Document of The World Bank Report No: ICR120543 IMPLEMENTATION COMPLETION AND RESULTS REPORT (Guarantee Numbers: G-2110-KE; G-2120-KE; G-2150-KE; G-2880-KE; G-2890-KE) ON A SERIES OF IDA PARTIAL RISK GUARANTEE IN THE AMOUNT OF US$146 MILLION AND EUR 15 MILLION (US$166 MILLION EQUIVALENT) IN SUPPORT OF THIKA POWER LIMITED, TRIUMPH POWER GENERATING COMPANY LIMITED, GULF POWER LIMITED, AND ORPOWER 4, INC. FOR THE PRIVATE SECTOR POWER GENERATION SUPPORT PROJECT IN THE REPUBLIC OF KENYA June 23, 2017 Energy and Extractives Global Practice Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective December, 31 2016) Currency Unit = Kenya Shilling K Sh 1.00 = US$0.0098 US$1.00 = K Sh 102.2 FISCAL YEAR July 1–June 30 ABBREVIATIONS AND ACRONYMS ABSA Barclays Africa Group AfDB African Development Bank CPS Country Partnership Strategy CSR Corporate Social Responsibility DFI Development Financial Institution DSCR Debt Service Coverage Ratio EHS Environment, Health, and Safety EIRR Economic Internal Rate of Return EPC Engineering, Procurement, and Construction EPP Emergency Power Producer ERC Energy Regulatory Commission FCOD Full Commercial Operation Date FIFO First in - First out FM Financial Management FSA Fuel Supply Agreement GDP Gross Domestic Product GoK Government of Kenya GPL Gulf Power Limited HFO Heavy Fuel Oil ICOD Interim Commercial Operation Date ICR Implementation Completion and Results Report IFC International Finance Corporation IPP Independent Power Producer ISR Implementation Status and Results Report KEMP Kenya Electricity Modernization Project KPLC Kenya Power and Lighting Corporation L/C Loan/Credit L/D Liquidated Damage LCPDP Least-cost Power Development Plan LIFO Last in - First out M&E Monitoring and Evaluation MIGA Multilateral Investment Guarantee Agency MoEP Ministry of Energy and Petroleum MoF Ministry of Finance MSD Medium-speed Diesel NPV Net Present Value O&M Operations and Maintenance OPIC Overseas Private Investment Corporation OPSQ Operations Policy and Quality OSL Ormat Systems Limited PAD Project Appraisal Document PDO Project Development Objective PPA Power Purchase Agreement PRG Partial Risk Guarantee TPGC Triumph Power Generating Company TPL Thika Power Limited Senior Global Practice Director: Riccardo Puliti Sector Manager: Sudeshna Ghosh Banerjee Practice Manager: Pankaj Gupta Project Team Leader: Laurencia Karimi Njagi ICR Team Leader: Mariano Salto REPUBLIC OF KENYA Private Sector Power Generation Support Project A. Basic Information........................................................................................................................ i B. Key Dates .................................................................................................................................... i C. Ratings Summary ........................................................................................................................ i D. Sector and Theme Codes............................................................................................................ ii E. Bank Staff ................................................................................................................................... ii F. Results Framework Analysis ..................................................................................................... iii G. Ratings of Project Performance in ISRs .................................................................................... v H. Restructuring (if any) ................................................................................................................ vi I. Disbursement Profile .................................................................................................................. vi 1. Project Context, Development Objectives and Design ............................................................... 1 1.1 Context at Appraisal ................................................................................................................. 1 1.2 Original Project Development Objectives (PDO) and Key Indicators ..................................... 3 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification ........................................................................................................................ 3 1.4 Main Beneficiaries .................................................................................................................... 4 1.5 Original Components ................................................................................................................ 5 1.6 Revised Components ................................................................................................................ 6 1.7 Other significant changes .......................................................................................................... 7 2. Key Factors Affecting Implementation and Outcomes .............................................................. 7 2.1 Project Preparation, Design and Quality at Entry ..................................................................... 7 2.2 Implementation ......................................................................................................................... 8 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization .......................... 9 2.4 Safeguard and Fiduciary Compliance ..................................................................................... 10 2.5 Post-completion Operation/Next Phase .................................................................................. 11 3. Assessment of Outcomes .......................................................................................................... 12 3.1 Relevance of Objectives, Design and Implementation ........................................................... 12 3.2 Achievement of Project Development Objectives .................................................................. 14 3.3 Efficiency ................................................................................................................................ 18 3.4 Justification of Overall Outcome Rating ................................................................................ 20 3.5 Overarching Themes, Other Outcomes and Impacts .............................................................. 20 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops ....................... 21 4. Assessment of Risk to Development Outcome ......................................................................... 21 5. Assessment of the Guarantee in support of the Project ............................................................ 22 5.1 Impact of the Guarantee in Mobilizing Private Sector Financing .......................................... 22 5.2 Role and Value of the Guarantee in Addressing Critical Risks and Improving the Overall Sustainability of the Transaction .................................................................................................. 22 5.3 Key Issues or Events that may Arise in the Future that Could Lead to a Potential Call on the Guarantee ...................................................................................................................................... 22 6. Assessment of Bank and Borrower Performance ..................................................................... 23 6.1 Bank Performance ................................................................................................................... 23 6.2 Borrower Performance ............................................................................................................ 24 7. Lessons Learned........................................................................................................................ 24 8. Comments on Issues Raised by Borrower/Implementing Agencies ......................................... 26 Annex 1. Project Costs and Financing .......................................................................................... 27 Annex 2. Project Development Outcomes Before and After the Level II Restructuring ............. 28 Annex 3. Economic and Financial Analysis ................................................................................. 32 Annex 4. Bank Lending and Implementation Support/Supervision Processes % ........................ 36 Annex 5. Beneficiary Survey Results ........................................................................................... 37 Annex 6. List of Supporting Documents ...................................................................................... 38 Annex 7. Financial Structure of the Project .................................................................................. 39 Annex 8. Contractual Framework for Guarantee Transaction ...................................................... 41 Annex 9. Summary of the Borrower’s ICR and/or Comments on Draft ICR............................... 44 Annex 10. M&E Utilization - Summary of Key Additional Issues .............................................. 48 Annex 11. Project Map ................................................................................................................. 49 A. Basic Information Kenya Private Sector Country: Kenya Project Name: Power Generation Support Project G-2880-KE; G-2890-KE; Project ID: P122671 GU Number(s): G-2150-KE; G-2110-KE; G-2120-KE ICR Date: June 23, 2017 ICR Type: Core ICR Kenya Power Lighting IDA Partial Risk Lending Instrument: Borrower: Company Ltd; Guarantees Government of Kenya Original Total US$166 million Disbursed Amount: n.a. Commitment: Revised Amount: US$135 million Environmental Category: Full Assessment (A) Implementing Agencies: Thika Power Ltd.; Triumph Power Generating Company Ltd.; Gulf Power Ltd.; and OrPower 4 Inc. Co-financiers and Other External Partners: IFC, MIGA, African Development Bank, Overseas Private Investment Corporation (OPIC), Barclays Africa Group (ABSA), Standard Bank B. Key Dates Revised/Actual Process Date Process Original Date Date(s) 08-Oct-2012 (Thika) 27-Mar-2013 Concept Review: 23-Feb-2010 Effectiveness: 15-Mar-2012 (Triumph) 27-May-2013 (Gulf) Appraisal: 08-Nov-2010 Restructuring(s): 04-Nov-2015 Approval: 28-Feb-2012 Mid-term Review: Closing: 31-Dec-2016 31-Dec-2016 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Negligible to Low Bank Performance: Satisfactory Borrower Performance: Satisfactory i C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Project at Quality at Entry No n.a. any time (Yes/No): (QEA): Problem Project at any time Quality of Supervision No n.a. (Yes/No): (QSA): DO rating before Moderately Closing/Inactive status: Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Power 100% 100% Theme Code (as % of total Bank financing) Infrastructure Services for Private Sector Development 100% 100% E. Bank Staff Positions At ICR At Approval Vice President: Makhtar Diop Obiageli Katryn Ezekwesili Country Director: Diarietou Gaye Johannes Zutt Practice Manager: Sudeshna Banerjee / Pankaj Gupta Lucio Monari / Pankaj Gupta Project Team Leader: Laurencia Karimi Njagi Karan Capoor / Ritin Singh ICR Team Leader: Mariano Salto ICR Lead Author Mariano Salto Alessandra di Renzo / Tendai ICR co-Authors Madenyika ii F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The Project Development Objective is to increase electricity generation through Independent Power Producers (IPP) in Kenya. Revised Project Development Objectives (as approved by original approving authority) Not revised (a) PDO Indicator(s) Actual Value Original Target Formally Achieved at Indicator Baseline Value Values (from Revised Target Completion or approval documents) Values Target Years Indicator 1: Electricity generated by IPPs available in Kenya’s Interconnected Grid (GWh/yr) Conventional: 766 Conventional: 454 Value (qualitative or 0 Dropped quantitative) Renewable: 277 Renewable: 851 Date Achieved 2015 Electricity generated through independent power producers (IPPs) available in Comments Kenya’s Interconnected Grid: Conventional (65% capacity factor down to 35% in Year 5) and Renewable (88% capacity factor) Indicator 2: Direct project beneficiaries (number), of which female (%) Value (qualitative or 0 202,169 (49.5%) Dropped 257,649 quantitative) Date Achieved 2015 The number of beneficiaries is computed as an extrapolation of the calculation of the beneficiaries in the PAD. Generation capacity in Kenya stands at 7,303 GWh before the subprojects coming on line. The subprojects’ additional generation will Comments represent a percentage increase that is to be fed into the Interconnected Grid, which Kenya Power and Lighting Corporation (KPLC) reports show has a customer base of 1,444,061. Direct beneficiaries are therefore the percentage of additional generation multiplied by the number of existing customers. Generation capacity (MW) of conventional generation constructed under the Indicator 3: project Value (qualitative or 0 250 New 250 quantitative) Date Achieved 2016 Comments Indicator 4: Annual average plant availability for each plant constructed under the project (%) Thika: 99 Value (qualitative or 0 85 New Triumph: 99.1 quantitative) Gulf: 98.7 Date Achieved 2016 The indicator measures the annual average plant availability in each plant; for Comments simplicity only one indicator was included because the same indicator applies to each plant individually. iii The 85% annual average plant availability corresponds to the minimum yearly availability requirement included in each PPA (b) Intermediate Outcome Indicator(s) Actual Value Original Target Values Formally Achieved at Indicator Baseline Value (from approval Revised Target Completion or documents) Values Target Years Indicator 1 Conventional generation capacity constructed at Thika Power (MW) Value (qualitative or 0 87 n.a. 87 quantitative) Date Achieved 2014 Comments Indicator 2 Conventional generation capacity constructed at Triumph Power (MW) Value (qualitative or 0 82 n.a. 83 quantitative) Date Achieved 2016 Comments Interim commissioning date (77 MW) achieved in August 2015. Indicator 3 Conventional generation capacity constructed at Gulf Power (MW) Value (qualitative or 0 87 n.a. 80 quantitative) Date Achieved 2015 Comments Indicator 4 Conventional generation capacity constructed at OrPower 4 (MW) Value (qualitative or 0 52 Dropped 52 quantitative) Date Achieved 2014 The indicator was dropped because the IDA Guarantee for OrPower 4 was Comments canceled as part of the Level II Restructuring approved in November 2015. Indicator 5 Commissioning test for Thika Power Value (qualitative or No Yes n.a. Yes quantitative) Date Achieved 2014 Comments Indicator 6 Commissioning test for Triumph Power Value (qualitative or No Yes n.a. Yes quantitative) Date Achieved 2015 Comments Interim commissioning of 77MW achieved in August 2015. Indicator 7 Commissioning test for Gulf Power Value (qualitative or No Yes n.a. Yes quantitative) Date Achieved 2014 Comments Commissioned at the end of December 2014. Indicator 8 Commissioning test for OrPower 4 iv Value (qualitative or No Yes Dropped Yes quantitative) Date Achieved 2014 The indicator was dropped because the IDA Guarantee for OrPower 4 was Comments canceled as part of the Level II Restructuring approved in November 2015. Incremental investment in generation (equity and debt) at Thika Power (US$, Indicator 9 millions) Value (qualitative or 0 146 n.a. 146 quantitative) Date Achieved 2014 Comments Incremental investment in generation (equity and debt) at Triumph Power (US$, Indicator 10 millions) Value (qualitative or 0 157 n.a. 155 quantitative) Date Achieved 2016 Comments Incremental investment in generation (equity and debt) at Gulf Power (US$, Indicator 11 millions) Value (qualitative or 0 108 n.a. 108 quantitative) Date Achieved 2014 Comments Incremental investment in generation (equity and debt) at OrPower 4 (US$, Indicator 12 millions) Value (qualitative or 0 280.5 Dropped 280.5 quantitative) Date Achieved 2014 The indicator was dropped because the IDA Guarantee for OrPower 4 was Comments canceled as part of the Level II Restructuring approved in November 2015. G. Ratings of Project Performance in ISRs Date ISR Actual Disbursements No. DO IP Archived (US$, millions) 1 21-Oct-2013 S S n.a. 2 07-Dec-2015 S S n.a. 3 16-Jun-2016 MS MS n.a. 4 29-Dec-2016 MS MS n.a. v H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Restructuring Disbursed at Reason for Restructuring and Approved PDO Date(s) Restructuring Key Changes Made Change DO IP in US$, millions The Restructuring Paper proposed a cancellation of US$31 million of the guarantee amount under Kenya Private Sector Power Generation Support Project approved for the OrPower4 Inc.’s Olkaria III Geothermal Plant expansion project (that is, 36 MW Plant 2 and 16 MW Plant 3) as well as cancellation of the associated IDA allocation of US$7.75 million. As 4-Nov-2015 No S S n.a. a result of the cancellation, there was a change in project components and cost, a change in the financing plan, a change in implementing agency, and a change in institutional arrangements. The project Results Framework was also revised to reflect the cancellation as well as the evolution of power supply conditions in the country. I. Disbursement Profile Not applicable vi 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal Country Context 1. Kenya suffered from chronic shortfalls of electricity supply for over six years, mainly due to (a) an overdependence on hydropower (over 50 percent of the installed capacity) that made the system especially vulnerable to droughts and (b) insufficient investment in electricity infrastructure to cope with the increase in demand. Figure 1. Kenya: Installed Capacity by Technology in 2010 (%) Source: World Bank. 2. Even though the system had 30 percent installed capacity in excess to cover peak demand, 1 available capacity on any given day was much lower and could vary on a weekly basis as some of the generation units were overhauled or out of service. In many weeks, more than 250 MW of capacity was unavailable resulting in load shedding of 30 MW to 50 MW despite the 120 MW of ‘emergency’ rental diesel generators brought to alleviate the supply shortage. 3. These emergency rental generators provided up to 8 percent of the installed capacity which contributed to cover up to 10 percent of the total peak in the system and over 14 percent of the overall generation. However, they were an expensive solution to the system, with wholesale tariff of approximately 32 U.S. cents/kWh. This critical situation put Kenya Power (KPLC) under an important financial pressure and took a toll on Kenya’s gross domestic product (GDP), reducing the rate of GDP growth approximately by 1.5 percentage points in 2011, and hampered the efforts for the access expansion program. 4. To turn this situation around, the Government of Kenya (GoK) prioritized the development of a diversified generation portfolio which provides higher security of supply, lower generation costs, and lower carbon energy mix over time. A cornerstone of the expansion plan was to develop geothermal generation supported by conventional thermal generators in key sections of the system where no other technology could provide firm capacity. Under this approach, three thermal power plants were strategically located close to the industrial areas in the suburbs of Nairobi where urgent capacity needs for baseload generation were required until about 2016–2017, when they were 1 In 2010, the total installed capacity in the system was 1,473 MW, while the simultaneous peak demand was 1,107 MW. 1 expected to transition from baseload to intermediate duty (lower dispatch), as cleaner and lower cost generation from large-scale geothermal plants entered the grid. 5. Nonetheless, there were challenges to attract independent power producers (IPPs) for the implementation phase of this diversification strategy. While the sector had gone through reforms and Kenya Power had been able to attract IPPs in the past, the traditional arrangement of reducing payment risks through letter of credits financed by Kenya Power’s balance sheet was becoming too onerous for the company and the sector. Further, the GoK had limited ability to provide a sovereign guarantee due to the tight fiscal space, leaving Kenya Power and Lighting Corporation (KPLC) with few options to provide the payment security, a crucial element to attract much needed IPPs. Rationale for World Bank Assistance 6. Implementing the proposed expansion plan would result in KPLC tripling the generation capacity from IPPs, with financing requirements of almost US$1 billion within an estimated time frame of 12–18 months. Amounts of this magnitude could not be mobilized for Kenya within such a short span of time on the basis of the traditional payment security (Letter of Credit financed on balance sheet) offered by KPLC to the IPPs without some form of credit enhancement. 7. To avoid the provision of sovereign guarantees, the World Bank Group supported the GoK through a Risk Mitigation Package designed to provide a holistic risk mitigation umbrella for the development of four IPPs: Thika Power (Combined Cycle 87 MW Medium-speed Diesel [MSD]), Triumph Power (Combined Cycle 82 MW MSD), Gulf Power (Single Cycle 80.3 MW MSD), and OrPower 4 (36 MW Geothermal Expansion, with an option for a further expansion of 16 MW). The package included IDA Partial Risk Guarantees (PRGs) 2 that have been used to backstop KPLC’s payment security obligation under the Power Purchase Agreements (PPAs), in the form of a Letter of Credit; Multilateral Investment Guarantee Agency (MIGA) support in the form of termination coverage, which was provided for commercial debt only to three IPPs (Gulf Power, Thika Power, and Triumph Power), and equity cover to one of them (OrPower 4) for Transfer Restriction, Expropriation, and War and Civil Disturbance; and International Finance Corporation (IFC) loans that were provided to Thika Power and Gulf Power. The project was able to leverage US$166 million of IDA resources to mobilize US$623 million of financing, including US$357 million in private investments and commercial lending. Contribution to Higher-level Objectives 8. The project was in line with Kenya Vision 2030, where the GoK has prioritized the development of a diversified portfolio of complementary electricity assets which are expected to result in a lower cost and lower carbon energy mix over time and with the Country Partnership Strategy (CPS) for 2010–2013 (Report #52521) by expanding electricity infrastructure based on the participation of the private sector as a key partner in development. More specifically, it contributed to the achievement of Outcome 1.2 (a) improving core infrastructure, especially in roads, electricity, and water supply. 2 There has been a modernization of the guarantee instrument, and the term PRG is no longer used. The project- based guarantee instrument used in this project falls into the category of Payment Guarantee. 2 1.2 Original Project Development Objectives (PDO) and Key Indicators 9. The PDO was to increase electricity generation through Independent Power Producers (IPP) in Kenya. The project’s PDO-level results indicators are presented in table 1. Table 1. Original PDO-level Results Indicators Indicator Baseline Value Original Target Values Electricity generated by IPP available in Kenya’s Conventional: 766 0 Interconnected Grid (GWh/yr) Renewable: 277 Direct project beneficiaries (number), of which female (%) 0 202,169 (49.5%) 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification 10. Through a Level II restructuring, approved on November 4, 2015, two changes were introduced in the project: (a) removal of the IDA Guarantee allocation for Subproject 4 (Olkaria III geothermal expansion project) (refer to section 1.6 for further details on changes in the project components) and (b) adjustments to the PDO-level results indicators to properly reflect a change in the power market circumstances. 11. As defined by the Project Appraisal Document (PAD), the PDO is “to increase electricity generation through independent power producers (IPP) in Kenya.” To measure the development impact of the project, the PAD included the two PDO-level results indicators: (a) Electricity generated by IPPs available in Kenya’s Interconnected Grid and (b) Direct project beneficiaries (number), of which female (percentage). These results indicators presented some weaknesses to measure the causality relationship between the World Bank’s intervention and the specific development objective of the project for the following reasons: (a) In the first indicator, the term ‘electricity generated’ makes a direct reference to the amount of hours the plants are dispatched by the market operator. In the PAD, the indicator is calculated assuming a certain dispatch (or capacity factor) figure for thermal and geothermal plants. However, the dispatch of plants in the system is an outcome of the market, therefore affected by many other variables beyond the simple availability of a plant to be dispatched by the system operator. (b) In the first indicator, the term ‘IPPs available in Kenya’s Interconnected Grid’ makes reference to all or any IPP beyond those included under the guarantee project; nonetheless, the baseline figure was set at 0 as if there was no previous IPP experience in the country, when Kenya already had 347 MW of capacity developed by IPPs at the time of project approval.3 Therefore, it becomes difficult to understand whether the target values represent the electricity developed by IPPs as a whole or only by the new IPP developed as part of the project. (c) In the second indicator, the assumption is that (i) beneficiaries are all Kenyan consumers and they benefit equally; —this is clearly a strong assumption because 3 Private Sector Power Generation Support PAD (Report 66363-KE; paragraph 12 - main section): “The sector’s sustainability and KPLC’s reliability as an offtaker have encouraged private participation in generation: Kenya’s previous efforts to attract IPPs successfully yielded 347 MW of installed power capacity over 12 years […].” 3 utilities usually have a wide set of different customers, all of them with substantially different consumption levels and patterns; (ii) there are no technical losses in the system; thus, the total incremental amount of electricity is directly translated into customer numbers; and (iii) all the electricity is consumed in Kenya (no power exports); while the level of export is quite small in Kenya, it could have been the case that the project would benefit electricity exports rather than internal consumption, as it is impossible to track how specific electrons flow over the grid. 12. After project approval, a lower-than-expected growth in the electricity demand led to acceleration of the pace at which the thermal plants in the project became peaking plants. This situation affected the original PDO indicators because they were based on a forecasted dispatch factor (relevant at approval due to the shortages in the system) that was not representative anymore of the market conditions at the time of monitoring, creating the need for an adaptive restructuring. Both PDO-level results indicators were replaced for indicators related to installed capacity (namely ‘Generation capacity of conventional generation constructed under the project [MW]’) and plant availability (‘annual average plant availability [%] for each plant: Thika, Triumph, and Gulf); also, any reference to the geothermal component was removed to consider the cancelation of subproject 4. These indicators have the benefit of providing a more accurate way to measure the project objective without the need to rely on estimated capacity factors that cannot be univocally linked to the PDO.4 Table 2. Revised PDO-level Results Indicators Formally Revised Original Target Values (from Indicator Baseline Value Target Values PADs) Indicator 1: Electricity generated by IPPs available in Kenya’s Interconnected Grid (GWh/yr) Conventional: 766 Value Dropped 0 Renewable: 277 Indicator 2: Direct project beneficiaries (number), of which female (%) Value Dropped 0 202,169 (49.5%) Indicator 3: Generation Capacity (MW) of conventional generation constructed under the project Value New 0 250 Indicator 4: Annual average plant availability for each plant constructed under the project (%) Value New 0 85 1.4 Main Beneficiaries 13. According to the PAD (paragraph 22), the “immediate Project beneficiaries are current and prospective electricity consumers, including the poor, who face unreliable service due to generation, cost and access constraints. Additional power generated by the sub-projects will help increase productivity and spur economic growth among these beneficiaries. The GoK is also a key beneficiary since the proposed IPPs will save Kenya tens of millions of dollars in annual fuel costs alone by displacing the emergency diesel projects.” 4 The dispatch level of a plant in a system depends not only on the availability of that particular plant but also on the rest of the plants in the system (supply side) as well as on the behavior of the demand, both of them not directly affected by the project. 4 1.5 Original Components 14. The project had four subcomponents, all aimed to achieve the PDO of increasing electricity generation through IPPs in Kenya. These subcomponents were processed together because they were all at an advanced preparation stage, had similar features in terms of project scope, and benefited from a harmonized security package offered by KPLC and the GoK through the IDA PRG risk coverage. Subcomponent 1: Thika Power Plant (Total cost US$146 million equivalent, IDA PRG US$35 million and EUR 7.7 million, IDA allocation of US$11.25 million equivalent) 5 15. This subcomponent supported the construction of an 87 MW medium-speed heavy fuel oil (HFO) power plant in the Thika area, near Nairobi. The total project cost was US$146 million and financing structured on a limited recourse basis with a debt equity ratio of 75:25, with US$110 million equivalent in debt and the balance of US$36 million equivalent in equity. Senior debt was mobilized in equal amounts of US$36 million equivalent through A loans from IFC and the African Development Bank (AfDB), and a commercial tranche from Barclays Africa Group (ABSA) Capital of South Africa. Subcomponent 2: Triumph Power Plant (Total cost US$157 million, IDA PRG US$45 million, IDA allocation of US$11.25 million) 16. This subcomponent supported the construction of an 82 MW medium-speed HFO plant at Kitengela near the Athi River, approximately 25 km from Nairobi. The total project cost was US$157 million, structured on a limited recourse basis with a debt equity ratio of 75:25 amounting to around US$118 million in debt and US$39 million in equity. Standard Bank of South Africa underwrote the entire debt financing of the subcomponent. Subcomponent 3: Gulf Power Plant (Total cost US$108 million equivalent, IDA PRG US$35 million and EUR 7 million, IDA allocation of US$11.25 million equivalent) 17. This subcomponent supported the construction of an 80.3 MW single cycle, medium-speed HFO plant on land adjacent to Highway A109 connecting Nairobi to Mombasa at Athi River Town, approximately 35 km from Nairobi. The total subcomponent cost was US$108 million. The financing was structured on a limited recourse basis consisting of 25 percent equity amounting to US$27 million equivalent and the balance of the debt financing consisting of 5 percent in subordinated debt through an IFC C loan amounting to US$5 million equivalent, an IFC A loan for US$22 million equivalent, and commercial financing for US$54 million equivalent from Standard Bank split between an IFC B loan tranche and a parallel loan. Subcomponent 4: Olkaria III Geothermal Plant 2 Expansion Project - Developed by OrPower 4 (Total cost US$212 million, IDA PRG US$26 million and subsequently increased to US$31 million for the option to add Plant 3) 18. This subcomponent involved combining a 36 MW expansion with an existing 48 MW baseload geothermal power plant (‘Plant 1’) at the Olkaria geothermal fields, increasing it to a 5 IDA ‘allocation’ refers to the amount of the PRG that counts against Kenya’s IDA allocation. 5 total installed capacity of 84 MW. This involved the additional development of the geothermal field and modification of the existing plant and integration of the new plant. The subproject is located within Hell’s Gate National Park in the Kenyan Rift Valley, 90 km northwest of Nairobi. A further expansion of a 16 MW plant (‘Plant 3’) was envisaged at the same site, for which OrPower 4 had an option that was exercised later. With the addition of Plant 3, the total aggregate geothermal capacity at Olkaria III reached 100 MW. Table 3. Key Characteristics of the Project’s Subcomponents Installed Investment IDA Guarantee Subcomponent Capacity Technology Input Fuel Cost (US$, (US$, millions) (MW) millions) MSD 1. Thika Power 87 combined HFO 146.0 45.0 Plant cycle MSD 2. Triumph Power 82 combined HFO 157.0 45.0 Plant cycle 3. Gulf Power MSD single 80.3 HFO 108.0 45.0 Plant cycle 4. Olkaria III 52 Geothermal Steam 212.0 31.0 (Plants 2 and 3) Total 623.0 166.0 1.6 Revised Components 19. Cancellation of subproject 4. Through a Level II restructuring approved on November 4, 2015, IDA cancelled US$31 million of the guarantee amount under this project that was approved for the OrPower4 Inc.’s Olkaria III Geothermal Plant expansion project (that is, 36 MW Plant 2 and 16 MW Plant 3). This included the cancellation of the associated IDA allocation of US$7.75 million. As a result of the cancellation, there was a change in project components and costs, a change in the financing plan, a change in the implementing agency, and a change in institutional arrangements. The project Results Framework was also revised to reflect the cancellation as well as the evolution of power supply conditions in the country (as described in section 1.3). 20. The expansion of the Olkaria III power plant was developed under the existing PPA between KPLC and OrPower 4 mainly to accommodate the phased increase in capacity up to 100 MW. In this context, the IDA PRG for subproject 4 was required as part of the pre-existing payment security obligation included in that PPA. However, in August 2015, the GoK and KPLC notified the World Bank that OrPower 4 had decided to remove the payment security requirements under the PPA. This obviated the need for the IDA Guarantee on this subproject. Two issues were instrumental for OrPower 4’s decision to drop the payment security requirement: (a) OrPower 4 had a long-standing commercial relationship with KPLC from the initial geothermal project (Olkaria III Plant 1 was developed in early 2000) that allowed to develop enough financial strength to ‘self-insure’ the development of the additional capacity, and (b) the signature of three other IDA PRGs provided enough comfort to OrPower 4 that an IDA intervention in the event of a payment default would benefit all IPPs and therefore, the risk was prudently mitigated. 6 1.7 Other significant changes 21. There are no other significant changes beyond those presented in section 1.6. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 22. Soundness of the background analysis. The long-standing support of the World Bank in the Kenyan power sector provided the opportunity for an extensive project preparation work before design and appraisal. The power plants covered by the project were included in the least-cost power development plan (LCPDP) of the Government; both thermal power plants and geothermal Olkaria III were subjected to international competition. The latter, however, was procured in early 2000 but was amended in the interim to allow for expansion. The World Bank’s intervention was key to support adequate financing terms and risk mitigation for each of the project’s subcomponents. The project was relatively complex in nature because it included the preparation of four different subprojects, with different sponsors; financiers; engineering, procurement, and construction (EPC) contractors; and a significant number of transaction documents to be reviewed by the team. Nonetheless, the preparation of the PAD took less than a year (preparation started in April 2011 and Board approval was reached on February 2012). The PAD presented a clear perspective of the power sector and each subproject, including its main counterpart, KPLC. Risks and mitigation measures were properly addressed in the PAD. The only significant deviation from the PAD was on the expected dispatch level. At the time of the appraisal, due to emergency needs in the system, the expectations were that the dispatch was going to be high in the first few years, transitioning to the peak power plants later. This was also one of the main reasons why the GoK was strongly supportive. In addition, the World Bank also incorporated lessons from outside Kenya in the project design; in particular, the project benefited from experience in Pakistan, Jordan, Bangladesh, and Cȏte d’Ivoire. 23. Assessment of the project design. Risk mitigation was at the core of project design. Despite positive previous experience with IPPs in Kenya, the risk perceptions of international investors in the country deteriorated due to (a) the civil disturbance that followed the December 2007 crisis in the wake of upcoming elections and (b) uncertainties caused by the global economic meltdown. During preparatory activities, it became increasingly clear that (a) KPLC would not be able to attract IPPs by providing the type of payment security it had provided in the past, and (b) given the tighter macroeconomic environment, the GoK was not in a position to offer sovereign guarantees for IPPs. In response, the World Bank Group designed a comprehensive Risk Mitigation Package for the project with the objective of providing a menu of alternatives to the IPPs that would allow them to select the appropriate tools to hedge project risks. The package included IDA PRGs6 that have been used to backstop KPLC’s payment security under the PPAs, in the form of Standby Letter of Credit; MIGA support in a form of termination coverage for commercial debt and equity cover for Transfer Restriction, Expropriation and War and Civil Disturbance; and IFC loans. Also, a robust contractual structure (see Figure 2) ensured sound 6 There has been a modernization of the guarantee instrument, and the term PRG is no longer used. The project- based guarantee instrument used in this project falls into the category of Payment Guarantee. 7 allocation of risks and responsibilities to these entities. Annex 8 presents the contractual framework for the project in detail. Figure 2. Lending Structure Note: FSA = Fuel Supply Agreement; L/C = Loan/credit; MoF= Ministry of Finance; O&M = Operations and maintenance. 24. Adequacy of the Government’s commitment. The GoK demonstrated strong commitment to the project and its development objectives. This is reflected by the fact that three new IPPs and expansion of an existing one were simultaneously developed following a competitive procurement. Developing four power projects at the same time required high levels of coordination between all stakeholders in terms of preparing all the necessary documentation, technical reports, procurement procedures, legal and commercial agreements, and so on. This project could not have achieved its objectives had the Ministry of Energy and Petroleum (MoEP) and KPLC not been strongly involved in all the project stages. 25. Assessment of risks. The World Bank task team identified seven risks (political, governance, technical, timely financial closure, regulatory, sustainability, and supply-demand balance) and appropriately mitigated them. Out of the seven identified risks, three have partially occurred (timely financial closure, regulatory, and supply-demand balance); however, the mechanisms envisaged to mitigate them (mainly through key provisions in the PPA) have operated satisfactorily; for further details on financial closure issues, refer to section 2.2, and on regulatory and supply-demand balance, refer to section 2.3. 2.2 Implementation 26. While all the plants supported by the project have been commissioned, thus increasing Kenya’s installed generation capacity by 302 MW by February 2016 (250 MW of thermal and 52 MW of geothermal), the time to develop each of the plants was not equal. 27. OrPower 4 and Olkaria III geothermal power plant Plant 2 and Plant 3 were the first units to be commissioned under the project, in April 2013 and in January 2014, respectively, in line with expectations; also, since commissioning, the plants have maintained high levels of plant availability. 8 28. Thika Power Plant was the second plant under the project to be commissioned. It is considered highly satisfactorily implemented because the plant was completed and commissioned earlier than expected in the PAD and in the commercial agreements. Also, since commissioning, the plant has maintained high levels of plant availability. Key to ensure the successful implementation of Thika Power Plant was the signature and declaration of effectiveness of the IDA PRG in a few months after Board approval, allowing the company to swiftly reach financial closure with its lenders and move ahead with the construction process. 29. Gulf Power Plant, the third IPP under the project to be commissioned, was satisfactorily implemented because the plant achieved Full Commercial Operation Date (FCOD) one month ahead of the scheduled FCOD in the PPA but later than the expected date in the PAD. Also, since commissioning, the plant has maintained high levels of plant availability. A key factor affecting the implementation of the plant was the delay in reaching financial closure with its lenders. This can be attributed to the fact Gulf Power Limited (GPL) was relatively new in developing power projects—this was their first such project—and the typical extensive list of condition precedent required in the loan agreement with lenders. 30. Triumph Power Plant, the last IPP under the project to be commissioned, was moderately satisfactory implemented because the commissioning of the plant was delayed, both in terms of the PPA schedule as well as the planned schedule in the PAD. While Triumph Power Generating Company (TPGC) had signed a PPA with KPLC on June 14, 2012, and the IDA PRG was signed in December 2012 and made effective in April 2013, the achievement of financial closure was delayed. This delay was mainly attributed to an increase in the project costs as a result of extra spare parts required by the lender’s technical advisor and by the methodology to fund the Debt Service Reserve Account. Furthermore, the plant was also affected by technical issues as the EPC contractor failed to complete the construction of TPGC within the agreed time line of 14 months in the PPA and missed the Interim Commercial Operation Date (ICOD) by about 10 months (achieved on August 2015) and the FCOD by a year (achieved on February 2016). TPGC also failed its first Full Commercial Operation Test in September 2015 because the steam turbine generating unit could not achieve the required 6 MW. Nonetheless, the plant’s availability since FCOD has been close to 100 percent, as the technical issues were resolved. 31. Also, as part of its monitoring activities, in September 2016, the World Bank supported the MoEP, Energy Regulatory Commission (ERC), KPLC, and the IPPs to reach an amicable agreement on arrangements to ensure that all thermal plants in the system are given a minimum dispatch—at around 12 percent—to avoid deterioration of the plants while minimizing the level of fuel costs passed through to consumers on a monthly basis by changing the HFO pricing from first in-first out (FIFO) to last in-first out (LIFO). This change was required primarily due to declining oil prices, which was not anticipated at the time the pricing formula was agreed. The parties also agreed to support the IPPs in its claims to the National Treasury to solve the change in taxation law, but, at the time of preparing this Implementation Completion and Results Report (ICR), the issue remains open. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 32. M&E design. At the appraisal stage it was determined that the data for monitoring project outcomes and results indicators were going to be generated by the IPPs in regular progress reports 9 required under IDA’s Project Agreement with each IPP and cross-checked with the relevant reports by KPLC required under IDA’s Indemnity Agreement with the GoK as well as through regular field visits until the expiry of each PRG. As presented in section 1.3, the original results indicators could have been better designed to be closely aligned with the PDO. 33. M&E implementation. Following the M&E designed at appraisal, IDA monitored and supervised the reports submitted by the IPPs required under IDA’s Project Agreement as well as the relevant reports provided by KPLC required under IDA’s Indemnity Agreement with the GoK and through regular field visits. The World Bank task team maintained a record of the monitoring activities through Management Letters, Aide Memoires after supervision missions, and annual Implementation Status and Results Reports (ISRs).7 However, biannual ISRs were recorded only in 2016 following Operations Policy and Quality (OPSQ) guidance. 8 34. M&E utilization. The continuous monitoring activities by the World Bank allowed addressing of several issues that arose during project implementation. These issues included the need to restructure the project, as described in section 1.3, as well as (a) inefficient plant dispatch that was technically affecting the durability of the generation equipment; (b) unforeseen effects not captured on the PPA of decreasing—rather than increasing—HFO prices; and (c) a change in taxation that needed to be settled as per PPA provisions (for further details on these issues, refer to annex 10). 2.4 Safeguard and Fiduciary Compliance Safeguards 35. The regular monitoring activities from the World Bank included the analysis of IPPs’ compliance with environmental and social regulations (OP 4.01 [Environmental Assessment] triggered at appraisal stage).9 The social safeguard policies—OP 4.10 (Indigenous Peoples) and OP 4.12 (Involuntary Resettlement)—were not triggered for these projects because all the three plants are situated within the EPC zone on land that has been legally acquired and encumbered by the State for industrial development; hence, no land acquisition or economic or physical displacement was necessary as a result of the projects. For these reasons, the social issues that were monitored included labor issues, corporate social responsibility (CSR) activities, and relations with the neighboring communities. While some anomalies were detected during the construction, commissioning, and regular operation process of the plants, no major social safeguard or environment, health, and safety (EHS) issues were identified. A summary for each plant is presented in the following paragraphs. 36. Thika Power Plant. No major safeguard and EHS issues for the plant were observed, including spills issues during the last plant visit by the World Bank, and no major EHS incidents 7 In 2014, an apparent issue in the World Bank’s Operations Portal prevented the team from successfully uploading the ISR for that year. 8 OPSQ (2014) “Preparing the ISR for Investment Project Financing for ISR” recommends, at minimum, biannual report preparation. 9 OP 4.04 (Natural Habitats) was also triggered at appraisal stage but only for the Olkaria III project; since the IDA PRG for this project was cancelled as part of the project restructuring, the need for monitoring the implementation of OP 4.04 was consequently voided. 10 have occurred. Thika Power conducts ambient air quality assessment regularly and the results recorded were below the guidelines values. The IFC standard limit on oil and grease has also consistently been met. During operation, the company gave priority for unskilled labor to the local community with special preference and attention being paid to female-headed households and single mothers. Similarly, for the sake of good neighborliness, the company initiated a relationship with the local community by approaching them to understand their needs. This resulted in support to a number of primary and secondary schools as well as support to local farmers with farm equipment and inputs. At the same time, the company staff became the market for the resultant farm produce, especially assorted vegetables. 37. Gulf Power Plant. No major EHS safety incidents have been reported on the site. The power plant has carried out statutory noise and air emissions tests in November 2015 and all pollutants measured were within regulatory limits. The highest noise levels recorded were 70.2 dBA, compared to the legal limit of 72 dBA. During operation, the company hired most of the unskilled labor from the local community. Also, for the sake of good neighborliness, both with the local community and the county at large, the company implemented a number of CSR activities including a medical camp and an HIV/AIDS awareness initiative as well as support to a local primary school. 38. Triumph Power Plant. The power plant did not experience major EHS issues. During operation, some labor issues were raised by local staff against their employer, the O&M contractor, concerning poor labor practices. The complaints included delayed payment and low salaries and general poor working conditions. Triumph Power brought these complaints to the attention of the O&M contractor. These complaints are part of the reasons given by Triumph Power for requesting a change of the O&M contractor. On the other hand, the change in the O&M contractor also raised some concerns from the plant employees over the possibility of losing their jobs. During meetings with the World Bank, the company management discussed the steps that were planned to be taken to ensure that the process of replacing the O&M contractor was carried out in accordance with the labor legislation in the country. These included adequate notice to employees on the anticipated changes and on acceptable severance packages. Fiduciary 39. The nature of the IDA PRG implied that no disbursements occurred under each subcomponent as long as no payments are required under the Letter of Credit as a consequence of a repayment default by KPLC. Since there was no IDA-financed procurement under any of the project subcomponents, the financial management (FM) arrangements were restricted to the supervision of the reporting arrangements. 2.5 Post-completion Operation/Next Phase 40. The post-completion operation of the project is cemented by the offtake agreements, supported by long-term IDA PRGs. The four IPPs developed under this project have signed a 20- year PPA with KPLC for the offtake of the entire generation capacity in each plant, whereas the IDA PRGs were signed for 15 years. Price formula in the PPAs includes capacity and energy payments, therefore providing comfort to the developers in terms of isolating the IPPs’ financial 11 sustainability from market risk (the dispatch risk is something the IPPs under the current market structure in Kenya cannot hedge). 41. From the operational perspective, an acceleration in the development speed of new geothermal plants combined with a lower-than-expected growth in the electricity demand changed the expected dispatch profile of the thermal plants, displacing them toward lower places in the economic merit order. The use of the thermal plants as peaking plants in the system revealed a small issue in the PPA, as the agreement lacks clarity on the technical constraints for the dispatch of generation units. This situation, only evident at very low dispatch levels, can lead to an early deterioration of the generation units as they operate in suboptimal conditions. In September 2016, the MoEP, ERC, KPLC, and the IPPs decided to ensure a minimum dispatch quota of approximately 12 percent for these plants (as mentioned in section 2.3) as a solution to prevent early deterioration of the generation equipment, at least until an improved programming/dispatch system is implemented by KPLC. 42. The World Bank will continue to monitor the IPPs and KPLC as part of the legal requirements included in the PRGs and Indemnity Agreement. At the time of preparing this ICR, the task team was monitoring the following open issues: (a) the decision by TPGC to replace its O&M contractor; the new O&M contractor is expected to become operative on April 2017; (b) the decision of National Treasury on some changes in taxation law that had a small but negative effect on the IPPs’ finances, for which compensation is sought based on PPA provisions; and (c) a soil claim issue between Gulf Power and KPLC. 3. Assessment of Outcomes 43. Note from the ICR team. The ICR team acknowledges that the rating of a Restructured IDA Guarantee Project presents some challenges as there are no disbursements (under normal circumstances) that can be used to ‘weight’ the ratings before and after the revised development objectives, as per ICR Guidelines.10 Therefore, the ICR team made an effort to analyze and present the achievement of both development indicators, before and after restructuring, while focusing on the overall achievement of the PDO at project closure. This is particularly relevant for the case of OrPower 4, as the project component was dropped because this IPP was developed without the need of an IDA PRG despite being available, which is considered a positive outcome of the project. 3.1 Relevance of Objectives, Design and Implementation Relevance of Objectives Rating: High 44. The relevance of the PDO, that is, “to increase electricity generation through Independent Power Producers (IPP) in Kenya” was and remains high for the GoK development priorities as well as for the World Bank CPS for FY14–18. The project contributed to higher-level objectives, as defined in Kenya’s Vision 2030 for economic and social development, and is aligned with the need for infrastructure improvement necessary to support Vision 2030. The project contributed to improve the quality of electricity supply and, by displacing expensive ‘emergency’ rental power 10 OPCS. 2014. Implementation Completion and Results Report: Guidelines. Appendix B. 12 plants, promotes greater competitiveness while ensuring security of electricity supply in Kenya. The project is also well aligned with Domain 1 of the CPS—Competitiveness and Sustainability— because it leverages the private sector participation in expanding power generation infrastructure to effectively promote security of supply on cost-competitive basis. Relevance of Design Rating: High 45. The design arrangement of the project was and still is technically very appropriate and highly relevant for the achievement of the project objectives. While the project was designed to mitigate non-payment risks from KPLC, it relied on technology selection from the LCPDP for Kenya developed contemporarily at the time of project preparation and the developers were competitively procured.11 Furthermore, the project, implemented as part of a holistic World Bank Group Risk Mitigation Package that included, in addition to the IDA Guarantee, MIGA Partial Risk Insurance (PRI) (only for the three thermal IPPs) and IFC loans (only for Thika Power Limited [TPL] and GPL), provided the developers with a complete menu of alternatives to hedge project risks in a cost-effective manner for KPLC and the GoK. 46. The project was complex in nature because it included preparation of four different sub- projects, with different sponsors, financiers, EPC contractors and a significant number of transaction documents to be reviewed by the team. Nonetheless, the preparation of PAD took less than a year and presented a clear perspective of the power sector of each individual sub-project including its main counterpart, KPLC. Also, risk and mitigation measures were properly addressed in the PAD: the design of a comprehensive Risk Mitigation Package for the project was defined, with the objective of providing a menu of alternatives to the IPPs that would allow them to select the appropriate tools to hedge project risks. 47. An efficient Project design was also favored by the timely monitoring activities by the World Bank task team, that allowed adjustments of the project design (through a Level II restructuring) to a change in the circumstances, keeping the project objective fully relevant (refer to section 1.3 for further details). Relevance of Implementation Rating: Substantial 48. The Rating is justified by the three main reasons: i) although all the plants supported by the Project have been commissioned, with additional 302MW of installed generation capacity by February 2016 for Kenya, the time to develop each of the individual plants was not equal, with significant delays in the starting of operations for Triumph Power Plant; ii) the delay in reaching financial closure for Gulf Power with its lenders; and (iii) the cancelation of the US$31 million guarantee amount for OrPower4 Inc.’s Olkria III geothermal project. 49. In particular, Triumph delayed its operations, both in terms of the PPA schedule as well as the planned schedule in the PAD, mainly attributable to an increase in the project costs as a result of extra spare parts required (see Section 2.2 on Implementation for details). Moreover, the plant 11 Except for the case of the geothermal expansion plant, where OrPower 4 already had a PPA in place which was amended to include the expansion in the generation capacity. 13 was affected by technical issues as the EPC contractor failed to complete the construction of TPGC within the agrees time line of 4 months in the PPA. Still, the plant’s availability since FCOD has been close to 100%, as the technical issues were resolved. 50. Key factor affecting Gulf Power Plant’s implementation was the delay in reaching financial closure with its lenders. This can be attributed to the fact that the company was relatively new in developing power project, and to the typical extensive list of conditions required in the loan agreement with lenders. 51. In addition, timely monitoring activities by the World Bank task team allowed adjustment of the project design, through a Level II restructuring, to a change in the circumstances. Despite the project restructuring, the World Bank has been able to keep the project objective fully relevant (refer to section 1.3 for further details). 3.2 Achievement of Project Development Objectives Achievement of PDO Rating: High 52. The PDO, “to increase electricity generation through Independent Power Producers (IPP) in Kenya”, was successfully achieved at project closure with four new IPPs and 302 MW of additional installed capacity. By structuring a comprehensive Risk Mitigation Package, the World Bank brought back confidence to private investors and leveraged private capital to develop much- needed power generation infrastructure in key areas of Kenya, reducing the use of more expensive rental generators to a negligible level. 53. Together with the addition of installed capacity with consequent better reserve margin for the Kenyan power system, the leverage of private sector resources to provide a public good is an important objective reached by the Project. In particular, private sector investors and lenders were able to finance the Project due to increased level of comfort brought through the World Bank Group involvement: in this sense, the use of IDA PRG provided a clear signal on the overall soundness of the Kenyan power sector, which contributed to create a secure environment for IPPs and private financiers. 54. Before the Level II restructuring in November 2015, the PDO-level results indicators were (a) Electricity generated by the IPPs initially proposed for an IDA guarantee (Thika, Gulf, Triumph, and Olkaria IV) and measured in GWh/year produced and available in Kenya’s Interconnected Grid and (b) Direct project beneficiaries (number), of which female (percentage). In particular, the three thermal IPPs were estimated to reach a yearly power production of 1,423 GWh from year 2 from their commissioning (notably, 2013) under the assumption that until 2016 they would have been operated as baseload plants with a 65 percent dispatch rate. Thereafter, during year 6 (2016) they were supposed to be operated as intermediate and peaking plants, with 35 percent load factor, once alternative lower-cost supply sources, including geothermal, were commissioned. With regard to Olkaria IV, the Ormat IPP was supposed to be operated as a baseload plant for the entire period of the project: its annual average capacity factor was therefore not expected to vary 14 markedly during the period of the PPA but stay in the range of 88 percent, with a total electricity production of 277 GWh.12 55. However, the power supply situation in Kenya improved significantly since the project was approved in 2012, with the country overcoming the electricity deficit that has plagued the energy sector for decades. In particular, the accelerated development of new geothermal capacity enhanced the country’s energy security by reducing the contribution of installed hydropower capacity – that is susceptible to weather changes – to about 35 percent of the installed capacity versus 48 percent in 2010/2011, and to 38.7 percent to the generation mix (47 percent in 2010/2011). Table 4. Evolution of Power Installed Capacity Mix in Kenya by June 2016 (Percentage) 2010/2011 2011/2012 2012/2013 2013/2014 2014/2015 2015/2016 KenGen Hydro 48.2 46.9 46.7 43.9 36.0 35.3 Thermal 16.3 15.4 14.8 14.1 11.5 11.3 Geothermal 9.5 9.4 9.0 13.6 21.4 21.3 Wind 0.3 0.3 0.3 0.3 1.1 1.1 IPPs Hydro 0.0 0.0 0.0 0.0 0.0 0.0 Thermal 17.0 16.0 16.0 19.0 23.0 23.0 Geothermal 3.0 3.0 5.0 6.0 5.0 6.0 Cogeneration 2.0 2.0 1.0 1.0 1.0 1.0 Biomass 0.0 0.0 0.0 0.0 0.0 0.0 Aggreko - Emergency Plants 4.0 7.0 7.0 2.0 1.0 1.0 Total Power Production 100.0 100.0 100.0 100.0 100.0 100.0 Table 5. Evolution of Power Production in Mix in Kenya by June 2016 (Percentage) 2010/2011 2011/2012 2012/2013 2013/2014 2014/2015 2015/2016 KenGen Hydro 47.1 45.1 53.3 44.8 35.8 38.7 Thermal 7.1 11.0 6.9 9.2 5.3 3.5 Geothermal 14.8 14.5 13.6 13.1 33.6 36.2 Wind 0.2 0.2 0.2 0.2 0.4 0.6 IPPs Hydro 0.0 0.0 0.0 0.0 0.0 0.0 Thermal 20.4 17.3 15.1 20.3 12.9 8.8 Geothermal 5.1 5.1 6.2 9.7 10.3 10.9 Cogeneration 1.2 1.3 0.9 0.6 0.2 0.0 12 This PDO indicator for the OrPower 4 Geothermal Plant was calculated considering exclusively the commissioning of Plant 2 (36 MW), and the guarantee was planned to be issued only for this power plant, since at the time of the PAD, OrPower 4’s management was not sure of the effective realization of Plant 3. Only in March 2014, the World Bank Board was notified that OrPower 4 had decided to develop Plant 3 and that the guarantee was supposed to be issued for this facility as well. 15 Biomass 0.0 0.0 0.0 0.0 0.0 0.0 Aggreko - Emergency Plants 3.7 5.0 3.2 1.1 0.7 0.5 Import 0.4 0.5 0.5 1.0 0.9 0.7 Total Power Production 100.0 100.0 100.0 100.0 100.0 100.0 56. In this context, the use thermal plants as peaking plants greatly accelerated compared to the initial estimations in the PAD. Overall, the dispatch rate for the three IPPs thermal facilities declined from over 65 percent in 2013/2014 to less than 15 percent in 2014/2015 and 7.3 percent in 2015/2016 (see Table 6); this reduction in the dispatch was further exacerbated by a lower-than- expected growth in electricity demand (peak demand compound average growth rate of 5.8 percent between 2010 and 2016, compared to an average growth rate of over 8 percent for the power installed capacity). 57. In November 2015, the project went through a Level II restructuring that implemented two changes: (a) cancellation of the US$31 million guarantee under the Kenya Private Sector Power Generation Support Project approved for OrPower 4 Inc.’s Olkaria III Geothermal Plant expansion project (36 MW Plant 2 and 16 MW Plant 3) and (b) cancellation of the associated IDA allocation of US47.75 million. No changes were introduced to the PDOs’ definition; however, changes were introduced on the PDO-level Results indicators as discussed in Section 1.3. After the restructuring process, the PDOs are set to be achieved exclusively through the realization of the three thermal IPPs’ power supply for a total of 250 MW, supported by the IDA Project-based Guarantees, with an availability rate of at least 85 percent. Table 6. Capacity Factors for All Power Plants in Kenya (Percentage) 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 KenGen KenGen Hydro 53.2 51.2 64.0 56.5 47.3 54.1 KenGen Thermal 32.3 45.8 30.6 40.6 24.5 17.0 KenGen Geothermal 86.5 83.9 82.4 55.5 73.3 83.4 KenGen Wind 39.6 32.7 31.1 7.9 16.9 25.4 IPPs IberAfrica - Thermal 76.0 74.2 62.3 57.9 20.8 13.5 Tsavo - Thermal 56.8 43.7 27.5 23.4 12.8 6.0 Thika Power - Thermal 0.0 0.0 0.0 59.6 30.6 9.2 Triumph Power - Thermal 0.0 0.0 0.0 0.0 0.7 11.3 Gulf Power - Thermal 0.0 0.0 0.0 0.0 8.5 1.1 Rabai Power - HFO 50.0 42.9 56.2 80.3 77.2 68.0 Imenti Tea Factory - Diesel 15.2 30.4 26.6 3.8 19.0 26.6 Mumias - Cogeneration 38.2 43.9 37.7 30.3 7.4 0.0 Biojoule Kenya Limited - Biogas 0.0 0.0 0.0 0.0 0.0 1.7 OrPower 4 Geothermal - Units 1, 2, and 3 88.5 93.2 62.1 88.3 99.1 97.2 - OrPower 4 Unit 2 88.0 88.0 88.0 88.0 - OrPower 4 Unit 3 88.0 88.0 88.0 16 OrPower 4 IPP Geothermal - Unit 4 0.0 0.0 0.0 0.0 0.0 50.8 Gikira Hydro - Hydro 0.0 0.0 0.0 8.9 35.5 42.2 Aggreko Emergency Plants 50.8 36.2 24.8 35.8 24.0 19.0 58. Table 7 shows and compares the PDO-level results indicators before and after the restructuring and their corresponding level of achievement. The analysis of the indicators after the restructuring is straightforward, as the new indicators were added in November 2015; thus, there was only one year to measure them. The results show that the PDO-level results indicators of the restructured Project were fully met, ahead of the project closing date of December 2016. In particular, 250 MW of thermal capacity has been developed and commissioned and associated investment of about US$431 million made by the IPPs that is supported by the IDA guarantees of an aggregate amount of US$135 million. In detail:  Triumph Power Plant. The power plant came on line in February 2016, with significant delay compared to initial expectations; nevertheless, its availability rate in 2015/2016 reached almost 100 percent, with a load factor of around 11 percent, due to the lower-than- expected peak demand and less expensive renewable sources displacing the new thermal capacity.  Thika Power Plant. The facility was the first one to be commissioned and the only one to come on line in time. Thika reached almost 100 percent of availability rate. During 2015/2016, it reached more than 9 percent dispatch rate, from over 50 percent in 2013/2014.  Gulf Power Plant (GPL). In 2015/2016, the power plant reached a very high availability rate (98 percent), compared to the end-of-project target value of 85 percent contained in the PAD after the project restructuring. Nevertheless, the average dispatch level for 2015/2016 for the facility was 1.1 percent compared to 8.5 percent in 2014/2015, given the plant being disadvantaged by dispatch preference for cheaper geothermal and hydro energy, during the last three years. Moreover, GPL ranked even lower in the economic merit order than other thermal plants with combine cycle units. Table 7. Achievement of PDO-level Results Indicators Unit 2012 2013 2014 2015 2016 Current Result Level Indicators Generation capacity (MW) Target MW N/A N/A N/A N/A 250 of conventional generation constructed by the project Achieved MW 0 0 87 244 250 Target % N/A N/A N/A N/A 85% Annul Average Plant Thika % 0% 0% 75% 95% 99% Availability Triumph % 0% 0% 0% 8% 99% Gulf % 0% 0% 0% 97% 99% Dropped Result Level Indicators Target (Thermal) GWh - 1,423 1,423 1,423 766 Target (Geothermal) GWh - 277 277 277 277 17 Total Target GWh - 1,700 1,700 1,700 1,043 Electricity generated through Thika + Gulf + Triumph GWh - - 454 298 160 IPPs available in Kenya's OrPower4 (Units 2 & 3) GWh - 69 339 401 401 interconnected grid Total Achieved - 69 793 699 561 Total Generation Potential - 69 911 1,864 2,569 Total Target - 332,134 332,134 332,134 202,169 Direct project beneficiaries Total Achieved - 13,644 156,804 138,217 110,930 (number) Total Generation Potential - 13,644 180,056 368,480 508,002 Source: World Bank estimations. Note: The 85 percent annual average plant availability corresponds to the minimum yearly availability requirement included in the PPA and high initial figures are observed simply because the equipment is brand new; ‘Generation potential’ is calculated using the installed capacity (MW) and the availability from each plant to obtain the maximum amount of GWh these plants could have generated without dispatch constraints; ‘Potentially Achieved’ is calculated using the ‘Generation potential’ figures. 3.3 Efficiency Rating: Substantial 59. Before the commissioning of the IPPs, KPLC had to rely on Emergency Power Producers (EPPs)13 to provide peak load and reserve margin. Nonetheless, despite these emergency rental generators, about 250 MW of capacity was not available in any even week, resulting in load shedding of 30 MW to 50 MW. Between July 2010 and July 2013—just before the commissioning of Thika Power Plant—KPLC purchased an average of 303 GWh per year from EPPs, around 5 percent of the annual national consumption. 60. During the first year of commissioning, Thika Power allowed a 41 percent decrease in fuel cost charges compared to those from EPPs (Figure 3). After the commissioning of the IPPs supported by the project, the contribution of EPPs decreased to 50 GWh (0.6 percent of total annual electricity consumptions) in 2016. 13 This concept involves short-term rental of small-/medium-size high-speed diesel generators inclusive of O&M service. This generation equipment usually presents lower thermal efficiency compared to other conventional generation alternatives. 18 Figure 3. Fuel Costs Comparison for Thika Power versus EPP Source: World Bank, based on MoEP data. 61. As a rule of thumb, the net monetary benefits arising out of the project can be estimated through an avoided cost analysis, that is, by multiplying the average fuel savings per kWh reached during fiscal year 2013/2014 to the total power generation from the thermal IPPs promoted by the project between 2014 and 2016. Table 5 shows that almost US$110 million in fuel costs have been saved due to the replacement of EPP by the IPPs in the project. Table 6. Fuel Cost Savings due to the Implementation of the Three Thermal IPPs 2013/2014 2014/2015 2015/2016 IPP Power Generation (GWh) 454 298 160 Total estimated savings (US$, millions) 54.4 35.7 19.2 Source: World Bank estimations. 62. In the appraisal stage, the economic analysis resulted in a base case economic internal rate of return (EIRR) of 25 percent and a net present value (NPV) of US$288 million for the three thermal IPPs and EIRR of 18 percent and an NPV of US$79 million for OrPower4 (the analysis did not provide a combined EIRR and NPV for the entire project). 63. In the ICR stage, a revised economic analysis was carried out following the methodology developed at appraisal but assuming that the thermal IPPs will operate as peaking power plants running at the minimum capacity factor of 12 percent agreed with KPLC and ERC, since it is expected that lower costs supply sources will continue to maintain thermal capacity in the margin (this is a conservative assumption as current dispatch level is close to 20 percent; future plant dispatch is largely dependent on hydrological conditions in Kenya) while the geothermal plant maintains its historical average capacity factor in the future. The results show that the project has an economic benefit (NPV) of US$66 million (at 12 percent discount rate) and an EIRR of 14 percent; these lower results compared to the appraisal stage mainly reflect the fact that the IPPs are expected to provide a lower capacity factor than initially estimated. 19 3.4 Justification of Overall Outcome Rating Rating: Satisfactory 64. Based on the assessment of sections 3.1 to 3.3, the overall outcome is rated as Satisfactory. The analysis showed that the project (a) is highly relevant (both at ICR and at approval) to achieve the GoK national development strategy (enshrined in Vision 2030) as well as with the World Bank’s CPS FY14–18; (b) has achieved the PDOs by supporting the development of 302 MW of installed capacity deployed by four private developers; and (c) by replacing expensive diesel rental plants, has achieved the PDOs efficiently. 65. Note from the ICR team. The ICR team acknowledges a mismatch between the Satisfactory rating given to the overall outcome in this ICR and the Moderately Satisfactory development outcome rating of the last two ISRs in 2016. The mismatch reflects the fact that the task team used the ISR ratings as a tool to bring to the GoK’s attention the delay in the commissioning of the Gulf and Triumph power plants as well as some other minor issues (like the deterioration in the generation equipment due to low dispatch level or the changes in taxes). This approach contributed to influence the World Bank’s sector dialogue and to support the long-term sustainability of the operation. While the task team used the ISR rating tool to signal ongoing issues to World Bank management and the GoK, the ICR evaluation and rating assessment is more adequate to reflect, at closure, (a) the complete achievement of the project’s objectives, (b) the high relevance of the project, and (c) the substantial efficiency in the achievement of the project’s objectives. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 66. The project has had a beneficial impact on poverty and social development through the creation of new jobs by the new thermal IPPs; the total new direct jobs added by Thika, Gulf, and Triumph are between 150 and 160 headcounts in total. In addition, the project supported the development of local communities by prioritizing the employment of local people (including females), promoting economic activities, and supporting schools and sanitary centers. 67. Also, by increasing cheaper generation capacity in a constrained system, the usual practice of load shedding was concluded and the overall reliability of the power sector was enhanced, effectively promoting the competitiveness of the Kenyan economy and benefiting all customers connected to the grid which would have been supplied by more expensive rental power plants or, worse, not supplied at all. (b) Institutional Change/Strengthening 68. At the time of project preparation, KPLC had limited ability to provide balance sheet- financed Letters of Credit and a GoK sovereign guarantee was envisaged as a potential solution to attract private investments in the sector. Instead, supported by the IDA guarantee and MIGA insurance, the GoK was able to only issue a Government Letter of Support, whereby it only backstops KPLC’s termination payment under the PPA and certain ongoing payment obligations. This option obviated the need for the GoK to provide a sovereign guarantee to both lenders and investors, for the full duration of the PPA and limited its provision to only counter-guarantee for 20 IDA guarantees, thereby substantially reducing its contingent liabilities. This innovative approach changed the way in which private sector interacts with state-owned utilities as the use of IDA PRG not only contributed to reduce contingent liabilities on the Government side but also provided a clear signal to the private sector on the overall soundness of the Kenyan power sector and on the economy as a whole. 69. In addition, by implementing the World Bank Group Risk Mitigation Package, KPLC was able to strengthen its profile as a credible offtaker, by either reducing the cost of providing payment guarantees or by avoiding them altogether, as in the OrPower 4 case, a situation that is almost unprecedented in Africa. This project also contributed to solidify the importance of financial sustainability across KPLC Senior Management which was subsequently consolidated with the successful implementation of the KPLC debt restructuring process supported by an IDA Guarantee (implemented under the KE Electricity Modernization Project [KEMP] - P120014). (c) Other Unintended Outcomes and Impacts (positive or negative) 70. The decision by OrPower 4 to drop payment security requirement in the PPA, despite being available, is quite unprecedented in Africa and created a positive milestone for KPLC in terms of business credibility. It also shows that IDA Guarantees have the ability to exceed the boundaries of the borrowers and provide comfort to private investors not directly covered by them. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 71. No beneficiary survey was carried out for the preparation of this ICR. 4. Assessment of Risk to Development Outcome Rating: Negligible to Low 72. Overall, the risk that the current development outcome cannot be maintained in the future is considered low, on the back of the continuous necessity of reserve margin in the Kenyan power system. The main factors that could hamper the achievement of the development objective comprise the following: (a) default on payments from the energy purchased by KPLC, (b) nationalization of IPPs, and (c) other significant breaches in the contract. 73. The risk of a potential call on the IDA Guarantee due to payment default from KPLC is considered low. KPLC financial performance has significantly improved since the debt restructuring process backed by an IDA Guarantee.14 The constant monitoring of KPLC finances by the World Bank team as part of this guarantee provides a strong risk mitigation for any default on payments that may affect the IPPs. In addition, the Kenyan power sector has promoted private sector participation in the power sector since 1996, where sectoral reforms introduced in the power sector over the years have been gradually enhancing private sector participation in power; the new Energy Bill, under development in the Parliament, seeks to enhance competition in the power sector for which additional private sector participation could be expected in the future. 14 Implemented through the KEMP (P120014), the debt restructuring process allowed KPLC to (a) reduce the weighted average cost of debt by approximately 1 percent; (b) double the weighted average loan maturity; and (c) comply with financial covenants, particularly the debt service coverage ratio set at 1.2x. 21 74. Finally, another risk element to development outcome could stem from some changes made in taxation law (refer to section 2.5 for further details). These changes have affected all power supply and other operators in the sector. However, KPLC and the MoEP have petitioned the National Treasury to review these tax arrangements for the energy sector. Although the current compensation claims for the three thermal IPPs are not substantial and the risk is estimated as low, changes in law and taxes present a risk to the financial integrity of the IPPs if they place substantial financial burden that is not mitigated by the Government in a timely way. 5. Assessment of the Guarantee in support of the Project 5.1 Impact of the Guarantee in Mobilizing Private Sector Financing 75. The impact of IDA support to mobilize private sector financing is considered high. Three out of the four IPPs were developed using IDA PRG (as well as other risk mitigation alternatives included in the World Bank Group Risk Mitigation Package) while the fourth one (OrPower 4) was developed without any payment security requirements under the PPA (despite being available), owing to its long-standing commercial relationship with KPLC, creating a positive precedent for KPLC in terms of business credibility. 76. The project was able to leverage US$135 million of IDA resources to mobilize US$623 million of financing, including US$357 million in private investments and commercial lending. The equity contribution was US$149 million funded by private project sponsors from their own resources. The total debt financing amounted to US$474 million, which was mobilized through IFC A, B, and C loans amounting to US$90 million, commercial loans of US$208 million (including the IFC B loan of US$27 million) from commercial lenders such as ABSA and Standard Bank (both of South Africa), with the balance of US$202 million from Development Financial Institutions (DFIs) such as Overseas Private Investment Corporation (OPIC) and the AfDB. 5.2 Role and Value of the Guarantee in Addressing Critical Risks and Improving the Overall Sustainability of the Transaction 77. IDA supported the investors and lenders to make generation project financially viable by underpinning its cash flow through a payment risk guarantee. The guarantee support also facilitated IDA’s ongoing involvement in the sector, through regular monitoring and mediation, ensuring amicable solutions to the issues that arose during implementation (for further details, refer to annex 10). 5.3 Key Issues or Events that may Arise in the Future that Could Lead to a Potential Call on the Guarantee 78. The ICR team did not find any additional issue other than the ones presented in section 4. 22 6. Assessment of Bank and Borrower Performance 6.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory 79. The PAD was well prepared with a detailed assessment of the situation of the power sector in Kenya and of the investments in the three thermal IPPs’ power plants and the Olkaria 4 Inc. geothermal power plant. The World Bank team was able to correctly identify the potential risks of the operation and worked together with IFC and MIGA to prepare an integral risk mitigation package that was satisfactory to effectively leverage the private sector participation. Only minor shortcomings were identified in the selection of PDO results indicators that were later replaced as part of the project restructuring (refer to section 1.3 for further details). (b) Quality of Supervision Rating: Satisfactory 80. The quality of the supervision is rated as Satisfactory. The regular monitoring activities allowed the task team to identify the need for a Level II restructuring of the project that was approved in November 2015, with the resultant reallocation of the IDA Guaranteed component for OrPower 4 to the Kenyan energy portfolio, as well as the replacement of the PDO-level results indicators. In addition, effective World Bank supervision activity contributed to facilitate the dialogue between the IPPs, KPLC, ERC, and the MoEP, allowing them to address the technical and regulatory issues affecting the IPPs and promoting the search of amicable solutions (as described in section 2.3). 81. Only minor shortcomings were identified in the file of monitoring activities. During the implementation process, the project team recorded the follow-up of the project progress and results achievement through regular Management Letters, Aide Memoires, and annual ISRs. 15 Biannual ISRs were recorded only in 2016. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory 82. Overall, the World Bank’s performance has been rated as Satisfactory. The World Bank solidified Kenya’s experience with IPPs and provided comfort to the market that KPLC was a credible offtaker and regulatory systems were responsive to evolving situations. 83. It has been noted that only annual ISR reports were provided until 2016; these are considered minor shortcomings since OPSQ guidance for ISR preparation recommends, at minimum, biannual report preparation. 16 Nonetheless, the World Bank supervision during the period remained sound. The World Bank task team, through its proactive engagement, ensured 15 In 2014, an apparent issue in the World Bank’s Operations Portal prevented the team from successfully uploading the ISR for that year. 16 OPSQ. 2014. Preparing the ISR for Investment Project Financing. 23 smooth implementation of the project. Moreover, any adverse issues that emerged during the project implementation and operation were diligently addressed by the World Bank task team. 6.2 Borrower Performance (a) Government Performance Rating: Satisfactory 84. Overall, the GoK’s performance in the project is rated as Satisfactory. The Government, represented in the project through the National Treasury, the MoEP, and KPLC, generally fulfilled its obligations and created the environment for the IPPs to operate efficiently. When issues emerged during the project, the MoEP and KPLC were willing to discuss and find a workable solution with the interested entities. Overall, the entities have shown positive commitment in the development of the new power capacity in Kenya and attention in supporting the three IPPs with some issues that emerged during the past year, particularly in the case of equipment deterioration due to low dispatch, fuel pass-through formula, and the change in the taxation law, that affect them. The fact that some of these issues remain unsolved at the time of preparing this report prevents the ICR team from providing the highest ranking for this category. (b) Implementing Agency or Agencies (Borrowers) Performance Rating: Satisfactory 85. The performance of the implementing agencies in the project is rated as Satisfactory. The project developers demonstrated competence in doing professional work to develop the IPPs, continuously cooperating with the task team, from project design to implementation, to ensure the project achieves its development objectives. Only minor shortcomings were identified, particularly on the delays in the commissioning of the Gulf Power and Triumph Power plants compared to the estimated dates in the PAD. These delays, while not affecting the achievement of the PDO, adversely affected the ability of the power sector to reduce the cost of fuel supply in the electricity tariffs further than initially estimated. (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory 86. The overall rating for the borrower performance is Satisfactory based on the ratings for the Government and implementing agency performances, as discussed above. 7. Lessons Learned 87. The implementation of the World Bank Group Risk Mitigation Package transformed the way in which the private sector interacts with state-owned utilities. Instead of using full GoK sovereign guarantee to enable much-needed private investments in the power sector, KPLC, through a combination of World Bank Group instruments and the GoK Letter of Support, was able to successfully conclude all four transactions. This arrangements benefited all parties—KPLC was able to meet payment security requirement under the PPA without using cash collateral, which it needed for other operational purposes; private sector investors and lenders were able to finance the project due to increased level of comfort brought through the World Bank Group involvement; and last but not least, the GoK was able to reduce its contingent liabilities since it only backstops KPLC’s termination payment under the PPA and certain ongoing payment obligations, thus 24 limiting its provision to only counter-guarantee for IDA guarantees. The use of IDA PRG provided a clear signal to the private sector on the overall soundness of the Kenyan power sector. The beneficiaries of the IDA PRG exceeded the boundaries of the PRG beneficiaries, providing comfort also to private investors not directly covered by them, as observed in OrPower 4. 88. The project also showed that a well-structured risk mitigation package with complementary World Bank Group instruments can effectively leverage the private sector, allowing a more efficient use of scarce public funding. It also assisted KPLC to cement its commercial relationship with developers, allowing the company to avoid any payment guarantee requirement in a PPA, something considered unprecedented in Africa. This project also contributed to solidify the importance of financial sustainability across KPLC Senior Management. 89. While the reduction in the dispatch of thermal plants compared to the expected scenarios in the PAD may have raised uncertainties over the economic feasibility of the project in the short term, the revised economic analysis has confirmed the importance of peaking plants in the system. This is particularly relevant now as Kenya enters a drought period that limits hydropower generation, where the role of the IPPs as peaking plants in the system has become evident. Unlike the past, Kenya now has plants readily available to provide generation capacity to support the system when required. 90. In the longer term, as the Kenyan power sector prepares to harness additional generation from intermittent renewable sources (particularly wind and solar), the role of these IPPs will be expanded not only to cover midterm variability on hydro resources but also short-term intermittency, typically observed in solar or wind farms. 91. The project was developed during an emergency in the system and the PAD clearly measured the negative impact of load shedding and emergency rental generators on the Kenyan economy. Furthermore, even with all the support received from the GoK and IDA, some plants suffered commissioning delays, prolonging the negative impact of the emergency on the consumers. Therefore, the need for continuous power sector monitoring and planning is a necessary condition to mitigate the risk that emergency situations became the drivers of generation investment in the country. To this extent, the availability of an LCPDP was key to identify where, when, and which technology was needed to be developed to overcome the most significant technical challenges in the power sector. It also showed that, unless regularly updated, the outcomes of the planning exercise can become quickly outdated, particularly in the cases of developing economies where the changes in demand parameters can substantially change across the years. A more regular planning update would have contributed to enhancing the development of the power system, potentially mitigating the risk of an emergency in the system or the dispatch issues observed in the IPPs. 92. Active monitoring of the project implementation by the World Bank task team has been key to project development. Identification of issues at the right time allowed the World Bank to approve a Level II restructuring in due time. Also, stakeholders in the project have praised the continued engagement by the World Bank staff to act as honest brokers, promoting the dialogue on the parties and reaching for consensus and amicable solutions to some of the issues that arose after project commissioning. 25 8. Comments on Issues Raised by Borrower/Implementing Agencies 93. The MoEP, on behalf of the GoK, reviewed the draft version of the ICR and commented that, (a) The report is consistent with the prevailing situation in Kenya; (b) The 250 MW developed through the project has continued to meet the project development’s objective in providing reserve capacity and enhancing energy security, particularly in the current drought situation that Kenya is experiencing; and (c) The National Treasury and the MoEP have agreed on the reimbursement schedule for the additional costs incurred by IPPs due to a change in tax law, as discussed in paragraph 31 (and annex 10) of this report. 26 Annex 1. Project Costs and Financing (a) Project Cost by Component (in US$, millions equivalent) Appraisal Estimate Actual/Latest Estimate Percentage of Components (US$, millions) (US$, millions) Appraisal Subproject 1: Thika Power Plant 146.00 146.00 100 Subproject 2: Triumph Power Plant 157.00 157.00 100 Subproject 3: Gulf Power Plant 108.00 108.00 100 Subproject 4: Olkaria III Geothermal 212.00 0.00 0 Plant 2 Expansion Total Baseline Cost 623.00 411.00 66 Total Financing Required 384.00 — — (b) Financing Appraisal Actual/Latest Type of Percentage of Source of Funds Estimate (US$, Estimate (US$, Cofinancing Appraisal millions) millions) Borrower Equity 149.00 102.00 68.5 Commercial Loans Loan 181.00 181.00 100.0 DFIs Loan 293.00 128.00 43.7 27 Annex 2. Outputs by Components Status of Outputs 1. In November 2015, the project went through a Level II restructuring that implemented two changes: (a) cancellation of US$31 million of guarantee amount under the Kenya Private Sector Power Generation Support Project approved for OrPower 4 Inc.’s Olkaria III Geothermal Plant expansion project (36 MW Plant 2 and 16 MW Plant 3) and (b) cancellation of the associated IDA allocation of US47.75 million. 2. As a result of the cancellation, there were no changes to the PDOs’ definition (“ to increase electricity generation through Independent Power Producers [IPP] in Kenya ”), no new safeguards policy was triggered, and there was no extension of the closing date for the project. However, considering the new market context, the two PDO-level results indicators based on electricity units generated from the power plants and dispatch levels for the four IPPs did not adequately capture the impact of the project during the operational phase. 3. Therefore, the project Results Framework was revised to reflect the cancellation of the guarantee for OrPower 4 Geothermal Plant 2 and Plant 3 as well as the evolution of the power supply conditions in the country. In particular, the PDO indicators “Electricity generated through IPPs available in Kenya’s interconnected system - Conventional (65 percent capacity factor to 35 percent capacity factor in Year 5), Renewable (88 percent capacity factor)” and “Direct project beneficiaries (number, of which female (percentage)” were no longer relevant and were removed as the IDA Guarantee was no longer needed for these expansions. 4. The following PDO indicators, which better capture the project’s progress toward achieving the PDO, were included after the restructuring process: “Generation capacity of conventional generation constructed under the Project (MW)” and “Annual Average Plant Availability (percentage) for each plant: Thika, Triumph and Gulf”. 5. Thus, the PDOs are set to be achieved exclusively through the realization of the three thermal IPPs’ power supply for a total of 250 MW, supported by the IDA Project-based Guarantees, with an availability rate of at least 85 percent. As shown in Table 2.1, all the new project indicators have substantially been achieved ahead of the project closing date of December 31, 2016. In particular, 250 MW of thermal capacity has been developed and commissioned and associated investment of about US$431 million made by the IPPs that is supported by the IDA guarantees of an aggregate amount of US$135 million. Below, a detailed analysis of the outcomes of each Subcomponent is carried on. 28 Table 2.1. PDOs before and after the Project Restructuring (November 2015) Year 1 Year 2 Year 3 Year 4 Year 5 2012 2013 2014 2015 2016 Before Project Restructuring PDO-level results indicators Total expected production (GWh) - Thermal IPPs (Thika, Triumph, Gulf) 0 1,423 1,423 1,423 766 IPPs’ Load Factor 0% 65% 65% 65% 35% - Renewables (OrPower 4) - only Unit 2 0 277 277 277 277 OrPower 4 Load Factor - only Unit 2 0% 88% 88% 88% 88% After Project Restructuring - November 2015 PDO-level results indicators restructuring Generation capacity for the three IPPs (MW) — — — — 250 Annual average plant availability for IPPs (%) — — — — — Thika Power — — — — 85% Triumph Power — — — — 85% Gulf Power — — — — 85% Effective Available Capacity (MW) Thermal IPPs (Thika, Triumph, Gulf) (MW) 0 0 87 244 250 Thika Power 0 0 87 87 87 Triumph Power 0 0 0 77 83 Gulf Power 0 0 0 80 80 Renewables (OrPower 4) 0 36 52 52 52 OrPower 4 Olkaria Unit 2 0 36 36 36 36 OrPower 4 Olkaria Unit 3 0 0 16 16 16 Effective Power Production Thermal IPPs (Thika, Triumph, Gulf) (MW) 0 0 454 298 160 Thika Power 0 0 454 233 70 Triumph Power 0 0 0 5 82 Gulf Power 0 0 0 60 8 Thermal IPPs’ Load Factors — — — — — Thika Power 0% 0% 60% 31% 9% Triumph Power 0% 0% 0% 1% 11% Gulf Power 0% 0% 0% 9% 1% Renewables (OrPower 4) (MW) 0 69 339 401 401 OrPower 4 Olkaria Unit 2 0 69 277 277 277 OrPower 4 Olkaria Unit 3 0 0 62 123 123 OrPower 4 Olkaria Load Factors — — — — — OrPower 4 Olkaria Unit 2 0% 88% 88% 88% 88% OrPower 4 Olkaria Unit 3 0% 0% 88% 88% 88% 29 Subcomponent 1: Thika Power Plant (Total cost US$146 million equivalent, IDA PRG US$35 million and EUR 7.7 million, IDA allocation of US$11.25 million equivalent) (i) Original Scope 1. This subcomponent supported the construction of an 87 MW medium-speed heavy fuel oil (HFO) power plant in the Thika area, near Nairobi. The total project cost was US$146 million and financing structured on a limited recourse basis with a debt equity ratio of 75:25, with US$110 million equivalent in debt and the balance of US$36 million equivalent in equity. Senior debt was mobilized in equal amounts of US$36 million equivalent through A loans from IFC and the African Development Bank (AfDB), and a commercial tranche from Barclays Africa Group (ABSA) Capital of South Africa. (ii) Outputs 6. The facility, which is today one of the most efficient diesel power plants in Kenya, was the first one to be commissioned and the only one to come on line in time. Thika reached almost 100 percent of availability rate. During 2015/2016, it reached more than 9 percent dispatch rate, from over 50 percent in 2013/2014. Subcomponent 2: Triumph Power Plant (Total cost US$157 million, IDA PRG US$45 million, IDA allocation of US$11.25 million) (i) Original Scope 7. This subcomponent supported the construction of an 82 MW medium-speed HFO plant at Kitengela near the Athi River, approximately 25 km from Nairobi. The total project cost was US$157 million, structured on a limited recourse basis with a debt equity ratio of 75:25 amounting to around US$118 million in debt and US$39 million in equity. Standard Bank of South Africa underwrote the entire debt financing of the subcomponent. (ii) Outputs 8. Triumph Power Plant came on line in February 2016 with significant delay compared to initial expectations; nevertheless, its availability rate at the end of June 2015/2016 reached almost 100 percent, with a load factor of around 11 percent, due to the lower-than-expected peak demand and the change in market conditions, with less expensive renewable sources displacing the new thermal capacity. Subcomponent 3: Gulf Power Plant (Total cost US$108 million equivalent, IDA PRG US$35 million and EUR 7 million, IDA allocation of US$11.25 million equivalent) (i) Original Scope 9. This subcomponent supported the construction of an 80.3 MW single cycle, medium-speed HFO plant on land adjacent to Highway A109 connecting Nairobi to Mombasa at Athi River Town, approximately 35 km from Nairobi. The total subcomponent cost was US$108 million. The financing was structured on a limited recourse basis consisting of 25 percent equity amounting to 30 US$27 million equivalent and the balance of the debt financing consisting of 5 percent in subordinated debt through an IFC C loan amounting to US$5 million equivalent, an IFC A loan for US$22 million equivalent, and commercial financing for US$54 million equivalent from Standard Bank split between an IFC B loan tranche and a parallel loan. (ii) Outputs 10. In 2015/2016, the power plant reached a very high availability rate (98 percent), compared to the end-of-project target value of 85 percent contained in the PAD after the project restructuring. Nevertheless, the average dispatch level for 2015/2016 for the facility was 1.1 percent compared to 8.5 percent in 2014/2015, given the plant being disadvantaged by dispatch preference for cheaper geothermal and hydro energy, during the last three years. Moreover, Gulf Power Limited (GPL) ranked even lower in the economic merit order than other thermal plants with combine cycle units. Subcomponent 4: Olkaria III Geothermal Plant 2 Expansion Project - Developed by OrPower 4 (Total cost US$212 million, IDA PRG US$26 million and subsequently increased to US$31 million for the option to add Plant 3) (i) Original Scope 11. This subcomponent involved combining a 36 MW expansion with an existing 48 MW baseload geothermal power plant (‘Plant 1’) at the Olkaria geothermal fields, increasing it to a total installed capacity of 84 MW. This involved the additional development of the geothermal field and modification of the existing plant and integration of the new plant. The subproject is located within Hell’s Gate National Park in the Kenyan Rift Valley, 90 km northwest of Nairobi. A further expansion of a 16 MW plant (‘Plant 3’) was envisaged at the same site, for which OrPower 4 had an option that was exercised later. With the addition of Plant 3, the total aggregate geothermal capacity at Olkaria III reached 100 MW. (ii) Outputs 12. The expansion of Olkaria III power plant was the first to be realized by independent power producers and developed under the existing PPA between KPLC and OrPower4 mainly to accommodate the phased increased in capacity up to 100MW. In this context, the IDA PRG for Subcomponent 4 was required as part of he pre-existing payment security obligation, included in that PPA. However, in August 2015, the GoK and KPLC notified that OrPower4 decided to remove the payment security requirements under the PPA. This obviated the need for the IDA Guarantee on this Subcomponent. 31 Annex 3. Economic and Financial Analysis 1. At appraisal, the four proposed IPPs were economically evaluated based on the following methodological approach: (a) the need for additional generating capacity was analyzed based on a forecast of electricity supply and demand; (b) the evaluation examined if the four IPPs provided the least-cost solution to generation expansion; and (c) the evaluation conducted a cost/benefit analysis to determine the economic viability of the investments in the incremental generation capacity, measured by the EIRR and the NPV of the project. 2. The economic analysis for the OrPower4 geothermal IPP was conducted separately from the three thermal IPPs. The base case analysis for the three thermal IPPs resulted in an EIRR of 25 percent and an NPV of US$288 million and of 18 percent and US$79 million for the geothermal plant expansion. 3. The purpose of this current analysis is to value the economic benefits of the electricity generated through the Kenya Private Sector Power Generation Support Project and compare that with the original EIRR estimations in the PAD. The methodology compares the differences between the business as usual scenario and the scenario in which an intervention in the form of the project is made. 4. The following elements are considered in the analysis for the calculation of the economic benefits: (a) The economic value of the output of the thermal IPPs in this present analysis was computed based on the avoided cost of EPPs selling emergency power to KPLC. This cost comprised a fixed monthly capacity payment estimated at US$20 per kW plus the economic cost of the diesel used for generation. The economic cost of the diesel has been adjusted based on the crude oil price forecast prepared by the World Bank Group. (b) Following the methodology in the PAD, the economic value of the output of the geothermal IPP was estimated as the long-run marginal costs of generation of the system. For this analysis, the long-run marginal cost of generation from 2017 onwards is assumed to be US$0.09 per kWh, based on Kenya’s LCPDP published in October 2016. 5. On the cost side, the following elements have been considered in the analysis: (a) Investment costs. Total investment costs for each of the IPPs have been taken from the PAD but the investment profile has been adjusted to reflect when the plants were actually commissioned. (b) Fuel cost. This measure represents the HFO consumed in generating power by the three thermal IPPs. Where historical information was available, it was used; for the projections, the HFO has been correlated to the average annual crude oil price forecasted by the World Bank Group. 32 (c) Fixed and variable O&M expenses. These costs remain unchanged from the amounts provided by the project sponsors in the original analysis at appraisal. (d) Capacity factor. For this analysis, the load factor used in the restructuring and actual/latest scenarios were 7 percent and 12 percent, respectively, reflecting the average peak load factor for the thermal IPPs at the time of restructuring, and the minimum dispatch quota 12 percent for the thermal IPPs starting in 2016. For the geothermal IPP, an 88 percent capacity factor has been assumed; this value is in line with the estimations made at project appraisal and with the actual output of the facilities after commissioning. 6. The results of the economic analysis show that the overall project is economically feasible with an NPV of US$66 million (at 12 percent discount rate) and an EIRR of 14 percent. The results of the revised economic analysis are more conservative compared to those obtained at the time of project approval (mainly due to the lower dispatch profile for the thermal plants and by the decrease in the long run marginal cost in the system) but still above the recommended threshold of 12 percent for the EIRR. Table 3.1. Summary of the Economic Analysis Results Revised Original (Base Case) NPV in US$, millions NPV in US$, Case (at 12% Discount EIRR (%) millions (at 12% EIRR (%) Rate) Discount Rate) Thermal IPPs 85 16 288 25 OrPower 4 geothermal IPP −19 10 79 18 Combined 66 14 n.a. n.a. 33 Table 3.2. Economic Analysis of the Thermal IPPs (in US$ mn) Total Energy Fiscal Year- Fxd Var Fuel Value Period Capex T-Cst Produced TotalBnft NtBfnt End O&M O&M Cost US$/kWh (GWH) 2010 2009/10 2011 2010/11 2012 2011/12 - - - 2013 2012/13 146 - - - 146 - 0.49 - (146) 2014 2013/14 265 2.0 3.7 77 348 454 0.34 155 (193) 2015 2014/15 - 6.0 2.4 51 60 298 0.34 102 43 2016 2015/16 - 6.0 1.3 21 28 160 0.49 78 50 2017 2016/17 - 6.0 2.2 29 38 263 0.39 102 64 2018 2017/18 - 6.0 2.2 31 39 263 0.40 106 66 2019 2018/19 - 6.0 2.2 32 40 263 0.41 107 67 2020 2019/20 - 6.0 2.2 32 41 263 0.41 108 68 2021 2020/21 - 6.0 2.2 33 41 263 0.42 109 68 2022 2021/22 - 6.0 2.2 33 42 263 0.42 111 69 2023 2022/23 - 6.0 2.2 34 42 263 0.43 112 70 2024 2023/24 - 6.0 2.2 35 43 263 0.43 113 71 2025 2024/25 - 6.0 2.2 35 43 263 0.44 115 71 2026 2025/26 - 6.0 2.2 36 44 263 0.44 116 72 2027 2026/27 - 6.0 2.2 37 45 263 0.45 118 73 2028 2027/28 - 6.0 2.2 37 45 263 0.45 119 74 2029 2028/29 - 6.0 2.2 38 46 263 0.46 121 75 2030 2029/30 - 6.0 2.2 39 47 263 0.46 122 75 2031 2030/31 - 6.0 2.2 39 47 263 0.46 122 75 2032 2031/32 - 6.0 2.2 39 47 263 0.46 122 75 2033 2032/33 - 6.0 2.2 39 47 263 0.46 122 75 2034 2033/34 - 4.0 1.4 25 31 172 0.46 80 49 2035 2034/35 - - - - - - - - Discount rate 12% 3 85 16% Scenario NPV EIRR Table 3.3. Economic Analysis of the Geothermal IPP (in US$ mn) Fiscal Year- Energy Produced Value Period Capex O&M T-Cst TotalBnft NtBfnt End (GWH) US$/kWh ohr 2009/10 2011 2010/11 2012 2011/12 211.5 211.5 (212) 2013 2012/13 3.4 3.4 69 0.160 11.1 8 2014 2013/14 3.4 3.4 278 0.160 44.4 41 2015 2014/15 3.4 3.4 278 0.160 44.4 41 2016 2015/16 3.4 3.4 278 0.160 44.4 41 2017 2016/17 3.4 3.4 278 0.090 25.0 22 2018 2017/18 3.4 3.4 278 0.090 25.0 22 2019 2018/19 3.4 3.4 278 0.090 25.0 22 2020 2019/20 3.4 3.4 278 0.090 25.0 22 2021 2020/21 3.4 3.4 278 0.090 25.0 22 2022 2021/22 3.4 3.4 278 0.090 25.0 22 2023 2022/23 3.4 3.4 278 0.090 25.0 22 2024 2023/24 3.4 3.4 278 0.090 25.0 22 2025 2024/25 3.4 3.4 278 0.090 25.0 22 2026 2025/26 3.4 3.4 278 0.090 25.0 22 2027 2026/27 3.4 3.4 278 0.090 25.0 22 2028 2027/28 3.4 3.4 278 0.090 25.0 22 2029 2028/29 3.4 3.4 278 0.090 25.0 22 2030 2029/30 3.4 3.4 278 0.090 25.0 22 2031 2030/31 3.4 3.4 278 0.090 25.0 22 2032 2031/32 3.4 3.4 278 0.090 25.0 22 Discount rate 12% 3 (19) 10% Scenario NPV EIRR 7. The financial performance of KPLC from 2011 to 2016 (Table 3.4) shows that the utility’s operations, capital adequacy, and liquidity have been sustained. KPLC has remained profitable 34 despite taking on increased debt to finance power system expansion. Between 2012 and 2015, KPLC’s financing costs tripled and eroded the public utility’s ability to continue investing in improving the quality and coverage of its services. During this period, the debt service coverage ratio (DSCR) fell below 1 to 0.25 in 2013 and 0.33 in 2015. The 1.77 DSCR recorded in 2014 was due to a 31 percent year-on-year increase in electricity sales and a 35 percent decrease in investing activities. 8. However, in 2016, as part of the KEMP (P120014), KPLC refinanced its expensive short- term debt with proceeds from a US$500 million low-cost commercial financing that was supported by a US$200 million IDA Guarantee. This debt restructuring process successfully allowed KPLC to (a) reduce the weighted average cost of debt; (b) increase the weighted average maturity of its debt; and (c) comply with the World Bank’s financial covenants, particularly its DSCR, which is expected to remain above 1.3 in the short and medium term. Table 3.4. KPLC Financial Ratios Financial Ratios 2011 2012 2013 2014 2015 2016 Operating Indicators Gross Profit Margin (%) 31.9 26.9 30.1 31.1 31.5 35.2 Operating Profit Margin 14.9 12.9 16.3 20.5 22.3 22.9 (%) Net Profit Margin (%) 5.8 4.8 5.1 6.1 7.0 5.5 Return on Total Assets 5.2 6.3 3.6 4.6 4.4 3.5 (%) Return on Equity (%) 10.7 10.6 9.5 11.8 12.1 9.1 Capital Adequacy Indicators DSCR 2.46 2.68 0.25 1.77 0.33 1.45 Debt to Equity 0.62 0.64 1.08 1.28 1.82 1.49 Debt to Assets 0.61 0.58 0.66 0.67 0.70 0.67 Liquidity Ratios Current Ratio 1.25 0.90 0.97 1.03 1.64 1.12 35 Annex 4. Bank Lending and Implementation Support/Supervision Processes % (a) Task Team Members Names Title Unit Responsibility/Specialty Lending Supervision/ICR Karan Capoor Senior Financial Specialist AFTEG Team Leader Ritin Singh Senior Operations Officer AFTEG Co-Team Leader Lead Financial Farida Mazhar FEUFS PRG Specialist Officer/Consultant Kyran O’Sullivan Senior Energy Specialist AFTEG Energy Economist Atsuko Okubo Senior Counsel LEGCF Counsel Harvey Van Veldhuizen Lead Environmental Specialist OPCQC Environmental Specialist Teuta Kacaniku Energy Financial Analyst FEUFS Energy Finance Specialist Monica Teresa Restrepo Senior Counsel LEGCF Counsel Daniel Murphy Senior Operations Officer AFTEG Operations Officer Mitsunori Motohashi Energy Specialist AFTEG Financial Specialist Vongy Rakotondramanana Energy Specialist AFTEG Power Engineer Senior Social Development Gibwa Kajubi AFTCS Social Development Specialist Specialist Noreen Beg Senior Environmental Specialist AFTEN Environmental Specialist Bassem Abou-Nehme Energy Financial Analyst FEUFS Energy Finance Specialist Alexandra Planas Energy Specialist/Consultant AFTEG Energy Specialist Procurement Jean Jacques Raoul AFTEG Procurement Specialist/Consultant Shirmila Ramasamy Counsel LEGFI Counsel Lily Wong Chun Sen Program Assistant AFTEG Program Assistant (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle US$, Thousands (including travel No. of Staff Weeks and consultant costs) Lending FY12 115.45 616,231.84 FY13 72.20 370,810.00 FY14 8.85 45,869.30 FY15 7.54 39,702.97 FY16 0.03 138.49 Total: 204.07 1,072,752.60 Supervision/ICR FY12 3.83 18,083.49 FY13 24.34 163,367.24 FY14 47.40 243,865.49 FY15 37.67 168,453.00 FY16 19.66 99,479.11 FY17 29.50 140,758.41 Total: 162.40 834,006.74 36 Annex 5. Beneficiary Survey Results (if any) Not applicable 37 Annex 6. List of Supporting Documents 1. Energy Regulatory Commission. Letter on Heavy Fuel Oil Stock and Dispatch of Medium Speed Diesel Power Plants, September 29, 2016. 2. Gulf Power Limited. 2017. Project Completion Report. 3. Kenya Power and Lighting Company. Annual Reports, 2012–2016. 4. Ministry of Energy and Petroleum. Information on Aggreko, Thika Power, and Iberaafrica Statement of Fuel Costs for 2013–14. 5. Energy Regulatory Commission. Letter on Heavy Fuel Oil Stock and Dispatch of Medium Speed Diesel Power Plants, September 29, 2016. 6. Thika Power Limited. Annual Environmental and Social Monitoring Report (AMR), 2013– 2015. 7. Thika Power Limited. Semi-annual Review of Operations, December 2013, June 2014, December 2014, June 2015, December 2015, June 2016. 8. Triumph Power. 2017. Project Completion Report. 9. World Bank. Management Letter, Aide Memoires, Back to Office Reports, and Implementation Status and Results Reports filed in WBDocs. 10. World Bank. 2012. Private Sector Power Generation Project - Project Appraisal Document. 11. World Bank. 2015. Private Sector Power Generation Project - Restructuring Paper. 12. World Bank. 2017. Commodity Price Forecast. January 2017. 38 Annex 7. Financial Structure of the Project 1. The overall cost of the project was around US$623 million equivalent with equity of US$149 million equivalent to be funded by private project sponsors from their own resources. The sponsors of the thermal IPPs are either local investors or originate from other emerging markets, thereby promoting South-South cooperation. OrPower is expected to fund its equity partly from Ormat group resources and partly from internally generated cash flow. The total debt financing amounted to the equivalent of US$474 million, which was mobilized through IFC A, B, and C loans amounting to US$90 million, and commercial loans of US$208 million (including the IFC B loan of US$27 million equivalent) from commercial lenders such as ABSA and Standard Bank (both of South Africa) with the balance of US$202 million from DFIs such as OPIC and AfDB (see table 7.1). The details of the lending structure for each of the four subprojects being developed by the IPPs are described in the following paragraphs. 2. Thika Power (total cost US$146 million equivalent, IDA PRG US$35 million, and EUR 7.7 million IDA allocation of US$11.25 equivalent). This subproject total cost was around US$146 million equivalent and financing has been structured on a limited recourse basis with a debt equity ratio of 75:25, with US$110 million equivalent in debt and the balance of US$36 million equivalent in equity. Senior debt will be mobilized in equal amounts of US$36 million equivalent through A loans from IFC and AfDB and a commercial tranche from ABSA Capital of South Africa. 3. Triumph Power (total cost US$157 million, IDA PRG US$45 million, IDA allocation of US$11.25 million). The total project cost was around US$157 million, structured on a limited recourse basis with a debt equity ratio of 75:25 amounting to US$118 million in debt and US$39 million in equity. Standard Bank of South Africa underwrote the entire debt financing of the subproject. 4. Gulf Power (total cost US$108 million equivalent, IDA PRG US$35 million, and EUR 7 million IDA allocation of US$11.25 million equivalent). The total subproject cost is equivalent to about US$108 million. The financing which will be structured on a limited recourse basis consisting of 25 percent equity (amounting to US$27 million equivalent) and the balance of the debt financing consisting of 5 percent in subordinated debt through an IFC C loan amounting to US$5 million equivalent, an IFC A loan for US$22 million equivalent, and commercial financing for US$54 million equivalent from Standard Bank split between an IFC B loan tranche and a parallel loan. 5. OrPower 4 (total cost US$212 million). The total cost of the expansion of Plant 2 will be around US$212 million, which will be funded by US$31 million of new equity injection through a US$165 million load from OPIC and the balance through internal cash flow. 39 Table 7.1. Project Financing Structure Thika Triumph Gulf Power OrPower 4 Total Power Power Financing Source - Amount (US$, millions equivalent) Equity 36 39 27 47 149 Debt 110 118 81 165 474 - IFC A 36 22 58 - IFC B (commercial loan) 27 27 - IFC C 5 5 - AfDB 37 37 - Commercial Bank 37 118 27 182 - OPIC 165 165 a Total 146 157 108 212 Note: a. The cost of the subproject Olkaria 4 refers exclusively to the implementation of the 36 MW Plant 2, since at the time of the PAD the decision about the plant realization was only made for 36 MW and not for the additional 16 MW. 40 Annex 8. Contractual Framework for Guarantee Transaction 1. Figure 8.1 gives a pictorial representation of the contractual structure for the Kenya Private Sector Power Generation Project. The contractual structure of the subprojects is consistent with industry practice for limited recourse project finance transactions. A summary of key project Agreements is described below. Figure 8.1. Project Contractual Structure and World Bank Group Risk Mitigation Structure Thermal IPPs 2. The PPAs provide for the IPPs to supply power to KPLC on the basis of ‘take or pay’ obligations based on 85 percent available plant capacity. If the IPP fails to deliver the power, then KPLC will not have any obligation to pay. If the IPP can deliver but KPLC does not dispatch, then KPLC would be obligated to pay capacity charges to the IPP. PPA payments consist of the following components: (a) a fixed capacity charge, (b) a euro or U.S. dollar consumer price index- linked escalating capacity charge depending on the currency of the PPA, (c) energy charge, and (d) fuel charge payable in U.S. dollars. 3. Government Letter of Support issued to each of the IPPs provides that the GoK undertakes to backstop KPLC’s payment obligations under the PPA in the case of a KPLC FM funding event if KPLC is unable to make the necessary payments under the PPA as a result of force majeure events affecting KPLC. The GoK could also undertake to compensate the IPP for any ‘political events’ affecting the IPPs. 41 4. FSA signed between each IPP and with fuel suppliers for the supply of fuel in accordance with fuel specifications stipulated in the PPA. The FSAs are for the term of the PPA. The fuel charges will be a direct pass-through, payable in U.S. dollars. 5. Land Lease Agreements. Each of the thermal IPPs entered into a Land Lease Agreement for the term of the PPA. The Land Lease Agreements are as follows: (a) Thika Power Plant - the land was acquired by KPLC from Agro Tropical and will be leased to TPL for 20 years for the duration of the PPA term; (b) Triumph Power Plant - TPGC executed two leases with the Export Processing Zones Authority in Kenya with respect to adjoining pieces of land for a term of 50 years; and (c) Gulf Power Plant - GPL entered into a lease agreement with KPLC, which owns the land on which the plant will be located, for use of the land over the term of the PPA. 6. EPC contract. The EPC contracts for the three thermal projects were a fixed-price turnkey contract with provisions for liquidated damages (L/Ds) for delays and for plant performance and incentives for early completion. 7. O&M contracts range from 5 to 10 years, with a five-year contract between Triumph Power Plant and XJ Engineering, a six-year contract between Thika and MAN Diesel and Turbo France SAS for the Thika Power Plant, and a ten-year contract between Gulf and Wartsila Eastern Africa Limited for the Gulf Power Plant. The scope of the contracts includes routine O&M of the plants, as well as services during scheduled outages and major overhauls, including the provision of spare parts. The performance parameters are stipulated in the contract in terms of availability, capacity, and efficiency and also L/Ds in case of noncompliance. 8. Direct Agreements signed between the lenders and KPLC and the GoK include customary clauses for such an agreement, including acknowledgement of security interests and enforcement rights under the Project Agreements for the benefit of the lenders. Olkaria IPP 9. Supply contract with Ormat Systems Limited (OSL), a group company. Under this contract, OSL is responsible for the conceptual design of the power plant and for the supply of materials and equipment from outside of Kenya. It also guarantees performance and warranty material and equipment. 10. Pricing, coordination, and security contract with Orda 6, a group company. Under this contract, Orda guarantees the plant costs and covers cost overruns for the supply, transportation, and construction works. 11. Amended and restated PPA. KPLC and OrPower entered into the original PPA in November 1998 with subsequent Supplemental Agreements dated July 2000 and the Second Supplemental Agreement dated April 2003 relating to the existing facility and the amended and restated PPA dated January 19, 2007, which amended all the prior agreements relating to the existing facility. In March 2011, KPLC and OrPower entered into an amended and restated PPA with the purpose of extending the agreement to cover the expansion facility as well as the option for Plant 3 which OrPower exercised. Under this agreement, OrPower was required to undertake the geothermal reservoir development; design, procure, construct, finance, test, commission the transmission interconnector; design, procure, construct, finance, test, commission, operate, and 42 maintain the generation facilities; and sell the net electrical output to KPLC. KPLC, on its part, undertook to purchase and pay for available plant capacity and net electrical output of the generation facilities in U.S. dollars. Unlike in the case of the thermal IPPs, KPLC has the obligation under the PPA to make capacity payments to the IPP in the event of both natural and political force majeure events regardless of availability of the plant (with some adjustments for capacity payments for the unavailable part of the plant in certain force majeure events). However, the cure period for termination for force majeure events is longer than for the thermal IPPs of 270 days following a 90-day consultation period. 12. Amended and Restated Security Agreement. Under an Amended and Restated Security Agreement dated March 2011, KPLC agreed to provide security in the form of Letters of Credit issued for a year (renewable) for payments due from KPLC for the existing facility and the expansion with a view for these to be replaced by an IDA Guaranteed L/C once IDA’s Board approval is obtained and the relevant L/Cs become due for renewal. 13. Government Letter of Support. The GoK issued its Letter of Support to Olkaria to compensate it for any shortfall in payments by KPLC in the event of a ‘political event’. The GoK will backstop any shortfall of KPLC’s payment obligation only in the event of a ‘political event’. However, unlike in the case of the thermal IPPs, the GoK has no obligation relating to a KPLC FM Funding Event for Olkaria III. 14. Nonetheless, on August 3, 2015, the GoK and KPLC informed the World Bank that OrPower4 Inc. had formally notified KPLC that it will not take the payment security provided by IDA Guarantee in consideration of other commercial arrangements negotiated with KPLC, the major one being removal of the payment security requirement under the PPA. 43 Annex 9. Summary of the Borrower’s ICR and/or Comments on Draft ICR Summary of Gulf Power Limited ICR Report 1. This report is prepared by Gulf Power Limited (GPL), a Special Purpose Vehicle (SPV) Incorporated in Kenya to implement the Athi River II Power Plant (ARPP2) Project in Athi River, Kenya. The report is prepared for the benefit of the World Bank with GPL being a beneficiary of a World Bank Guarantee. The aim of this report is describing GPL’s experience with respect to the project. 2. ARPP2 is an 80.32 MW HFO Fired Medium Speed Diesel power plant that was developed following a tender launched by the Kenya Power and Lighting Company (KPLC) in April 2009. 3. ARPP2 is one of three MSD Power Plants tendered for in the above mentioned tender. GPL opted to take on Wärtsilä Finland Oy as the Engineering, Procurement and Construction Contractor (EPC) owing to their vast experience in MSD both Globally and Nationally having supplied equipment to four other MSD plants in the country. Wartsila Eastern Africa, a member of the world wide Wärtsilä network was selected as the Operations and Maintenance contractor (O&M) owing to their equally vast experience in the region since 1997. 4. This report is aimed at covering the following: (a) Assessment of the Project’s objectives, design, implementation and operational experience; (b) Assessment of the outcome of the project against the agreed objectives; (c) Evaluation of GPL’s own performance as well as that of the Bank and/or of the other partners during the preparation and implementation of the project with special emphasis on lessons learned that may be helpful in the future; (d) Description of the proposed arrangement for future operation of the project. 5. It is understood that the findings of this report are to be included in the overall Implementation Completion and Results (ICR) Report to be prepared by the World Bank regarding this project. Summary of Triumph Power ICR Report 6. Triumph Power Generating Company Limited (Triumph Power), an Independent Power Producer, has constructed and commissioned an 83 MW Diesel Power Plant in Athi River area in Kenya. The company will be selling electricity to Kenya Power and Lighting Company Limited for a period of twenty years as per agreed terms and conditions of Power Purchase Agreement, the PPA. 7. Triumph Power engaged XJ International Engineering Corporation of China as Engineering, Procurement and Construction contractor. 44 8. XJ International Engineering Corporation is also currently responsible for the operations and maintenance of the facility under and O&M agreement. 9. Triumph Power is the owner of an 83 MW Heavy Fuel Oil (HFO) fired power plant, located at Athi River, Kenya. The plant consists of ten medium speed engine sets with electrical and mechanical auxiliaries. Waste heat from the engine sets is recovered by ten heat recovery steam generators (HRSG), with steam centrally collected and converted to electricity with a steam turbine generator plant. 10. Triumph Power entered into a Power Purchase Agreement (PPA) with the Kenya Power & Lighting Company Limited (KPLC) on 14 June 2012 for the sale of electricity generated by the plant. 11. The plant was built by XJ International Engineering Corporation (XJIEC) under an Engineering, Procurement and Construction (EPC) contract with Triumph Power, initially entered into on 9 December 2011. 12. Triumph Power further entered into an Operations and Maintenance (O&M) agreement with XJ (also referred to as the operator) on 14 November 2012, for an operating period of ten years 13. Since mid 2015 all the thermal power plants are experiencing very low dispatch levels as Energy Regulator is tasked with lowering the cost of electricity as alternative sources of power supply have evolved. 14. The project construction completion is delayed by nearly one year, which had impacted adversely the revenue generation and debt service capabilities of the project. 15. Interim commercial operation (ICO) was achieved on 6 August 2015 and, Full commercial operation was achieved on 3 Feb 2016. 16. Currently, the operator (XJIEC) is able to satisfy the dispatch demand regime; however, the confidence in operator’s ability to perform with ten engines with Steam Turbine Generating Unit in full load operation is a concern TPG has. 17. With the foregoing concern, TPG had requested the lenders’ approval to replace the current Operator. The Lenders have taken a lot of time to grant the consent; especially the majority lender ICBC. Triumph power is seriously concerned about the performance of the operator, equipment failures and general premature deterioration of the power plant in the interim period. 18. On 2nd December 2016 TPG had signed the O&M agreement with MAN Diesel and Turbo Ltd., as the replacement contractor to provide O&M services. The new O&M contractor will be at site on 5th March 2017. Summary of Thika Power Limited ICR Report 19. Thika Power Limited did not submit an ICR Report to the Bank. 45 Comments of the Ministry of Energy and Petroleum to the Draft ICR 46 47 Annex 10. M&E Utilization - Summary of Key Additional Issues 1. Equipment deterioration due to low dispatch. The use of the thermal plants as peaking plants in the system revealed a regulatory gap in the PPA. In the current PPA, there is a lack of clarity on the technical constraints for the dispatch of generation units. This situation, only evident at very low dispatch levels, can lead to an early deterioration of the generation units as they operate in suboptimal conditions. 2. Decreasing HFO prices. Price formulas in the PPA allow a fuel cost pass-through on FIFO basis. This approach is reasonable when there is an expectation of increasing fuel prices but when the trend reverses (as it unexpectedly happened in 2014), the companies become uncompetitive because they are forced to offer electricity using the historical cost of the fuel they have in stock. 3. Change in law. Two changes in taxation law generated small financial claims from the IPPs: (a) the Finance Act 2016 reinstated a withholding value added tax regime, and (b) a Railway Development Levy was applied on all imported materials. These changes, not foreseen at the time of PPA signature, triggered the change in the law clause in the agreement, even though the financial impact of the change is considered marginal and poses low to negligible financial risk to the IPPs. 4. As part of continuous dialogue between the World Bank’s team and the stakeholders in the project, in September 2016, the MoEP, ERC, KPLC, and the IPPs reached an amicable agreement on arrangements to ensure that all thermal plants in the system are given a minimum dispatch—at around 12 percent—to avoid deterioration of the plants while minimizing the level of fuel costs passed through to consumers on a monthly basis by changing the HFO pricing from FIFO to LIFO. The parties also agreed to support the IPPs in their claims to the National Treasury to solve the change in the taxation law. 48 Annex 11. Project Map 49