Any Hydro 10–1,000 2,000–5,000 0.1–0.3 3,000–8,000 Biomass gasifier 50–150 2,000–3,000 0.1–0.3 3,000–6,000 Wind hybrid 1–100 2,000–6,000 0.2–0.4 2,000–2,500 Solar hybrid 1–150 5,000–10,000 0.4–0.6 1,000–2,000 Distribution System LCOE Distribution Type Voltage Level (US$/km) Required Length Low-voltage 400 V 5,000–8,000/km 30 customers/km Average connection cost: $350/customer; average distribution cost: $200/customer. Medium-voltage 33 kV 13,000–15,000/km Total ($/kWh) 0.25–1 Source: IED Reference Costs for Green Mini-Grids. may be incorporated into the project to achieve other the spread of fixed costs, especially capital costs, across goals, such as grid parity in terms of tariffs or greater a larger pool of customers with diverse peak-load pro- adoption of electric irrigation. files. For example, since productive users need electricity during the day and households’ peak load is in the evening, Financing of Development the system peak load should be lower than the sum of individual peak loads. However, load balancing requires All six projects analyzed shared two common issues: an analysis of load profiles to optimize supply, and the (i) making projects financially viable and (ii) providing level of additional benefit depends on the proportion of funding for viable projects. Several ways have been iden- capital costs in total costs and the load matching between tified to make projects financially viable. To benefit from customers. The utilities—owing to their larger-capacity economies of scale, capacity for local generation can be cross-subsidization and ability to spread costs over a wider increased beyond the level of local demand, and surplus customer base—are usually in a better position to do so. power can be sold to the grid. This option is particularly As detailed for the Mwenga and KTDA projects, sell- relevant in countries that have introduced FiT programs ing power to more reliable customers, such as the utilities, set above the utility’s avoided costs. Selling excess power increases a project’s viability since anchor customers are makes it possible to lower the per-megawatt cost, but assumed to be better payers. This is especially true in relies on the ability to sell excess generated power. For countries where clear schemes for renewable energy FiTs example, the capacity of Tanzania’s Mwenga mini-hydro have been introduced with dedicated funding. Although mini-grid is greater than what the tea estate requires; relying on the utility still depends on its ability to afford therefore, the surplus is sold to the utility and nearby rural payments, the anchor-customer approach has reduced customers. the risk of the utility’s non-payment by giving certainty Another option, as done for the main grid exten- on tariffs. sion projects in Zambia (Mkushi and Mwomboshi), is to Finally, the role of subsidies to cover certain costs require the beneficiaries to partially finance projects and should be highlighted. All of the distributed schemes ana- share the development costs with major customers. In lyzed in this chapter have received subsidy payments to this way, farmers partially contribute to the capital costs decrease the level of cost recovery through retail tariffs. in exchange for receiving power. A further option is load This approach contributes to ensuring maximum capacity balancing across beneficiary categories, which enables development, increasing the project’s NPV, improving Lessons from Ongoing Power-Agriculture Integration Projects 63 tariff affordability for customers, and attracting private- to be financially unviable, developers can be encouraged sector participation. Subsidies are particularly necessary to expand their customer base to capture additional sub- for most privately developed, small-scale projects under sidies, prioritizing smaller customers close to each other 5 MW. By subsidizing household connections, which tend rather than larger ones. endnotes 1. The analysis presented in these case studies is indicative only and not a comprehensive feasibility study. 2. The only exceptions are projects based on quite expensive sources of power generation for small demand loads. 3. SAGCOT aims to facilitate the development of seven agribusiness clusters along the southern corridor of Tanzania’s Southern Highlands. 4. This comprises 3 MW from a 66 kV line into Zambia, 5 MW from a mini-grid in Sumbawanga, and a 2.6 MW mini-grid in Mpanda; both are isolated, diesel based mini-grids operated by TANESCO. 5. ECA and Prorustica estimates, consistent with the SAGCOT investment blueprint, constructed from own analysis and various official sources. 6. Other products such as cassava and livestock are also likely to demand electricity for processing, but for the sake of simplicity, are not included in the calculations here. 7. According to Tanzania’s national census, Rukwa had 1 million inhabitants in 2012. 8. The cost calculations consider all capital and operating expenditures; the calculations are based on ECA analyses conducted for small-scale systems in Kenya, Tanzania, and elsewhere in Sub-Saharan Africa. 9. Assumes that the factory operates 16 hours per day, 6 days a week for 10 months out of the year. 10. EU funds were through the African, Caribbean, and Pacific Group of States facility; and REA funding was supported by the World Bank’s Tanzania Energy Development and Access Project (TEDAP). 11. Assumes that the area’s power demand from irrigation is 1 kW per ha and average irrigating hours per year are about 1,900 (with a 15 percent load factor), representing in part the seasonality in demand for irrigation. 12. Assumes that the average mill has a power demand capacity of 400 kW and operates 5,000 hours per year. 13. US¢9.29 per kWh under a FiT. 14. Assumes a diesel generation cost of US¢60 per kWh (KTDA) and an overall tariff decrease of 5 percent annually. 15. Since this analysis focuses on the impact of an anchor load on household electrification, we restrict it to grid-connected households. 16. Observed for mini-grid development in Kenya. 17. If we assume that 50 percent of the 30,000 households connected are from SHS, then the household net benefits increase to US$14 million and the overall NPV to $17.2 million. 18. Apart from the mitigating impact of seasonal variation, the ability to sell excess power to the grid also helps invest in large genera- tion capacity and reduces costs due to economies of scale in generation. 19. Enabling small-scale, private power generation and distribution requires clear regulations and purchasing processes (e.g., PPAs and FiTs); regulations in Tanzania are relatively transparent in this regard. 20. The difference between WTP and WTA is a measure of the total surplus generated by the electricity sale/consumption. Opportunities to Harness Agriculture Load for Rural Electrification Chapter 5 W hat is Sub-Saharan Africa’s potential for the power load of bulk water pumping and infield irrigation harnessing power-agriculture synergies for systems were made.1 rural electrification? This chapter considers For each farming type, the production volume was this question, using a simulation model and used to calculate the milling load for the area, based on case studies from Ethiopia and Mali—two countries that assumptions about the proportions of milled production. exhibit a range of innovative options moving forward to With total milling volumes, the total load requirement 2030. Before turning to the case studies, the chapter for milling was estimated, based on the load characteris- presents a hypothetical case illustrating the conditions tics of an assumed average mill. Household and business under which power demand from agriculture could be connections for the given area were also estimated, based economically viable. on assumptions about a consistent population density and members per household, connection rate, household power consumption, and proportion of this load for busi- Simulation of Power Demand ness consumption. in a Stylized Agricultural Setting The stylized analysis from the simulation model helps to determine the general features of power demand from A simplified simulation model was developed to analyze agricultural areas. Based on the average power demand the relationship between agricultural activity, power from agricultural sources, the results show that a fairly demand, and the geographic area that a power supply large area of coverage would be required to aggregate would serve (table 5.1). The model assumed a theoretical sufficient electricity demand from customers; based on circular area around the generation source, with electric- the model assumptions, a 50 km radius area would, on ity consumers distributed uniformly throughout. Further average, aggregate 60 MW of demand. simplifying assumptions were made about what percent- In the simulation, as in the case studies, irrigation age of this area was under cultivation and the proportions accounts for a substantial proportion of power demand split between small-scale and commercial farmers. The from agriculture (figure 5.1).2 The irrigation power load electricity demand from each of the two farmer groups is dependent on choice of crops and availability of bulk were estimated separately, with differing proportions of water. Some systems with a large body of available area under irrigation and yields (on rainfed and irrigated water nearby the infield irrigation system may require summer and winter crops). The model assumed that there little bulk water pumping; however, in cases where were two crops: summer maize and winter wheat. Across water must be pumped into storage before utiliza- Sub-Saharan Africa, maize is a common summer crop on tion, additional electricity is required. As such, total both mixed-used commercial and small-scale farms. In observed power loads for irrigation are in a range of the winter months, irrigated wheat is commonly grown. 0.5 kW–2.0 kW per ha. Based on the areas under irrigation, assumptions about 64 Opportunities to Harness Agriculture Load for Rural Electrification 65 Table 5.1: Assumptions for Typical Area/Agricultural Activity/Power Demand Model Assumption Basis Small-scale Value Commercial Value Overall Value Proportion of total land area Observations of other 25 under cultivation (%) large-scale production areas Proportion of farming type Observations of other 70 30 within cultivated area (%) large-scale production areas Proportion of irrigated land Observations of other 20 50 (%) large-scale production areas Summer crop yield (rainfed) Maize yields 1.5 6 (MT/ha) observed Summer crop yield (irrigated) Maize yields 4 8 (MT/ha) observed Winter crop yield (irrigated) Wheat yields observed (not 2 5 (MT/ha) grown without irrigation) Proportion of crop milled Observations of other 25 80 (%) production areas Irrigation load requirement Average, based on 0.3 1.0 (kW/ha) schemes observed Milled load Average mill, 200 (kW) consultant calculations Hours of operation (hrs/day) Average mill 16 Days of operation (hrs/year) Average mill 313 Population density (per km2) Comparison with other 50 countries People per household (no.) Comparison with other 5 countries Household connection rate Comparison with other 50 (%) countries Peak household consumption Various household power- 0.3 (kW) consumption studies Business load as proportion Various rural business 50 of household load (%) power-consumption studies Source: ECA and Prorustica (2015). The relatively low load for processing suggests that (figure 5.2). By contrast, the impact of the proportion of the machinery used for typical post-harvest processing crop processed is relatively low, especially as this load is operations (e.g., mills) does not require large amounts already minimal. of electricity, in part, because of the small size; also, it may be in operation for fewer hours in a year. Thus, most crop-processing loads are fairly small for the volume Simulation Study 1. Ethiopia: Power processed, with the exception of such activities as sugar Generation from Sugar Estates processing, which provides much or all of its own power. The total power load for a given area is highly sen- Sugarcane is an important crop in Ethiopia (map D.7). sitive to the assumed area under commercial irrigation, Indeed, the Ethiopian Sugar Corporation (ESC) aims reiterating the importance of irrigation to power loads to increase national annual production nearly eightfold 66 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 5.1: Power demand and Figure 5.2: Sensitivity of power load breakdown for a given area radius to changes in percent of commercial irrigation 80 70 Peak power load (MW) 60 60 Peak power load (MW) 50 40 40 30 20 20 0 10 5 10 15 20 25 30 35 40 45 50 0 Area radius (km) 0 10 20 30 40 50 Irrigation Processing Households Businesses Area radius (km) 15% 35% 50% Source: ECA and Prorustica (2015). Source: ECA and Prorustica (2015). Table 5.2: Ethiopia: Power Generation from Sugar Estates Project Overview Self-generation of power from bagasse and sale of power surplus to the main grid. Commodities Sugar. Description Sugar processing and irrigation are the largest sources of electricity demand. Irrigation makes it possible to extend the sugarcane production season and therefore smooth the annual profile of both production and processing. Processing and refining are the most power consuming activities in the sugar estate. Typically, a sugar processing plant can produce enough electricity from bagasse to meet its own electricity demand, and sell excess power to the grid. The viability of connecting such processing plants to the grid depends on the amount of excess power produced, the cost of producing it relative to other sources, and additional customers that can be connected. Financial From the utility’s perspective, extending the grid to the sugar estate is not financially viable—the net Viability present value (NPV) is negative because the utility does not benefit from sales to the estate, which self-supplies; from the sugar estate’s standpoint, the project is highly profitable (US$139 million). Economic The economic NPV for the whole period is positive ($367 million), thus justifying development of the Viability project. within five years. To do so, the government has launched Power Demand the Sugar Development Programme, with the objective of upgrading existing estates and commissioning new ones Agriculture (Irrigation). Traditional sugarcane production (table 5.2). is heavily water dependent. Irrigation ensures year-round This simulation analyzes a representative example of production of the crop and therefore a smoothing of the power-agriculture integration on sugar estates in Ethiopia. annual profile of processing activity. This means that sugar Sugar estates have the potential to generate power facilities operate throughout the year with a consistent from bagasse, a natural by-product of sugar refining. electricity demand. Hypothetically, the potential electricity generation is Irrigation is also a major source of power demand in enough to cover the electricity needs of the refinery and the sugarcane production process. In Ethiopia, irrigated associated facilities and sell the surplus to the main grid or land is expected to increase from 1,500 ha to 9,000 ha other supply schemes. over 20 years. The associated power demand from Opportunities to Harness Agriculture Load for Rural Electrification 67 irrigation over the same period is expected to rise from the sugar estate, depending highly on production volume. 0.8 MW to 4.7 MW,3 with power consumption increasing Considering forecasts in terms of yield rates and produc- from 2,340 MWh to 14,040 MWh (table 5.3).4 tion increases, the power requirements for processing irri- Agriculture (Processing and refining). Processing gated sugarcane will amount to 6,300 MWh in year 1 of and refining are the most power-consuming activities in the hypothetical model, rising to 37,800 MWh five years later. For processing rainfed production, power consump- Table 5.3: Total Power Demand tion will increase from 3,150 MWh to 18,900 MWh over from Agriculture and Residential/ the same 20-year period (figure 5.3). Commercial Loads Beyond processing, refining activities also consume power for centrifuging raw sugar and crystallization. Over Power Capacity Energy Demand the 20-year period, electricity consumption from refining Demand (MW) (MWh/year) is estimated to rise from 300 MWh to 1,800 MWh, while Demand Source Year 1 Year 20 Year 1 Year 20 power load will increase from 0.09 MW to 0.54 MW.5 Irrigation 0.8 4.7 2,340 14,040 Staff housing. In addition to agricultural needs, sugar Processing 2.9 17.5 9,450 56,700 estates also require power for staff housing and other Refining 0.1 0.5 300 1,800 supporting activities. Given that the average household electricity consumption in rural Ethiopia is about 0.10 kW Residential 0.1 1.5 384 3,783 (including staff (increasing to 0.15 kW by year 20),6 total electricity housing) demand from staff housing is estimated at 0.02 MW in Commercial 0.1 0.6 278 2,780 year 1, increasing to 0.21 MW by year 20. Residential/Commercial demand. In this model, Source: ECA and Prorustica (2015). the area is not yet connected to the grid, but a 30-km Figure 5.3: Estimated energy demand and peak load, by sector a. Energy demand 100,000 Electricity demand (MWh) Irrigation 80,000 Processing Refining 60,000 Sta housing 40,000 Rural households Nonresidential 20,000 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years b. Peak load 20 Irrigation Processing Peak load (MWh) 16 Refining 12 Sta housing 8 Rural households Nonresidential 4 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years Source: ECA and Prorustica (2015). 68 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa grid extension is finalized once the sugar factory is built. Table 5.4: Sugar Factory Power Thanks to the proximity of houses and the factory, the Generation in Years 1 and 20 electrification rate rises sharply to 85 percent by year 4. Electricity Electricity As the population grows from 28,500 to 50,386 by Generation Demand year 20,7 with household growth following the same Sugarcane Capacity (MW) (MWh/year) trend,8 the total electricity load from rural households will reach 1.3 MW by year 20. For commercial activities Processing Type Year 1 Year 20 Year 1 Year 20 surrounding the sugar estates, consumption is expected Irrigated 3.8 22.5 12,600 75,600 to increase from 278 MWh in year 1 to 2,780 MWh by Rainfed 5.8 35.0 6,300 37,800 year 20 (table 5.3).9 Total 5.8 35.0 18,900 113,400 Source: ECA and Prorustica (2015). Power Supply Options and Commercial Arrangements Bagasse is commonly used to generate electricity in sugar Beyond meeting its own power needs, the sugar factories. It is mainly used as a boiler fuel to generate steam factory can generate surplus power.10 This supports the to meet the sugar factory’s heating and power needs. The development of estate activities, especially irrigation, level of net electricity generation assumes (i) a bagasse before enough on-site bagasse has been produced. It generation potential of 29 MT for every 100 MT of sugar- also covers shortfalls in power generation during planned cane produced and (ii) a 70 kWh generation capacity for annual maintenance when the mills are not operating every MT of sugarcane. Since irrigated and rainfed process- (May–September) (table 5.5). ing of sugarcane do not occur simultaneously, the power The capital cost of extending the grid line 30 km to capacity of generation equals the maximum capacity of the the sugar estate and surrounding villages is US$2.4 million two, that is 47 MW by year 20 (table 5.4). (table 5.6). Table 5.5: Net Power Generation from Sugar Factory by Year 20 Power Capacity Hours of Energy Demand Agricultural Activity Demand (MW) Operation/Year (MWh/year) Irrigation (A) 4.7 3,000 14,040 Processing (B) 17.5 3,360 56,700 Refining (C) 0.5 3,360 1,800 Total demand 22.7 72,500 Power generated during processing (D) 35 113,400 Net power surplus D − (A + B + C) 12.3 40,900 Source: ECA and Prorustica (2015). Table 5.6: Capital Cost Assumptions for Grid Connection Cost Component No./Distance (km) Unit Cost Cost (million US$) 230 kV shunt/line/transformer (thousand $/unit) 15 25 0.4 Associated switchgear (thousand $/unit) 1 120 0.1 33 kV line (thousand $/km) 50 14 0.7 11 kV line (thousand $/km) 120 10 1.2 Total 2.4 Source: ECA and Prorustica (2015). Note: Costs estimates are based on those for similar projects in Ethiopia’s 2014 Electrification Master Plan; cost assumptions include connecting villages along the power line (i.e., 33 kV and 11 kV lines and transformers). In reality, the estate may feed back power to the villages from the substation. Opportunities to Harness Agriculture Load for Rural Electrification 69 Currently in Ethiopia, however, no sugar factory From the sugar estate’s perspective, the combina- exports its power to the grid because of the country’s tion of heating and power from bagasse combustion is (i) low electricity tariffs and (ii) unclear regulations on a fundamental asset for sugar processing and refining. conditions of exporting power to the main grid. A feed-in- The project’s financial viability depends on the following tariff (FiT) proposal, which aims to provide incentives to factors (table 5.8): private investors, is expected to become law in 2016 and ºº Capital costs, linked to development of the whole should clarify those conditions; thus, under future devel- estate, including land improvement, buildings and opment plans, power sold to the grid will be at the FiT. It equipment, and staff housing. is unlikely that sugar estates will sell directly to residential ºº Production costs, including employee wages, seeds, customers; this will be left up to the electricity utility. harvesting, loading, transport, maintenance, and electricity costs. Financial Analysis ºº Expected revenues from sugar sales and power sales. The project’s financial viability can be analyzed separately The project is highly profitable for the sugar estate, from the respective standpoints of the utility and the with a NPV of US$139 million. As mentioned above, the sugar estate. From the utility’s perspective, extending the large financial benefits for the sugar estate create ample grid to the sugar estate is not financially viable; the esti- scope for a negotiated arrangement of capital cost sharing mated NPV is negative, at US$ –1.5 million (table 5.7). to improve the utility’s financial viability. The viability is driven by the amount of power purchased by the utility, the margin between retail tariff and the price at which electricity is purchased from the sugar fac- tory (possibly the FiT), and the cost of extending the grid. Table 5.8: Sugar Estate Capital Costs, The price at which the utility purchases power from Assumptions for Production Costs, the independent power producer (IPP) is confidential. In and Revenues the absence of actual data, it is assumed that the utility tariff margin is US¢1 per kWh, which amounts to 40 per- Component Value cent of the domestic tariff.11 Capital costs (million US$) The project is not viable for the utility, in large part Land improvement ($3,500/ha) 41.9 because it does not benefit from sales to the estate, which Buildings and equipment 80.5 self-supplies. Subsidies would thus be required for project Staff housing ($5,000/house) 7.0 development. Given the significant financial benefits Present value of total capital costs 129.4 that will accrue to the sugar estate from the project, one option could be to have the sugar estate contribute to Production costs capital costs. Average wage ($/month) 100 Permanent employees (months/year) 12 Temporary employees (months/year) 7 Table 5.7: Financial Analysis from the Seeds costs ($/ha) 515 Utility’s Perspective Harvest cost ($/MT) 6 Present Value Loading cost ($/MT) 2 Component (million US$) Transport to sugar mill ($/MT) 3 Net revenue from sales 2.7 Maintenance (% of capital expenditure) 3 Expenses (Opex, losses, depreciation) 1.8 Present value of total production costs 311 Capital cost 2.4 (million US$) NPV −1.5 Revenue (million US$) IRR (%) 7.6 Present value of sugar sales 573 Present value of exported power to the grid 6 Source: ECA and Prorustica (2015). Present value of total revenues 579 Note: The discount rate is 10 percent over the 20-year period; of total capital costs, operating costs account for 3 percent, while Sources: Agritrade; ECA and Prorustica (2015); ESC; IEA; losses and depreciation each account for 5 percent. National statistics. 70 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Economic Analysis over land ownership; despite the government’s ability to make quick investment decisions regarding state-owned The project’s total economic benefits, estimated at about property, identifying large tracts of high quality agricul- US$410 million, comprise household energy cost savings, tural land is difficult in Ethiopia. Third, regulations on sugar estate profits, job creation, and import substitution exporting power to the grid must be clarified by defining (table 5.9). tariff rates that guarantee investors a price for selling The economic NPV over the period, about generated power from bagasse to the utility. Finally, US$367 million, equals the sum of the net social bene- selling power to the utility carries off-taker risk; delayed fits linked to the electrification project (figure 5.4), the payments for power sold or even payment defaults would financial NPV, and the present value of the sugar estate greatly impact the sugar factory investor. investment cost (table 5.10). Various factors could hinder the development of such agriculture-power schemes in Ethiopia. The first one is Simulation Study 2. Mali: Mini-Grid funding availability for grid extension; however, given the Expansion for Productive Users project’s associated economic benefits, funding from the government, development partners, or even cost sharing Mali is a regional success in rolling out private mini-grid with the sugar estates could be sought. Second, for green- concessions for rural electrification (map D.8). field development, investors face issues about uncertainty Table 5.9: Net Economic Benefits of Grid Extension to the Sugar Estate Benefits Year 1 Year 5 Year 20 Household energy savings Electrification rate (%) 21 85 85 Households electrified (no.) 1,479 6,752 9,964a Savings from grid electrification per household ($/month) 17 Total savings on energy consumption (million $) 0.025 0.12 0.17 Incremental income to the sugar estate Production revenues (million $) 14.8 74.2 89.2 Production costs (million $) 11.0 39.5 46.7 Sugar estate’s profit (million $) 3.8 34.7 42.5 Sugar estate jobs created Monthly salary ($/month) 100 Permanent jobs created (no.) 933 4,663b 5,595 Temporary jobs created (no.) 1,588 7,939 9,527 Total salaries (million $) 2.2 11.1 13.4 Non-sugar jobs created Jobs created (no.) 1,260 6,301 7,561 Salaries paid (million $) 0.13 0.63 0.76 Import substitution New production of sugar (MT) 42,000 210,000 252,000 Value of import substitution (million $) 1.3 6.3 7.6 Total economic benefits (million $) 7.5 52.9 64.4 Source: ECA and Prorustica (2015). a. The difference in the number of connected households between years 5 and 20 is related to population growth, which is expected to increase by 2.89 percent. b. Assumes 0.37 permanent job and 0.67 temporary job (working 7 months a year) created by hectare—Estimation based on the number of employees in Metehara sugar factory in Ethiopia. Opportunities to Harness Agriculture Load for Rural Electrification 71 Figure 5.4: Net social benefits of grid extension to sugar estate (years 1–20) 60 50 Economic benefits (million $) Household energy saving 40 Incremental income for sugar estate 30 Jobs created in sugar estate 20 Non-sugar job creation 10 Import substitution 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Years Source: ECA and Prorustica (2015). Table 5.10: Economic Net Present Value In 2015, it has 255 operating concessions, with a total of Extending the Grid to the Sugar installed capacity of 22 MW. However, mini-grid opera- Estate tors face key challenges, including the saturated capacity of their schemes and low revenues, which hinder invest- Value ment in capacity expansion. Limited power-generation Item (million US$) capacity has constrained the mini-grids’ ability to supply Financial NPV of Ethiopian Electric Power −1.5 households and serve productive users. The current ser- Corporation (EEPCO) vice level—limited daily hours (typically in the evenings) Present value of investment cost of sugar −41.9 and tariffs that are higher than on-site diesel generators estate (usually above US$0.50 per kWh)—are inappropriate for Net social benefits 410.0 meeting agro-industry power requirements. As a result, Economic NPV 367 productive users in off-grid areas use their own diesel generators as a more competitive power supply option Source: ECA and Prorustica (2015). Note: The discount rate is 10 percent over the 20-year period. (table 5.11). Table 5.11: Mali Mini-Grid Expansion for Productive Users at a Glance Project Overview Capacity expansion of an existing hybrid mini-grid (diesel-solar PV) to serve productive users. Commodities Agro-industrial activities. Description The Koury mini-grid is reaching a point of near saturation as generation capacity is fully taken up by existing household demand. However, small-scale commercial and agro-industrial activities in Koury (milling, water pumping, and bakeries) present significant opportunities for supplying unmet power demand. Attracting powered small businesses as mini-grid customers would require incentives to (i) lower tariffs, (ii) supply electricity during the daytime, and (iii) replace manual equipment with electricity powered machinery. Financial From the perspective of SSD Yeelen Kura, the rural energy services company, the Koury mini-grid is Viability in a fragile financial situation. However, the capacity expansion project is profitable, thanks to a higher payment rate, additional revenues, and proportionally low capital expenditure and operating expense (with a NPV of €103,000). Economic The economic NPV for the expansion project is slightly negative (−€18,000) as no significant savings Viability are expected from agro-industrial customers, who currently use individual diesel generators. However, the project could become economically viable if other economic, environmental, and social benefits were considered (e.g., reduction in CO2 emissions, reduced reliance on imported fuels, and exposure to price fluctuations). 72 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Based on a representative example of an existing mini- private mini-grid projects, an expansion project has been grid, this simulation study analyzes how agro-­industrial designed to assess the viability of supplying agro-industrial activities may improve mini-grids’ financial viability, loads. The simulated study also evaluates the potential for while benefiting from a more sustainable and competi- adding value to agricultural activities in rural areas through tive source of electricity. Based on the potentially lower mini-grid supplied power. Powered agricultural activities costs of hybrid solar photovoltaic (PV) projects, the study can indeed improve rural communities’ revenues and explores the potential for attracting agro-industrial power therefore potentially increase mini-grid operators’ profit demand to mini-grids. Given that there is no precedent (box 5.1). for tying medium- or large-scale industrial processing to Box 5.1: Isolated Mini-Grid Systems in Mali: Existing and Potential Power DemanD In Mali, large-scale irrigation schemes are gravity fed, with electric power used only for small diesel or petrol-­ powered pumps. Four key commodities that could benefit from greater access to electricity are mango, rice, shal- lot, and shea kernel. Mango. Mali’s Bamako and Sikasso regions are particularly favorable for growing mango. But to export larger vol- umes, Mali must handle various issues related to market transport and product handling, notably reliance on cold chains (e.g., fixed and mobile chilling facilities). Considered a production hub, Sikasso would be the logical location to set up a temperature-controlled mango packing house. Areas outside Sikasso not yet connected to the main grid have limited potential for extending or replacing cold-chain packing-house facilities; such areas are mainly served by isolated mini-grids or diesel gensets. An alternative value chain to fresh mango is processing mango pulp or nectar. Mali has only lightly exploited this value chain due to the lack of transforming infrastructure, irregular sourcing from small-scale farmers, and dis- tance to markets. Excess mango production can be used for dried mango or canning. However, high start-up costs and working capital would be required; this is not economically viable, given Mali’s low margins and small scale. Rice. Mali is a net importer of rice. Its rice production system uses gravity-based irrigation without mechanized bulk water pumping or infield irrigation. On the processing end, rice milling (husking) occurs throughout small- scale private milling operations, using both diesel-powered mobile or fixed husking machines and fixed-site mills. However, Malian milled rice is of low quality, with a high volume of broken rice. In some high production areas connected to the main grid (e.g., the 100,000 ha Office du Niger), larger-scale, fixed-site mills have been devel- oped with higher quality rollers that reduce broken rice, thereby adding value to the volume of rice sold. In addition to pure processing activities, post-hulling bran-hull biomass is used to generate power for the mill and related activities, as well as lighting on the premises and for staff housing facilities. Shallot. Mali could potentially become a major West African exporter of shallot, thanks to favorable growing conditions. Shallot is grown on small-scale farms across the country, and 90 percent of production ends up in local urban markets. Shallots can be provided fresh or variously processed (e.g., dried, crushed, or machine sliced, [potentially] using solar drying panels or improved solar heaters). Electricity is required for only two processes: (i) pounding and drying and (ii) slicing and drying. Since consumers prefer the fresh form of shallot, the market for transformed shallots is limited, and higher pro- duction costs induced by processing cannot be justified. The main opportunity is extending the market season for fresh shallot, capturing value from price fluctuations due to reduced market volumes. More efficient stocking and drying techniques would make fresh shallot available 4–6 months beyond the regular growing season and Opportunities to Harness Agriculture Load for Rural Electrification 73 over a year for its dried form. Because storage and drying processes require small amounts of power, there is little opportunity for power to add value to the commodity’s value chain, especially in areas not yet connected to the main grid. Shea kernel. Mali is a minor market player in kernels and butter, capturing less than 10 percent of global demand. Penalized for poor quality and yield, unreliable supply, and higher costs, Malian kernel exporters can hardly com- pete with other West African producing countries. Vegetable oil firms in Europe, India, and Japan dominate the global market, while West Africa accounts for only a handful of industrial extraction facilities, some of which work on a toll basis for global companies. Though Malian farmers have an incentive to produce higher quality kernels, they have little incentive to expand their kernel processing capacity, given the limited potential benefits (Derks and Lusby 2006). Manual processing of shea fruit includes kernel removal from pits; drying, moulding, and grinding kernels into paste; and kneading paste into separate solids and oils. These activities could benefit from mechanization, but weighed against the required investments, the benefits are not obvious, especially given the low labor costs and limited access to capital. Sources: FAO and Authors. Power Demand from Mini-Grids schemes that require water pumping rely on decentralized pumps spread over large areas. In Mali, households consume 90 percent of mini-grid electricity, which is mainly used for lighting, with peak load occurring during evening hours. The Koury mini-grid, Power Supply Options and Commercial located in a rural community of Yorosso circle (cercle) Arrangements in the Sikasso region, is operated by SSD Yeelen Kura, The Koury mini-grid is reaching a near saturation point as a private operator that manages 21 concessions12 and generation capacity is fully taken up by current demand. has started to hybridize its mini-grids with solar PV. More than 20 percent of the generated electricity is from In 2012, Yeelen Kura added 100 kWp of solar PV to diesel generators (figure 5.6). The variable cost of thermal the existing 112 kW of thermal capacity, making power generation, at €0.40 per kWh,13 and the cost of direct available 10 hours a day (typically from 3 p.m. to 1 a.m.). consumption (below €0.20 per kWh) suggest the advan- Because of the mini-grid demand profile, the solar tages of expanding solar PV capacity. output produced by PV generators is stored in batteries, Notably, expansion of solar PV could enable the elec- which increases energy losses and capital expenditure tricity provision for productive activities since they require (figure 5.5). power mainly during the daytime. Direct consumption of The Koury mini-grid currently supplies 180 MWh per solar output would (i) avoid energy losses in the battery year, mostly for households. Out of 3,371 households bank and (ii) reduce the battery bank size relative to living in the area, 556 are already connected to the mini- capacity of the solar PV generator. grid, at an average consumption level of about 24 kWh per To attract businesses as mini-grid customers, incen- month. tives would be needed to (i) lower tariffs, (ii) supply The opportunities for supplying unmet power demand electricity during the daytime, and (iii) replace manual from small-scale commercial and agro-industrial activi- equipment with electricity powered equipment. Figure 5.7 ties in Koury are significant. Although such activities rely shows the impact of adding the daytime loads of pro- mainly on their own diesel or petrol engines or genera- ductive users, along with a 50 kWp matching capacity tors, they represent a total potential energy demand of expansion of the solar PV system (totaling 150 kWp) on 7,755 kWh per month—about a 50 percent addition to the Koury mini-grid load profile.14 the existing energy production of the mini-grid power This capacity expansion is assumed to fall under plant (table 5.12). Irrigation is not expected to play a the existing rural electrification program of the Malian significant role for the mini-grids, given that most irriga- Agency for Development of Household Energy and tion in Mali utilizes gravity fed schemes, and small-scale Rural Electrification (AMADER) and therefore benefits 74 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure 5.5: Koury mini-grid: Electricity consumption patterns 100 90 80 Load 70 Solar output 60 50 kW 40 Solar output Loads supplied not directly 30 by battery bank used goes to and/or genset 20 battery bank 10 0 6:00 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00 Public, 5% Commercial & industrial, 5% Residential, 90% Source: SSD Yeelen Kura. Table 5.12: Potential Addition of Small Agro-Industrial Activities and Other Businesses Typical Energy Consumption Total Consumption Business Type Number (kWh/month) (kWh/month) Milling or grinding (maize, rice, shea kernel) 6 300 1,800 Water pumping 2 300 2,520 Bakery (electric mixer) 1 300 450 Mechanical workshop (welding, grinding, drilling) 2 1,260 300 Media center (computer, printer) 1 450 135 Petrol station (pumps) 1 150 300 Small shops (refrigerators, freezers, TV, lighting) 10 135 2,250 Total 7,755 Source: GERES and SSD Yeelen Kura. Opportunities to Harness Agriculture Load for Rural Electrification 75 Figure 5.6: Energy generation profile from capital expenditure subsidies, with ownership of at Koury site infrastructure remaining with the government and the operator regulated under contract. 600 Taking a conservative approach, it is assumed that agro-industrial customers’ willingness to pay will be 500 capped at the costs of running individual diesel gensets. This implies that the tariffs needed would be lower than 400 current household tariffs. kWh/day 300 Financial Analysis 200 From the perspective of SSD Yeelen Kura, the current 100 financial situation of the Koury mini-grid is somewhat precarious (table 5.13, figure 5.8). Although operating 0 expenses are covered by revenues, the 20 percent capital expenditure contribution of the private operator is not 4 14 14 5 5 5 t-1 1 1 -1 v- c- n- b- recovered through tariffs. In order to achieve a 10–15 per- ar Oc No De Ja Fe M cent return, the project receives up to 80 percent of Diesel genset output (kWh/day) capital expenditure subsidy from the government. Equity Average solar output (kWh/day) investment and reinvestment in capacity expansion and Max solar output (kWh/day) replacement of major parts (e.g., batteries and gensets) Source: SSD Yeelen Kura. cannot be recovered. Figure 5.7: Koury mini-grid profile: Additional commercial and industrial loads 160 140 120 Total load 100 Solar ouput New commercial/ 80 kW industrial load 60 40 20 Daytime loads supplied directly by solar 0 6:00 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00 Public, 12% Commercial & industrial, 28% Residential, 60% Source: SDD Yeelen Kura. 76 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Economies of scale, daytime energy use, and falling consumption from the mini-grid. Largely as a result of solar PV prices imply that the expansion project could be the significant capital subsidies, the expansion in gener- attractive as it allows for additional revenue with relatively ation capacity is financially viable from the perspective low capital expenditure and operating expense. The oper- of SSD Yeelen Kura, with a positive NPV (table 5.14). ating costs will marginally increase due to higher expenses However, if viewed from the perspective of AMADER or in maintenance and administration, but will be offset by a the Government of Mali, the asset owners, the financial lower level of generation losses due to direct consumption returns are negative (essentially including the subsidy of solar power (reducing the need for storage) and lower costs in the calculation). use of thermal generation. Along with the capital subsidy to the developer, this implies a lower average tariff and Table 5.14: Financial Analysis of Capacity creates the incentive for new customers to switch from Expansion of Koury Mini-Grid their current diesel generators to daytime electricity Item Amount Commercial and industrial customers served (no.) 20 Table 5.13: Current Financial Situation Average total consumption (MWh/year) 80 of Koury Mini-Grid Average retail tariff (€/kWh) 0.40 Item Amount Payment rate (%) 90 Households served (no.) 556 Additional revenues (€) 28,800 Average total consumption (MWh/year) 160 Operating costs (€)a 5,600 Average retail tariff (€/kWh) 0.55 Capital costs before subsidy (€)b 189,000 Payment rate (%) 80 Capital costs after 80% subsidy (€) 37,800 Revenues (€) 70,500 Project cash flows NPV after subsidy (€)c 103,000 Operating costs (€)a 55,400 Project IRR (%) 56 Capital costs before subsidy (€)b 831,000 a. Including the cost of fuel and increased maintenance and Capital costs after 80% subsidy (€) 166,200 administrative expenses; excluding depreciation. NPV after subsidy (€) (259,700) b. Including an additional investment of 50 kWp of solar PV; assumes no additional expense in the distribution network. a. Including corporate overhead and fuel, maintenance, and c. Additional parameters affecting cash flows and thus the administrative expenses; excluding depreciation. calculation of NPV include (i) reinvestment in batteries (every b. Including the cost of solar and diesel powered generation and 6 years) and inverters (every 12 years), which are not subsidized; battery storage, as well as costs of the distribution network, civil (ii) increased fuel costs, given a PV system degradation rate of and electrical works, and engineering; current (2015) costs are 0.5 percent per year; and (iii) a 10 percent weighted average used (i.e., €5,300 /kWp, excluding the distribution network). cost of capital (WACC). Figure 5.8: Operating expense and capital expenditure distribution a. Operating expenses b. Capital expenditures Other O&M, 12% Distribution, 30% Other, 32% Operator overhead, 52% Fuel costs, 36% Storage, 20% Solar PV, 18%     Opportunities to Harness Agriculture Load for Rural Electrification 77 Similar to most other mini-grid projects in Mali, the same. No significant benefits are expected to accrue to Koury mini-grid is not financially viable without large agro-industrial customers as most would not save sig- subsidies. While capacity expansion to integrate commer- nificantly on electricity costs by switching from margin- cial and agro-industrial loads would improve the financial ally more costly individual generators to the mini-grid. performance slightly, it is unlikely to be enough to make This is unlikely to lead to an expansion in processing the grid financially sustainable without subsidies. Some activity and thus would have little associated economic measures that could improve mini-grid performance benefits, as reflected in the slightly negative economic include implementing better load management practices NPV for the expansion project (−€18,000). However, to reduce energy storage needs, reducing administrative including additional economic, environmental, and social expenses, and enhancing revenue collection through pre- benefits that are not quantified (e.g., reduction in CO2 paid meters and remote monitoring. Despite these poten- emissions and other pollutants or reduced reliance on tial improvements, the profitability for hybrid solar-diesel imported fuels and exposure to price fluctuations) could mini-grids would require a revision of the subsidy struc- make the project economically viable with a positive ture and current tariff levels. NPV. Benefits could also accrue to the agriculture To reach financial viability while serving productive sector if it has suppressed electricity demand, which can users, capital expenditure subsidy requirements, under be met much easier through mini-grid capacity expan- assumptions for a greenfield mini-grid similar to Koury, sion rather than expansion in the size of the individual would have to reach 96 percent of a one-off capital generator. expenditure subsidy for initial development and replace- ment of major parts. With more optimistic assumptions Main Inferences and Institutional (e.g., a better load management to reduce solar PV Arrangements losses, improved revenue collection, and lower battery-­ replacement costs), the subsidy requirement could be In order for the potential large-scale opportunities to inte- reduced to 77 percent of capital investment.15 grate productive users into Mali’s mini-grids to succeed, several major barriers need to be overcome (box 5.2). Economic Analysis Available financing for rural electrification is a crucial issue for both AMADER and the mini-grid operators. Solar PV capacity expansion to supply productive users Insufficient and uncertain availability of funding for capital has limited economic benefits. For households and cost grants has limited AMADER, while private operators existing customers, the cost of supply would remain the cannot afford to scale up on their own. Box 5.2: Large-Scale Opportunities for Power-Agriculture Integration in Mali Agribusiness development in Mali could have a critical impact on job creation and poverty reduction. With over 40 million ha of arable land and an irrigation potential of 560,000 ha, Mali’s agribusiness sector could benefit from favorable agro-ecological conditions and regional food demand. But constraints along the agribusiness value chain (e.g., lack of access to energy and other basic infrastructure, lack of access to finance, and poor sector gov- ernance) limit its development. Beyond developing a value-chain strategy, a spatial approach is promoted to boost productivity growth, diversification, and value addition. Since Mali is a vast country, the creation of growth poles, clusters, and trade corridors in the agribusiness sector has real significance. In the Sikasso region, conversion of the Randgold Resources–operated Morila gold mine into an agro-industrial cluster is an example of opportunities to realize large-scale power-agriculture integration. Currently, the mine’s power demand is covered by cumulative available capacity of about 26 MW, with 187,000 MWh of potential production from 10 diesel generators. Once closed and replaced by the agropole in 2017, esti- mated power needs may drop to 8–10 MW (Randgold Resources estimate), and Randgold Resources plans (continued) 78 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Box 5.2: Continued to hybridize the generation plant and set up a mini-grid aiming to power medium-voltage agribusiness activities, including the following: ºº Henhouse (installed capacity of 130 kW with a monthly consumption of 21,000 kWh). ºº Juice production and packaging (installed capacity of 1 MW for 4,000 bottles per hour and 30–60 packets per minute). ºº Air-conditioned logistic facility (installed capacity of 20 kW with a monthly consumption between 700 kWh in freshness period and 1,250 kWh in peak season). ºº Slaughterhouse (installed capacity of 100 kW with a daily consumption of 2,200 kWh). ºº Fish preservation units (installed capacity of 200 kW per unit). ºº Carton packaging unit (installed capacity of 2.5 MW). ºº Other activities (e.g., aquaculture, mango production, and beekeeping). The mini-grid also aims to connect 100 small- and medium-sized enterprises (SMEs) that require low-­ voltage, unitary power below 30 kW for transforming and cooling crops (e.g., cereal, shea kernel, and vegetables). Powering SME activities will also facilitate the connection of 15,000 surrounding households and community facilities. This integrated solution optimizes the use of infrastructure to support large-scale agro-industry projects and secure raw materials and supply inputs through a partnership between smallholders and large players. It can also play a role in bringing rural power to the surrounding community. Source: Randgold. One way to improve the financial viability of mini- differentiated tariffs by customer type or time of use grid operators would be through diversification of the would allow operators to cross-subsidize between cus- service offering to include other energy solutions (e.g., tomer categories. Finally, access to capital for productive stand-alone systems).16 Also, clear regulations with scope users is critical. Indeed, agribusiness players willing to con- for tariff-setting flexibility would improve the ability and nect to mini-grids will have to invest in electric machinery incentives for supplying productive customers. In addition, to replace manual equipment. endnotes 1. The highly stylized setting of the model is thus less appropriate for considering such value chains as milk, poultry, and even floricul- ture, which have a different spatial distribution of production. While it is possible to adapt the model to these and other settings, it is considered beyond the scope of the present analysis and left for future work. 2. Based on the model assumptions, irrigation load demand is about one-and-a-half times that of all other power demand combined. 3. Assumes an average mill requires 35 kWh to process 1 MT of sugarcane. For other sugar estates in Africa, per hectare power demand could be significantly higher if the potential for gravity fed flood irrigation is not as high. 4. Assumes 3,000 irrigation hours per year. 5. Assumes that the same operating hours as for processing are applied and that a modern inverter driven batch centrifugal con- sumes about 1 kWh per MT of sugarcane processed. 6. Using a metric of 0.37 employees per ha and considering 4 workers per house (with no family), there are 233 houses in year 1, which rise to 1,399 houses from year 6 onward. 7. The area occupied (300 km2), average rural population density (94 people per km2), and average population growth rate (2.9 per- cent per year) are used to estimate the surrounding population. Opportunities to Harness Agriculture Load for Rural Electrification 79 8. On average, each household has 5 members; the number of households totals 5,700 in year 1, rising to 10,077 by year 20. 9. Assumes that total power demand is half that of residential demand and that nonresidential consumers use electricity roughly 4,368 hours a year. 10. A grid connection is essential for exporting power. 11. Domestic tariff is US¢2.3 per kWh. 12. 2015. 13. Analysis was done in Euro (€) currency since the local currency (CFA Francs) is pegged to the Euro, using a diesel price of 650 FCFA per liter (1€ per liter); consumption of 0.33 liters per kWh; and 20 percent in auxiliary losses, lubricants, and other main- tenance costs. 14. Assumes no need for further investments in the distribution network or additional diesel generators. 15. Assumes that integration of at least one-third of daytime commercial and industrial loads, 10 percent reduction in solar PV losses from current levels through better load management, 90 percent revenue collection, 20 percent reduction in administrative expenses, and a 20 percent reduction in battery replacement costs within the next 4–5 years due to battery technology development. 16. Partnerships with suppliers of solar pumps or solar mills could also be attractive since many operators are progressively building on an expertise in solar PV technologies. Conclusions Chapter 6 T his chapter highlights the study’s key findings gives a sense of the investment in generation capacity on Sub-Saharan Africa’s potential for leveraging that will be required to meet agricultural needs and the complementary investments in agriculture and addition to rural electricity demand that is expected owing electricity to contribute to the region’s rural pov- to the agriculture sector. erty reduction; these include overall results of the study For the 13 agricultural value chains selected, electric- and case studies, along with key learnings from the com- ity demand could increase by 2 GW by 2030, represent- mon challenges encountered by the case study projects ing nearly half of the 4.2 GW of potential incremental (chapters 4 and 5). It then recommends steps that can be increase in electricity demand from agriculture. Among taken to maximize the joint benefits of expanded electric- the value chains examined, poultry has the largest per ity access and increased value added along the agricultural hectare electricity demand. Together, maize, rice, and value chains. cassava account for 83 percent of total incremental demand in agro-processing to 2030. The largest source of electricity demand for the 13 commodities is commercial Key Findings irrigation, which has the greatest potential to develop large power loads across a range of farm sizes. Overall results This study finds that creating opportunities to piggyback Case Study Findings viable rural electrification onto local agricultural devel- The case studies show that power supply options for agri- opment depends on a variety of site-specific factors culture and rural electrification benefit from economies (e.g., scale and profitability of agricultural operations, of scale. Small-scale power systems (less than 5 MW), crop, terrain, type of processing activity, and other local which may provide a useful source of power service for conditions). Rural electrification opportunities will be agricultural processing and household connections, are best created by agro-processing activities that generate rarely financially viable without subsidies.1 When financial electricity demand close to rural population centers, gen- viability is not a key driver (or constraint), a full range of erate adequate income to cover electricity supply costs, activities can benefit from electric power. Once economic are sufficiently large in relation to household demand, and benefits are considered, a strong case can be made for have relatively low seasonal variation. providing effective subsidies to cover gaps in financial By 2030, electricity demand from agriculture is viability. estimated to double from its level today, to about 9 GW. The case studies also confirm that irrigation consti- Between 2016 and 2030, irrigation is expected to provide tutes the largest power demand from agriculture; without about three-fourths of the incremental demand (3.1 GW), it, demand from agricultural activities (except sugar with agro-processing accounting for the remainder processing) tends to be small. Large land areas are needed (1.1 GW). The overall magnitude of electricity demand to support a major irrigation load. Economic viability is 80 Conclusions81 likely for all except the most expensive sources of power because of untested procedures and lack of precedents, generation for small loads. Power supplies generate pro- notably concerning retail tariff approbation. portionally high economic value, primarily through social Another major barrier to development is the lack and indirect economic benefits. of clear electrification plans (e.g., Tanzania and Kenya). Among the agriculture schemes examined, only Information about future developments of the national large-scale development of irrigation-based agriculture grid and concession protection is crucial for dispelling and sugar estates could justify a large grid connection on a developers’ reluctance and avoiding potential friction purely financial basis. Their requirements—not all of which from tariff differences between customers. The case of are readily available in Sub-Saharan Africa—include rela- large-scale, mini-grid development in Mali shows how tively clear and empty land with good quality soils, reliable regulation and strong government buy-in can, despite supplies of sufficient water, and high quality physical and large subsidies, allow for development (chapter 5, case market infrastructure. Suitable commodities include those study 2). This example also illustrates that clear power typically cultivated on large-scale farms: maize, wheat, regulations are a necessary, but insufficient, condition for sugar, rice, soybean and barley. successful project development. For example, Tanzania’s The projects show that successful integration of agri- Mwenga mini-hydro mini-grid—one of the first projects culture and power system development requires physical of its kind to deal with regulations about water rights, land and market infrastructure to facilitate market access for access, import laws, and building permits—has entailed inputs and produce. In Zambia, for example, the strate- significant delays. This experience highlights the need to gic location of the Mkushi farming block has improved extend regulations beyond the power sector to include its development viability. The farming block is situated related sectors (e.g., trade, water, land, and environmental alongside the main T2 Highway and Tazara Railway, which management). connect Lusaka and the Copperbelt in Zambia to Tanzania For every case study analyzed, the technical and and on to the Dar es Salaam commercial port, providing financial capacity of key institutions—the utility, regula- access to markets for both inputs and produce (chapter 4, tor, and rural energy agency—to implement and permit case study 3). In Tanzania, the site of the Mwenga mini-­ development is perceived as a challenge. The weak finan- hydro generator is situated far from the main TANZAM cial status of the utilities prevents them from being able Highway between Dar es Salaam and the Zambian border; to develop financially viable projects without external sup- however, the Tunduma, Mufindi Tea Estates, which drove port. Furthermore, their cash-strapped situation increases the mini-grid’s development, is located only 10–15 km the risk of nonpayment for the power supplied by private from the main road (chapter 4, case study 2). developers, which negatively impacts project costs and Key learnings from common challenges. The main tariffs and, as a result, power affordability. If feed-in-­ barriers faced by the case study projects are linked to tariffs (FiTs) are not capped at the utility’s avoided costs, the regulatory environment, electrification planning, and the situation could worsen, further deteriorating the institutional and financial capacity. To succeed, projects utility’s viability. From the perspective of power-sector must be implemented within a stable legal environment regulators, the extra cost and delays resulting from inex- that imposes requirements and provides protection. The perience in negotiating various supply arrangements may right degree of regulation must then be found. Viewing be a hindrance to developing private power generation, the absence of regulations as an opportunity to reduce distribution, and supply. costs increases risks considerably because of uncertainty. In Tanzania, grid extension planning is generally a Light-handed regulation of small-scale electricity systems transparent and efficient process, largely included in is generally more favorable to developers and operators. the Power System Master Plan. Although grid densi- In Tanzania, the small power producer (SPP) framework fication is currently the priority for the Rural Energy allows private operators to function as power distributors Agency (REA),3 grid extension projects, such as the one and retailers, charging fully cost-reflective tariffs.2 This in Sumbawanga, are also part of the plan, considering type of regulation should tackle the economic barriers the potential economic benefits. However, TANESCO of unaffordability and uneconomic supply. In Kenya, (Tanzania Electric Supply Company Limited) has a fragile developers have been reluctant to pursue the opportunity financial situation, which has consequences for new proj- to implement electricity distribution and retail schemes ect investments. 82 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa As the mini-hydro project illustrates, dealing with the commercial farmers. Given the extra profits potentially social and environmental considerations that any project generated by a more reliable power connection, 10 large-­ of this nature raises (e.g., water resource management, scale farmers agreed to fund half of the capital costs. forestry, village lands, land acquisition, and environmental Beyond these key success factors, some hurdles still management) is still lacking in transparency and coordi- need to be overcome. The inability of national generation nation. Both the regulatory framework and the processes capacity to support higher peak load and the resulting for project development are open to political interference. load shedding create a major risk for farmers. In response, Coupled with transmission planning, generation capacity backup diesel solutions were bought to secure produc- must be developed sufficiently and consistently to support tion, and irrigation activities were carefully planned to grid extension. avoid under-voltage. Even though the irrigation project in Tanzania generally provides developers clear guidance Mwomboshi will increase peak load slightly, it will require on tariffs, concession security, and system registration; an increase in national capacity in order to reduce risks. however, the Mwenga experience shows that application Conscious about the critical role played by agriculture in of the SPP framework, particularly in setting tariff levels, Zambia’s economy, central authorities are actively intend- continues to place unnecessary pressure on developers. ing to expand the national installed generation capacity so For mini-grid developers, especially those that sell power as to limit shortages and load shedding.4 to TANESCO, the risk comes more from the off-taker. In Kenya, small-scale, private-sector renewable Late payments create financial pressure for the opera- energy projects have had little success, despite the large tor. Third-party support can therefore help by providing number of FiT applications, owing to their high devel- bridging loans. Land access, another obstacle for project opment and transaction costs. Although permits for developers, can be overcome by developing mutually sym- self-generation are straightforward and allow industrial biotic relationships with the local community and district firms, notably in the agribusiness sector, to lead renew- authorities and gaining their support. Project develop- able energy projects, it may take up to three years to ment is still a complex process. The developer, Rift Valley acquire licensing and securing of land. The power reg- Energy (RVE), expects to sign about 3,000 agreements ulator is working to streamline licensing procedures for to access land over which its network runs. projects relying on FiTs. Also, land and way-leave issues Tariff affordability for consumers continues as one can be mitigated thanks to the involvement of project of the most critical issues for mini-grid development. beneficiaries. Although RVE is free to set up its tariffs under the SPP A second major concern in Kenya is related to the framework, pressure from social and political interests private sector’s involvement in electricity distribution and continues to make it difficult to do so. The profitability supply. Currently, Kenya Power and Lighting Company of projects is therefore supported by significant capital (KPLC) is the only licensed company undertaking distri- subsidies. bution and supply activities. The regulatory framework In Zambia, favorable conditions have facilitated the is still unclear on whether other companies are legally design and implementation of the Mkushi farming block allowed to enter this business. Other obstacles concern and the Mwomboshi Irrigation Development and Support tariffs and subsidies. Although not explicitly required Project (chapter 4, case studies 3 and 4, respectively). At under the regulations, retail tariffs cannot be higher than a national level, the Mkushi grid-extension process was KPLC’s tariff schedule. This principle could jeopardize the efficient and transparent; the Zambia Electricity Supply financial viability of any small-scale initiative. Moreover, Corporation (ZESCO) led the feasibility study, with the subsidies are not available for private companies. support of a consulting company. Also, land management was clarified by the 1995 Land Act, which gave investors more visibility and reduced the risks of long-term projects. Recommended Actions to Promote In addition, some solutions were put in place to improve Power-Agriculture Integration the financial feasibility of both projects. To overcome the utility’s cash-strapped situation, the investment costs of Power utilities in Africa, like those elsewhere in the world, grid extension in Mkushi were shared between ZESCO and often focus exclusively on their own business, rarely Conclusions83 venturing outside their limited realm of expertise. But can work both ways; that is, electricity companies can a narrow institutional approach—focused only on wires, prioritize certain regions with existing or potentially high poles, and consumer billing—means that many of the levels of agricultural production, while rural development potential development benefits from electricity remain or agricultural agencies can also target areas that will be unrealized. When used by a combination of households, able to take advantage of the many possible productive commercial businesses, industry, and agriculture, electric- use impacts of electricity. The benefits of breaking down ity provides a wide array of benefits and revenue. Ignoring institutional barriers between power, agriculture, and rural these broader possibilities not only limits the possible development programs result in higher revenues for the benefits for communities and the country overall; most utility companies and higher levels of development for importantly, it neglects the potential revenue for power regions and countries. producers from the increased electricity sales. Promote Farmers’ Productivity Improve Institutional Coordination For their part, the electricity companies can promote In order to realize their full potential as providers of internal units responsible for demand-side management electricity service, power companies need to engage with and encourage the productive and efficient use of elec- related programs to develop complementary strategies. tricity. Productive use units can be responsible for pro- In the case of agriculture-power integration, this means moting the adoption of productivity enhancing machinery establishing electricity expansion strategies in collab- in agriculture, from planting to irrigation and harvest. oration with rural development, agriculture, and other Such units can coordinate with other organizations, such institutions and agencies. as farmer associations, nongovernmental organizations Such complementary strategies can take several (NGOs), and various other local- and regional-level orga- forms. One is to provide electricity to those rural areas nizations already working closely with farmers to increase with the most potential for commercial activities, which is productivity. typically the case. For example, electricity can be prior- The barriers to farmers’ productively using electricity itized in areas with a large irrigation potential, combined in rural areas are relatively easy to overcome. They typ- with access to markets for agricultural goods. Machinery ically include a lack of simple knowledge about available used in agricultural production, including small threshers, machinery, lack of a local vendor, and inability to purchase can be promoted as part of a package to encourage elec- machinery on credit. Given the high expense of using tricity use in agriculture. For areas receiving electricity diesel-powered engines for grain processing, campaigns for the first time, agricultural fairs can be set up by local could be developed by local governments to promote the governments to demonstrate the possible machinery that substitution of electricity for diesel engines among farm- can be used in agriculture. ers in areas just gaining access to electricity. In many countries of Sub-Saharan Africa, lines of Integrate Planning of Power, Agriculture, credit to farmers and other agricultural entrepreneurs and Rural Development could be augmented by local banks so as to enable the adoption of new machinery (e.g., irrigation pumps, mills, Coordination with related institutions and agencies can and small stationary threshers). In many cases, existing also benefit the electricity companies. Once a rural devel- lines of credit are mainly for seed and other supplies opment agency realizes that an area is to receive electric- provided at the beginning of the growing season, with ity, it may make plans to include those communities in loans paid off after harvest. The electricity companies its program, meaning that the region would have access could work with banks and other credit agencies to set to electricity in conjunction with other inputs important up credit lines specifically for the purchase of electric for rural development. Thus, institutional cooperation machinery. 84 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa endnotes 1. Exceptions may include hydropower and biomass. Under favorable geographical conditions, low-cost hydropower can be provided; also, biomass can support agricultural activities, but seldom beyond those of the agriculture estate. 2 Especially for systems under 100 kW, for which no approval is required from the Energy and Water Utilities Regulatory Authority (EWURA), Tanzania’s sector regulator. 3. Tanzania’s rural electrification planning is led by the REA, with the operational support of TANESCO and support of development partners. The July 2014 National Electrification Program Prospectus identified key development centers for connection to the main grid, which will not be effectively initiated before 2016. While the prospectus suggests that some flexibility in identifying additional centers could be considered in order to develop synergies between power and agriculture, such uncertainty can be unhelpful to plan- ners of rural electrification projects. 4. In addition to these technical issues, environmental considerations must be taken into account. The impacts of these projects on the environment, especially those that involve dam construction, have a non-negligible significance. Annexes Annex A: Business Models for Agricultural Development A ttaining productivity increases by focusing on villages can be linked to water and power supplies at low small-scale agriculture and small- and medium- marginal cost. In cases where nucleus farms and out- sized agribusiness enterprises, as compared grower schemes incorporate community-owned land on to larger scale commercial systems, is a major a leasehold basis, local residents can be given an equity challenge. Larger scale farming provides economies of scale share in the farming business, as well as access to low-cost in production and input supply, including finance. This is irrigation. Likewise, farmer producer associations can be particularly observable for relatively large, uneven invest- integrated into commercial value chains through out- ments (e.g., machinery, irrigation, and electricity instal- grower or contract farming models. lation) or working capital needs. Smaller farms tend to be Other evolving agribusiness models enable the less efficient when collateral requirements affect their “crowding in” of both public and private investment into ability to raise working capital (Collier and Dercon 2009). defined areas of a country. Due to economies of scale, However, this does not mean that one farming system farmers and agribusinesses are most likely to be success- should entirely preclude the other as there are examples ful when they are located in proximity of each other and of successful crop-specific, small-scale projects, partic- related service providers. Such programs as the Southern ularly in the higher value commodities. Meeting growing Agricultural Growth Corridor of Tanzania (SAGCOT) demand will require improved performance of informal is focusing initially on 5–6 clusters within the southern value chains and their linkage with formal value chains corridor where there is potential, over time, for profitable to gain much needed capital, knowledge and skills, and groupings of farming and processing to emerge.1 Each market contacts. Achieving this will require a more flexible cluster requires investment along the full agriculture value approach to farming systems, currently being evaluated, chain. Some of these investments are public goods (e.g., whereby farming is seen as a business, with small-scale rural infrastructure and electrification) that must come farmers and their communities forging stronger linkages from the government and its development partners; with modern agribusiness. The key is to ensure economies others can expect to earn a financial return and will come of scale around aggregated small-scale farmer models from the private sector (figure A.1). linked to larger commercial agribusiness. For example, Building on existing operations and planned invest- new integrated small-scale farmer models are being ments, the clusters are likely to bring together agricultural tested in northern Ghana with the development of a research stations, larger nucleus farms and ranches with commercially run, professionally managed maize farmers outgrower schemes, commercially focused farmer associa- association, Masara N’Arziki. Such small-scale farmers tions (like those described above), irrigated block-­farming associations are being developed with the technical help operations, processing and storage facilities, transport and and financial support of commercial inputs and commod- logistics hubs, and improved “last mile” infrastructure to ity marketing companies; Masara N’Arziki currently has farms and local communities. more than 10,000 small-scale members producing over When occurring in the same geographical area, these 100,000 MT of maize for local and regional markets. investments result in strong synergies across the agri- Other models that create scale include the nucleus culture value chain, helping create the conditions for a farm hub and outgrower models. These allow small-scale competitive, low-cost industry. Similar corridor programs and emergent farmers to benefit from access to infra- are operational in Mozambique (e.g., Beira Agricultural structure, including irrigation, lower cost inputs, process- Growth Corridor), while others, such as the Lakaji ing and storage facilities, finance, and markets. Adjacent Corridor in Nigeria, are still in the design stage. 88 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Figure A.1: Example of an agribusiness cluster Source: SAGCOT Investment Blueprint, AgDevCo, and Prorustica. The aim of creating simultaneous coordinated The forces driving the evolution of the design and investments can also be found in the concept of growth development of these types of programs are the demands poles. Rather than being oriented around addressing of modern agribusiness and commercial agriculture for identified market failures, growth pole projects center on new technology, finance, and logistics. To ensure their exploiting opportunities that already exist. The underlying success, larger agricultural systems are needed, be assumption about the benefits of growth poles is that they stand-alone commercial farming and agribusiness they increase market size so that it becomes profitable enterprises or those linked to business focused, integrated for firms to invest, with the resulting higher wages and small-scale organizations. All of these agricultural systems economies of scale. Notable agriculture-related growth require viable and reliable power sources. The primary pole programs include those now being developed in power requirement of commercial agribusiness clusters is Burkina Faso (e.g., Bagre Growth Pole Project) and the irrigation, which can increase yields, reduce risk, and allow Democratic Republic of the Congo (e.g., Western Growth for winter cropping and post-harvest processing and stor- Poles Project). The Western Growth Poles Project also age activities; locating these activities closer to production includes development of a special economic zone to can reduce transport costs and allow for increased value provide land equipped with critical infrastructure and a capture closer to the point of production. more conducive business environment for investors and With a focus on particular regions for agribusiness private-sector operators. development in place, the aim of governments should be Annex A: Business Models for Agricultural Development89 to encourage anchor investments that require reliable related investments into the region to exploit the sources of power. Building up a critical mass of such ­ value-chain opportunities and economies of scale. These investments should lead to a trigger point, whereby activities, in turn, will lead to opportunities to electrify investments in grid extension and cluster electrification local businesses and community customers, whose low are financially and economically feasible. Reaching this levels of power consumption would not otherwise have tipping point will allow for the “crowding in” of additional justified electrification. endnote 1. Kilimo Kwanza Executive Committee, Investment Blueprint (Dar es Salaam: SAGCOT, 2011). Annex B: Agriculture Fuels for Power Generation I n addition to providing demand for power, certain agri- crushed produces nearly 3 MT of wet bagasse. The high culture activities provide a supply of power. Agricultural moisture content of bagasse, typically 40–50 percent, is products that may be used as fuels for power genera- detrimental to its use as a fuel. For electricity production, tion can be categorized as direct burning fuels or fuels it is stored wet, and the combination of the mild exother- that are the product of chemical conversions. This annex mic reaction resulting from the degradation of residual outlines three of the more common forms of power supply sugars, along with exposure to air, light, and heat, dries the from agricultural activities. bagasse pile slightly. Bagasse is used primarily as a fuel source for sugar mills. When burned in quantity, it produces sufficient heat Biomass energy to provide both electricity and heat (including steam) to supply all the needs of a typical sugar mill, with Biomass is biological material derived from living or energy to spare. At some sites, surplus electricity is sold to decaying organisms. In the context of biomass energy, the third parties (including feeding in to main grids). term often refers to plant-based material; however, bio- mass can apply equally to animal- and vegetable-derived material. As it is growing, biomass takes carbon out of the Biogas atmosphere, and returns it as it is burned. Biomass for energy can include a wide range of materials. High-value Anaerobic digestion is a natural process, whereby plant material, such as good quality large timber, is unlikely and animal materials (biomass) are broken down by to become available for energy applications. However, microorganisms in the absence of air. The process begins resources of residues and waste could potentially become when biomass is placed inside a sealed tank or digester. available, in quantity, at relatively low cost. In the con- Naturally occurring microorganisms digest the biomass, text of Sub-Saharan Africa, the main categories include which releases a methane-rich gas (biogas) that can be agricultural residues from harvesting and processing and used to generate renewable heat and power. The remain- high-yield crops grown specifically for energy applications. ing material (digestate) is rich in nutrients, so it can be Plant-based material includes wood (sawmill waste), nut- used as a fertilizer. shells, agricultural wastes (e.g., rice husks), corn stover, A biogas plant can be fed with such crops as maize and cassava peels. silage or biodegradable wastes, including sewage sludge An assessment for the West African Economic and (animal and human) and food waste. Monetary Union (UEOMA) countries suggests that agri- Four types of technology can be used to convert cultural residues amount to about 10 metric tons (MT) of the chemical energy found in biogas into electricity. In stubble per ha of maize, 5 MT of dry matter per ha of sor- biogas conversion, the chemical energy is converted into ghum, 4 MT of straw, 2.5 MT of bran per ha of rice, and mechanical energy in a controlled combustion system. 2 MT of tops per ha of groundnut and cowpea (UEMOA The mechanical energy activates a generator, producing 2008). In many countries, these are sources for tradi- electrical power. Gas turbines and internal combustion tional, as well as modern, utilization of biomass energy. engines are the most common technologies used in this type of energy conversion. Bagasse At the village level, biogas plants can be built to con- vert livestock manure into biogas and slurry, the fermented Bagasse—the fibrous matter that remains after sugar- manure. For small-scale farmers, the technology is feasible cane or sorghum stalks are crushed to extract their for those with livestock producing 50 kg of manure per juice—is used as a biofuel in many sugar estates around day, an equivalent of about 6 pigs or 3 cows. This manure is the world. In sugar production, every 10 MT of cane collected and mixed with water and fed into the plant. Annex C: Description of Processing Activities Post-Harvest and Primary Commercial-scale mills are usually found along main Processing roads with access to national grid power supplies. Diesel power supplies are too expensive for commercial operators Cleaning drying. Many of the basic drying techniques rely to remain competitive, and other sources of power can be on solar energy through sun drying (e.g., such cereals as unreliable. In many countries, a mill may have a backup wheat and maize). Slightly more rigorous drying technolo- diesel generator to compensate for the unreliability of gies use energy input for heating boilers; this energy may national grid supplies. be in the form of electricity, but often is biomass (farm Cold storage. Control temperature storage is used waste) or liquefied petroleum gas (LPG). The latter tech- to reduce the temperature of foods and flowers post-­ niques are more common for fruits, vegetables, and meats harvest. Cooling or chilling a food product reduces the risk with a high moisture content (i.e., about 60–80 percent) of bacterial growth and allows longer storage of produce which must be reduced to a range of 10–25 percent to without spoilage.1 In principle, this process enables farmers prevent spoilage. in relatively remote locations to harvest and store pro- Milling. Mills are used for processing in the value duce for shipment to large demand centers beyond the chains of maize, wheat, and rice. Smaller mills may be local markets (including exports). A cold chain is thus a powered with diesel or electricity, and larger units with necessary asset for many high-value agricultural products electricity only. For maize, the main choice of milling is (e.g., milk and dairy products, fish and other seafood, fruit either a plate mill or hammer mill (often supplied by India and vegetables, meat and prepared foods) and high-value and China, and increasingly from local craftsmen). The horticulture and floriculture industries, especially those plate mill can grind both wet and dry products, while the that are export-oriented. Large storage hubs are often hammer mill is restricted to dry products. Hammer mills centrally located at transportation centers; however, more are the more prevalent of the two although plate mills localized facilities are often necessary since products are popular in West Africa and Sudan and operate with a deteriorate quite rapidly post-harvest and must be cooled/ greater component of shear than compression. As a rule dried or processed immediately.2 While grid power is of thumb, about 1 kW can mill 25–30 kg of produce per more cost effective, alternative energy sources, including hour. Hammer mills have a power requirement in a range solar power, can be used.3 For commodities transported of 2–50 kW, while motor-driven plate mills generally fresh to market, cooling systems are often temporary or demand less power; 0.5–12 kW is usually sufficient. Larger movable, with commodities packed straight into refriger- scale hammer mills, with a capacity of 4.5–5 MT per hour, ated reefers before being moved within days. Reefers can have a power consumption of approximately 75 kW; for be plugged into any power supply for the short term, and, fully integrated milling systems, with a capacity range once in transit, are often powered with diesel gensets. of 2.5–25 MT per hour, power demand is 120–650 kW. Cassava processing. Roots and tubers (e.g., cassava, These systems can operate year-round, often at nearly potatoes, and yams) have high moisture content, which constant rates. makes them hard to store and bulky to transport. Cassava The power demand of wheat mills ranges from 20 kW is the most perishable of the roots and tubers and can for smaller units up to 600–700 kW for larger ones. deteriorate within a couple of days of harvesting. This Small-scale rice mills can remove the hard husk and polish implies that cassava is mostly sold in processed form, and the kernel. A full rice processing production line (exclud- processing facilities and machinery need to be located at ing the polisher), with a daily output of 20–30 MT, has a relatively short distances from the agricultural lands. The total power demand of approximately 38 kW, whereas a more important traditionally processed products include processing line with polishers requires 60–90 kW. dried chips, flours/starches, and gari. Most small-scale 92 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa chippers and graters are petrol driven, with capacities of that could adversely alter food properties or deactivate 1 MT per hour and a power drive of 3.5 hp, equivalent to enzyme action and optimize the retention of certain 2.6 kW. Large-scale cassava factories are usually located quality factors at minimum cost, including such processes in the vicinity of cassava farms. as pasteurization (e.g., of milk and some fruit juices) and Meat processing. The core processing equipment sterilization. Heat exchangers are used on a wide variety consists of hoists for lifting, which can be operated manu- of products, including pasteurization of cheese, milk, ally or electrically; meat grinders; bowel cutters; cooking and other beverages; ultrahigh temperature sterilization; vats; smokehouses; and chillers. Refrigeration is generally bottled water treatment; and heating of soups, sauces, the most energy-intensive activity in meat-processing and starches. facilities. Other uses of electricity include on-site water Canning, bottling, and packaging. A growing num- pumping for washing, electrical elevators, and hoists and ber of foods are packaged to increase their shelf life. stunning guns, with scalding tanks (electrical heating) Prior to packaging (or canning or bottling), food may be for pig processing. Modern abattoirs consume energy in processed (by juicing, peeling, or slicing) to increase value livestock holding; slaughtering and processing; monitoring and prevent deterioration (through pasteurization, boiling, and testing; cleaning; and packing. refrigeration, freezing, or drying). Each of these processes Oil extraction. Oil extraction from a variety of creates demand for electricity. Packing requires electric- oilseeds (e.g., sunflower, soybean, sesame, palm oil, and ity to run machines for vacuum sealing, heat sealing, and groundnut) results in significant value addition to the final bottling; in larger facilities, electricity is needed to power product. While smaller scale extraction is done using a conveyor belts, as well as to run filling, weighing, wrapping, manual press, larger scale commercial systems use motor- boxing, coding, and printing equipment. ized presses that rely on electric input. Oil filter presses Many of Sub-Saharan Africa’s canning and bottling are used for larger, electricity-powered oil-extraction factories are situated in areas where electric power is systems for sunflower, groundnut, and soybean. Once available and reliable.4 Modern packing lines require cleaned and de-hulled, the seed is placed under increas- reliable electricity supplies to operate efficiently. As with ing pressure as it is conveyed through a tapered chamber other secondary processing plants, packaging plants are (expelling). Mini extruders, typically with a capacity of often supplied with main grid power. The power require- 125 kg per hour, require a power drive of about 10 kW, ments for juicing and canning is quite low. For example, while 400 kg per hr power requirements are approxi- a juicing machine that can process up to 5 MT of raw mately 23 kW. Capacity depends on the quality and type fruit per hour may have a peak power load of 5–22 kW. of seed (e.g., groundnut capacity is 120–180 kg per hour, A canning machine with a per-hour capacity of 250 compared to sunflower capacity of 280–320 kg per hour cans (approximately 125 kg) has a power-load range of using a similar 15–18.5 kW motor). 5.5–7.5 kW.5 Given the scale efficiencies of larger facili- ties, it is difficult to extrapolate to determine the load of a much larger commercial plant without information on the Secondary Processing capacity and power demand. Thermal treating. Thermal treating of foods (either heating or cooling) is necessary to destroy microorganisms endnotes 1. Rapid chilling—also known as flash freezing—lowers this risk even further. 2. For some products, the shelf life may be diminished by a factor of eight times the length of delay between harvesting and cooling. 3. With peak demand during daylight hours matching the generation profile of solar power, freezing systems can be switched off overnight when outside temperatures are cooler. 4. Notable canned foods prevalent in Sub-Saharan Africa include pineapple, grapefruit, and tomato. 5. References come from data on plants available for sale on Alibaba. Annex D: Maps of Case Study Project Areas Map D.1: Tanzania: Power and Agriculture in the Sumbawanga Agriculture Cluster 94 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.2: Tanzania: Mwenga Mini-Hydro Mini-Grid Annex D: Maps of Case Study Project Areas95 Map D.3: Zambia: Mkushi Farming Block 96 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.4: Zambia: Mwomboshi Irrigation Development and Support Project Annex D: Maps of Case Study Project Areas97 Map D.5: Kenya: Oserian Flowers and Harnessing Geothermal Power 98 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.6: Kenya Tea Development Agency Holdings Mini-Hydro Mini-Grids Annex D: Maps of Case Study Project Areas99 Map D.7: Ethiopia: Sugar Estates 100 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa Map D.8: Mali: Power Network and Agricultural Districts References ACET (African Center for Economic Transformation). Collier, P., and S. Dercon. 2009. “Africa Agriculture in 2013. The Soybean Agri-Processing Opportunity in Africa. 50 Years: Smallholders in a Rapidly Changing World?” Accra, Ghana and Washington, DC: African Center for Expert Meeting on How to Feed the World in 2050. Economic Transformation. Rome: Food and Agriculture Organization of the United Nations (FAO). Alexandratos, N., and J. Bruinsma. 2012. World Agriculture: Towards 2030/2050. ESA Working Paper da Silva, C., D. Baker, A. Shepherd, C. Jenane, No. 12-03. Rome: Food and Agriculture Organization of and S. Miranda-da-Cruz. 2009. Agro-Industries for the United Nations (FAO). Development. Rome: Food and Agriculture Organization of the United Nations (FAO) and United Nations Banerjee, S., Z. Romo, and G. McMahon et al. 2015. The Industrial Development Organization (UNIDO). Power of the Mine: A Transformative Opportunity for Sub- Saharan Africa. Directions in Development; Energy and Deininger, K., and D. Byerlee. 2011. Rising Global Interest Mining. Washington, DC: World Bank Group. in Farmland: Can It Yield Sustainable and Equitable Benefits? Agriculture and Rural Development. Washington, DC: Barnes, D. (ed.). 2007. The Challenge of Rural World Bank. Electrification: Strategies for Developing Countries. Washington, DC: RFF Press. Derks, E., and F. Lusby. 2006. Mali Shea Kernel: Value Chain Case Study. MicroREPORT #50. Washington, DC: ———. 2014. Electric Power for Rural Growth: How United States Agency for International Development Electricity Affects Life in Developing Countries. Second (USAID). edition. Washington, DC: Energy for Development. Development Initiatives. 2015. “Aid to the Agricultural Barnes, D., H. Peskin, and K. Fitzgerald. 2003. “The Sector in Sub-Saharan Africa Doubles, 2003–12” (http:// Benefits of Rural Electrification in India: Implications for devinit.org/#!/post/aid-agricultural-sector-sub-saharan- Education, Household Lighting, and Irrigation.” Draft africa-doubles-2003-12). paper prepared for South Asia Energy and Infrastructure. World Bank Group, Washington, DC. Diao, X., J. Thurlow, S. Benin, and S. Fan (eds). 2012. Strategies and Priorities for African Agriculture: CAADP (Comprehensive Africa Agriculture Economywide Perspectives from Country Studies. Development Programme). 2012. “About CAADP” Washington, DC: International Food Policy Research (http://www.nepad-caadp.net/about-us. Accessed Nov. 7, Institute (IFPRI). 2012). ECA (Economic Consulting Associates) and Prorustica. Cervigni, R., R. Liden, J. Neumann, and K. Strzepek. 2015. Power and Agriculture in Africa—Landscape analysis. 2015. Enhancing the Climate Resilience of Africa’s London: Economic Consulting Associates Limited. Infrastructure: The Power and Water Sectors. Africa Development Forum. Washington, DC: World Bank. FAO (Food and Agriculture Organization of the United Nations). 2005. Irrigation in Africa in Figures—AQUASTAT Chu, J. 2013. Creating a Zambian Breadbasket: “Land Survey. FAO Water Reports. FAO Land and Water grabs” and Foreign Investments in Agriculture in Mkushi Development Division. Rome: Food and Agriculture District. Land Deal Politics Initiative (LDPI) Working Organization of the United Nations (FAO). Paper 33. Brighton, UK: Institute of Development Studies. 102 Double Dividend: Power and Agriculture Nexus in Sub-Saharan Africa ———. 2009. Agribusiness Handbook—Sugar Beet White Rome: International Fund for Agricultural Development Sugar. FAO Investment Centre Division. Rome: Food and (IFAD). Agriculture Organization of the United Nations (FAO). Marshall, A. 1890. Principles of Economics. London: ———. 2014. Crop Prospects and Food Situation. No 1. Macmillan and Co., Ltd. Global Information and Early Warning System on Food Porter, M. 1990. “The Competitive Advantage of and Agriculture. Trade and Markets Division. Rome: Food Nations.” Harvard Business Review, March–April. and Agriculture Organization of the United Nations (FAO). Poulton, C., G. Tyler, P. Hazell, A. Dorward, J. Kydd, and M. Stockbridge. 2008. “All-Africa Review of Experiences Giordano, M., C. de Fraiture, E. Weight, and J. van der with Commercial Agriculture: Lessons from Success Bliek (eds.). 2012. Water for Wealth and Food Security: and Failure.” Background paper for the Competitive Supporting Farmer Driven Investments in Agricultural Water Commercial Agriculture in Africa (CCAA) Study. World Management. Synthesis Report of the AgWater Solutions Bank Group (WBG) and Investment Centre of the United Project. Colombo, Sri Lanka: International Water Nations Food and Agriculture Organization (FAO). Management Institute (IWMI). Schaffnit-Chatterjee, C. 2014. Agricultural Value Hazel, P., C. Poulton, S. Wiggins, and A. Dorward. 2007. Chains in Sub-Saharan Africa. Frankfurt: Deutsche Bank The Future of Small Farms for Poverty Reduction and Research. Growth. 2020 Discussion Paper No. 42. Washington, DC: International Food Policy Research Institute (IFPRI). Sebastian, K. 2014. Atlas of African Agriculture Research and Development: Revealing Agriculture’s Place in Africa. Hosier R., et al. forthcoming. Setting the Scene for Washington, DC: International Food Policy Research Regional Dialogue: Southern Africa Energy Water Nexus Institute (IFPRI). Issues. Washington, DC: World Bank Group. Seck P., A. Touré, J. Coulibaly, A. Diagne, and Hussain, M., K. Malik, J. Kapika, and C. Etienne. forth- M. Wopereis. 2013. Africa’s Rice Economy before and after coming. Southern Africa Energy Water Nexus Issues. the 2008 Crisis. Cotonou, Benin: Africa Rice Center. Washington, DC: World Bank Group. Staatz, J. 2011. “Enhancing Agricultural Productivity.” In IEA (International Energy Agency) and World Bank. Agribusiness for Africa’s Prosperity, edited by Kandeh H. 2015. Sustainable Energy for All 2015—Progress Toward Yumkella, Patrick M. Kormawa, Torben M. Roepstorff, and Sustainable Energy (June). Global Tracking Framework Anthony M. Hawkins, 58–86. Vienna: United Nations Report. Washington, DC: World Bank (http:// Industrial Development Organization (UNIDO). trackingenergy4all.worldbank.org/~media/ GIAWB/GTF/ Documents/GTF2105-Full-Report.pdf). Standard Bank Research. 2014. “Rise of the Middle Class in Sub-Saharan Africa” (http://www.blog.standardbank International Trade Center. 2014. “Kenya’s Cut Flower .com/node/61428). Export to Reach USD 1 Billion” (http://www.intracen.org). UEMOA (West African Economic and Monetary Union). Ivanic, M., and W. Martin. 2014. Short- and Long-Run 2008. Sustainable Bioenergy Development in UEMOA Impacts of Food Price Changes on Poverty. Policy Research Member Countries. Working Paper No. WPS 7011. Washington, DC: World Bank Group. United Nations. 2013. World Population Ageing 2013. Department of Economic and Social Affairs, Population Korwama, P. 2011. “Agribusiness: Africa’s Way Out of Division. New York: United Nations. Poverty.” Making It: Industry for Development, June 15. United Nations Industrial Development Organization World Bank. 2008. World Development Report 2008: (UNIDO). Agriculture for Development. Washington, DC: World Bank. Krugman, P. 1991. Geography and Trade. Cambridge, MA: MIT Press. ———. 2009. Awakening Africa’s Sleeping Giant: Prospects for Commercial Agriculture in the Guinea Savannah Zone Livingston, G., S. Schonberger, and S. Delaney. 2011. Sub- and Beyond. Directions in Development; Agriculture and Saharan Africa: The State of Smallholders in Agriculture. Rural Development. Washington, DC: World Bank. References103 ———. 2011a. The World Bank Group Framework and WBG (World Bank Group). 2015. Enabling the Business IFC Strategy for Engagement in the Palm Oil Sector. of Agriculture: 2015. Progress Report. Washington, DC: Washington, DC: World Bank. World Bank. ———. 2011b. Irrigation Development and Support Project ———. 2016. Enabling the Business of Agriculture 2016: (P102459) Project Appraisal Document. Washington, DC: Comparing Regulatory Good Practices. Washington, DC: World Bank. World Bank. ———. 2011c. Leveraging Investments by Natural Resource You, L. 2008. “Irrigation Investment Needs in Concessionaires. Infrastructure Policy Notes. Washington, Sub-Saharan Africa.” Background Paper 9, Africa DC: World Bank. Infrastructure Country Diagnostic. World Bank Group, Washington, DC. ———. 2013. Growing Africa: Unlocking the Potential of Agribusiness. Washington, DC: World Bank. You L., C. Ringler, G. Nelson, U. Wood-Sichra, R. Robertson, S. Wood, G. Zhe, T. Zhu, and Y. Sun. ———. 2015. Africa’s Pulse (October 2015). Washington, 2009. “Torrents and Trickles: Irrigation Spending DC: World Bank Group. Needs in Africa.” Background Paper 9 (Phase II), Africa ———. 2016a. High and Dry: Climate Change, Water, and Infrastructure Country Diagnostic. World Bank Group, the Economy. Washington, DC: World Bank. Washington, DC. ———. 2016b. Energizing Agriculture: Enhancing Efficiency in Agriculture. Washington, DC: World Bank. The majority of households and enterprises in rural Africa cope without electricity, compromising socio-economic welfare and firm productivity. Africa, characterized by low electricity consumption and ability to pay, makes rural electrification commercially unviable. Agriculture as the most important value added industry in rural areas presents a significant opportunity to improve commercial viability of grid and off- grid projects. This study explores the nexus between power and agriculture, challenges in scaling-up, and recommendations to harness this opportunity.