Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00005264 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-50110 and IDA-53070) ON A CREDIT IN THE AMOUNT OF SDR 65 MILLION (US$ 102 MILLION EQUIVALENT) TO THE Government of Mozambique FOR THE National Water Resources Development Project May 20, 2021 Water Global Practice Africa Eastern and Southern Region CURRENCY EQUIVALENTS (Exchange Rate Effective June 30, 2020) Currency Unit = New Meticais MZN 69.41 = US$1 US$1.37 = SDR 1 FISCAL YEAR July 1 - June 30 Regional Vice President: Hafez M. H. Ghanem Country Director: Idah Z. Pswarayi-Riddihough Regional Director: Mark Lundell Practice Manager: Catherine Signe Tovey Task Team Leader(s): Pierre Francois-Xavier Boulenger ICR Main Contributor: John Factora ABBREVIATIONS AND ACRONYMS AF Additional Financing AfDB African Development Bank ARA Administração Regional de Água (Regional Water Administration) ARA-Sul Regional Water Administration -- South COVID-19 Corona Virus Disease 2019 CPS Country Partnership Strategy CUT Conta Única de Tesouro (Single Treasury Account) CWRAS Country Water Resources Assistance Strategy DNA Direcção Nacional de Águas (National Directorate of Water) Direcção Nacional de Abastecimento de Água e Saneamento (National Directorate DNAAS for Water Supply and Sanitation) Direcção Nacional de Gestão de Recursos Hídricos (National Directorate for Water DNGRH Resources Management) DRI Department of International Rivers DSS Decision Support System DUAT Direito de Uso e Aproveitamento de Terra (Land Use Right) EA Environmental Category EIRR Economic Internal Rate of Return EMP Environmental Management Plan EPP Emergency Preparedness Plan ERRP Emergency Resilient Recovery Project ESF Environmental and Social Framework ESIA Environmental and Social Impact Assessment ESS Environmental and Social Safeguards ESSP Environmental and Social Service Provider FSL Full Supply Level GMMA Greater Maputo Metropolitan Area GMWSEP Greater Maputo Water Supply Expansion Project GoM Government of Mozambique Ha Hectares ICR Implementation Completion and Results Report IEG Independent Evaluation Group INE Instituto Nacional de Estatísticas (National Institute of Statistics) ISR Implementation Status and Results Report Km Kilometers LiDAR Light Detection and Ranging LRP Livelihood Restoration Plan m Meters M&E Monitoring and Evaluation mASL Meters Above Sea Level Mm3 Million Cubic Meters Ministerio das Obras Públicas, Habitação e Ricursos Hídricos (Ministry of Public MOPHRH Works, Housing, and Water Resources) MS Moderately Satisfactory MT Mozambican Metical MU Moderately Unsatisfactory NPV Net Present Value NWRDP National Water Resources Development Project O&M Operations and Maintenance OP/BP Operational Policy / Bank Procedure ORAF Operational Risk Assessment Framework PAD Project Appraisal Document PAMT Project Administration and Management Team PAP Project Affected Persons PARP Poverty Reduction Action Plans PDO Project Development Objective PHRD Policy and Human Resources Development Programa Nacional de Desenvolvimento de Águas (National Water Development PNDA Program) PoE Panel of Experts PPA Project Preparation Advance PPCR Pilot Program for Climate Resilience RAP Resettlement Action Plan SADC Southern African Development Community SDR Special Drawing Rights SSA Sub-Saharan Africa TA Technical Assistance TSDA Technical Services and Dam Safety WRSS Water Resources Sector Strategy W2P Willingness to Pay WTP Water Treatment Plant Yr Year ZAMCOM Zambezi Watercourse Commission Table of Contents DATA SHEET .......................................................................................................................... 1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 6 A. CONTEXT AT APPRAISAL ......................................................................................................6 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION .............................................................9 II. OUTCOME .................................................................................................................... 12 A. RELEVANCE OF PDO........................................................................................................... 12 B. ACHIEVEMENT OF PDOs (EFFICACY) ................................................................................... 14 C. EFFICIENCY ........................................................................................................................ 23 D. JUSTIFICATION OF OVERALL OUTCOME RATING ................................................................. 25 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 25 A. KEY FACTORS DURING PREPARATION ................................................................................ 25 B. KEY FACTORS DURING IMPLEMENTATION.......................................................................... 27 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 32 A. QUALITY OF MONITORING AND EVALUATION (M&E) ......................................................... 32 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE.................................................. 33 C. BANK PERFORMANCE ........................................................................................................ 34 D. RISK TO DEVELOPMENT OUTCOME .................................................................................... 35 V. LESSONS AND RECOMMENDATIONS ............................................................................. 37 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 39 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION TEAM ............... 52 ANNEX 3. PROJECT COST BY COMPONENT ........................................................................... 54 ANNEX 4. EFFICIENCY ANALYSIS ........................................................................................... 55 ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS ... 59 ANNEX 6. EVOLUTION OF INTERMEDIATE RESULTS INDICATORS .......................................... 61 ANNEX 7. STATUS OF IMPLEMENTATION OF ENVIRONMENTAL AND SOCIAL SAFEGUARDS .. 63 ANNEX 8. GOM LETTER ON SAFEGUARDS BUDGET, HANDOVER CERTIFICATE ...................... 65 ANNEX 9. SUPPORTING DOCUMENTS ................................................................................. 69 The World Bank Water Resources Development (P107350) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P107350 Water Resources Development Country Financing Instrument Mozambique Investment Project Financing Original EA Category Revised EA Category Full Assessment (A) Full Assessment (A) Organizations Borrower Implementing Agency The National Directorate of Water (DNA, Direcção Government of Mozambique Nacional de Águas) Project Development Objective (PDO) Original PDO The Project Development Objective is to strengthen the management of national water resources and increase the yield from Corumana dam to augment water supply for the Greater Maputo Metropolitan Area. PDO as stated in the legal agreement The objective of the Project is to strengthen the development and management of national water resources and increase the yield of the Corumana Dam to augment water supply for the greater Maputo metropolitan area. Page 1 The World Bank Water Resources Development (P107350) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing 70,000,000 69,992,707 61,666,291 IDA-50110 32,000,000 26,966,213 25,248,435 IDA-53070 Total 102,000,000 96,958,920 86,914,726 Non-World Bank Financing 0 0 0 Borrower/Recipient 3,000,000 0 0 Total 3,000,000 0 0 Total Project Cost 105,000,000 96,958,920 86,914,726 KEY DATES Approval Effectiveness MTR Review Original Closing Actual Closing 15-Sep-2011 31-Jan-2012 15-Dec-2014 15-Dec-2017 30-Jun-2020 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 27-Sep-2013 5.38 Additional Financing Change in Results Framework Change in Components and Cost Reallocation between Disbursement Categories Change in Disbursements Arrangements Change in Legal Covenants 08-Dec-2017 40.09 Change in Results Framework Change in Components and Cost Change in Loan Closing Date(s) Change in Implementation Schedule 12-Nov-2019 68.39 Change in Results Framework Change in Components and Cost Change in Loan Closing Date(s) Cancellation of Financing Reallocation between Disbursement Categories Change in Implementation Schedule Page 2 The World Bank Water Resources Development (P107350) KEY RATINGS Outcome Bank Performance M&E Quality Moderately Satisfactory Moderately Satisfactory Modest RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 12-Jun-2012 Satisfactory Satisfactory 3.57 02 21-Feb-2013 Satisfactory Moderately Satisfactory 4.17 Moderately 03 10-Nov-2013 Moderately Unsatisfactory 6.12 Unsatisfactory Moderately 04 09-Apr-2014 Moderately Unsatisfactory 8.38 Unsatisfactory Moderately 05 05-Jul-2014 Moderately Unsatisfactory 13.77 Unsatisfactory Moderately 06 20-Feb-2015 Moderately Satisfactory 18.46 Unsatisfactory 07 09-Dec-2015 Moderately Satisfactory Moderately Satisfactory 28.90 08 21-Jun-2016 Moderately Satisfactory Moderately Satisfactory 31.66 09 27-Dec-2016 Satisfactory Moderately Satisfactory 33.12 10 22-Jun-2017 Satisfactory Moderately Satisfactory 35.74 11 20-Jan-2018 Satisfactory Satisfactory 40.84 12 03-Oct-2018 Satisfactory Satisfactory 50.80 13 04-Apr-2019 Moderately Satisfactory Moderately Satisfactory 55.53 14 02-Oct-2019 Moderately Satisfactory Moderately Satisfactory 63.03 15 31-Mar-2020 Moderately Satisfactory Moderately Satisfactory 76.01 Page 3 The World Bank Water Resources Development (P107350) SECTORS AND THEMES Sectors Major Sector/Sector (%) Public Administration 5 Other Public Administration 5 Social Protection 4 Social Protection 4 Transportation 5 Other Transportation 5 Water, Sanitation and Waste Management 86 Public Administration - Water, Sanitation and Waste 17 Management Other Water Supply, Sanitation and Waste 69 Management Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Social Development and Protection 4 Social Inclusion 4 Other Excluded Groups 4 Fragility, Conflict and Violence 0 Forced Displacement 0 Urban and Rural Development 33 Urban Development 33 Urban Infrastructure and Service Delivery 33 Page 4 The World Bank Water Resources Development (P107350) Environment and Natural Resource Management 154 Climate change 91 Mitigation 37 Adaptation 54 Water Resource Management 63 Water Institutions, Policies and Reform 63 ADM STAFF Role At Approval At ICR Regional Vice President: Makhtar Diop Hafez M. H. Ghanem Country Director: Laurence C. Clarke Idah Z. Pswarayi-Riddihough Director: Jamal Saghir Mark R. Lundell Practice Manager: Jonathan S. Kamkwalala Catherine Signe Tovey Pierre Francois-Xavier Task Team Leader(s): Marcus J. Wishart Boulenger ICR Contributing Author: John Factora Page 5 The World Bank Water Resources Development (P107350) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL A.1. Country Context 1. Between the 1992 peace agreement and the 2011 appraisal of the National Water Resources Development Project (NWRDP), Mozambique recorded impressive socio-economic achievements. The introduction of a multiparty system, deep economic reforms, and strong external support facilitated a successful transition to a market economy. As a result, Mozambique became one of the few examples in Sub-Saharan Africa (SSA) to achieve a combination of high growth and significant poverty reduction. Real GDP averaged eight percent from 1993 to 2011, making the country the fastest growing non-oil economy in SSA, and the national poverty headcount fell from 69 to 54 percent from 1996 to 2003. However, Mozambique started from a very low base and was one of the world's poorest countries. Furthermore, the 2008-09 household survey suggested that poverty reduction had plateaued at around 55 percent. 2. The Project Appraisal Document (PAD) describes water as a primary constraint to sustained economic growth and poverty reduction.1 Most of the sectors that contribute to the Mozambican economy are either directly dependent on secure, sustainable water availability or are indirectly affected by droughts or floods. While the country’s water resources compare reasonably well with other countries in similar climatic zones, the country is highly dependent on transboundary river basins. Nine important river basins are shared with other countries, including the Zambezi and the Limpopo, with Mozambique in the vulnerable downstream position in all but one. Furthermore, the country is one of the most exposed to risks from weather-related hazards, including floods, cyclones, and droughts. Low levels of infrastructure, including one of the world’s lowest water storage capacity, have hindered both the country’s ability to mitigate climate risks and to meet increasing water demand for consumption, irrigation, and other uses. 3. The projected shortage of bulk water supply was a key concern for the Greater Maputo Metropolitan Area (GMMA). At appraisal, the GMMA had a population of about 2.5 million people growing at a rate of 2.7 percent per year and contributing over 30 percent share of the national GDP. Water supply to the GMMA came almost exclusively from the Pequenos Libombos Dam on the Umbeluzi River, which had already been fully committed. The Master Plan for the Greater Maputo Water Supply System indicated that the quickest and least costly solution to increasing bulk water was by completing the Corumana Dam. The rock-filled dam, with a height of 45 meters (m), a crest length of 3,050 m, and a designed storage volume of 1,240 million cubic meters (Mm3), was situated 90 kms from Maputo. Built in 1989 for irrigation, hydropower, and flood control, the dam had been operating seven meters below the designed full supply level (FSL) due to a lack of financing to install the spillway gates. Installing the gates would increase the maximum FSL, increasing the reservoir’s storage capacity to serve as the next source of water to the GMMA. A.2. Theory of Change 4. The Government of Mozambique (GoM) viewed the NWRDP as part of the first phase of a broader multi- donor program to carry out the country’s water sector policies, strategy, regulations , and investments. Under the leadership of the Ministry of Public Works, Housing, and Water Resources (MOPHRH) and its National Directorate of 1 Project Appraisal Document (PAD) for the National Water Resources Development Project (Report No. 63794-MZ), August 29, 2011. Page 6 The World Bank Water Resources Development (P107350) Water (DNA),2 the NWRDP sought specifically to build water resource management capacity at the regional and national levels, while developing a pipeline of key investments. The NWRDP’s flagship activity was the completion of the Corumana Dam to augment bulk water supply, while the downstream intake, water treatment, and conveyance system would be supported by the Greater Maputo Water Supply Expansion Project (GMWSEP).3 In addition, the Pilot Program for Climate Resilience (PPCR) was to co-finance the upgrade of the country’s hydro-climatic information management system. 4 Figure 1 depicts the theory of change (ToC), which illustrates the inferred links between the project’s interventions, outputs, and outcomes, along with underlying assumptions, as described in the PAD and the additional financing (AF) project paper.5 A.3. Project Development Objectives and Outcome Level Indicators 5. The project development objective (PDO) was to strengthen the development and management of national water resources and increase the yield of the Corumana Dam to augment water supply for the Greater Maputo Metropolitan Area. The PDO as stated in the PAD was the same as what was in the Financing Agreement. The PDO was to be monitored and measured through the following outcome-level indicators: (1) Percent of the river basins covered by water resources management plans; (2) Percent of hydro-climatic stations operational and reporting; (3) Incremental reservoir yield available from Corumana Dam for GMMA. A.4. Components Component A: Water Resources Management (Planned: US$27 million: US$15 million from IDA, US$10 million from PPCR, and US$2 million from GoM counterpart funding. Actual: US$16.15 million from IDA). 6. The objective of this component was to enhance capacity at the national and regional level to address the challenges of water resources management in Mozambique and to strengthen the framework for water resources development. The component was to provide support to: (a) increase the capacity for negotiations, monitoring and compliance with international waters instruments; (b) strengthen the capacity of the GoM’s regional water authorities to develop strategies and studies to ensure the sustainable and equitable development of water resources in the country’s river basins; and (c) develop and implement mechanisms to facilitate comprehensive water allocation, licensing, revenue and compliance with monitoring measures. The PPCR was also expected to co-finance the project to develop a hydro-climatic information management system and related technical assistance (TA). These activities were to be supported through the provision of: i) consultant services and TA; ii) goods and equipment, including hydro-climatic and water quality equipment, bulk meters, computers, vehicles and office equipment; iii) works to establish hydrometeorological stations; iv) carrying out training and capacity building activities to the sector; and, v) incremental operating costs to support the DNA and the regional water authorities (ARAs). It also included support for the Panel of Experts (PoE) for dam safety and environmental and social aspects. 2 In July 2015, the institutional framework for MOPHRH was revised and DNA was divided into two directorates, the National Directorate for Water Resources Management (DNGRH) and the National Directorate for Water Supply and Sanitation (DNAAS). The ICR uses DNA and DNGRH interchangeably. 3 The Greater Maputo Water Supply Expansion Project (P125120) was approved on July 25, 2013, for US$178 million. 4 The Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049) was approved in April 25, 2013, for US$15 million. 5 National Water Resources Development Flood Response Project Additional Financing (P146098, Report No. PAD702), September 2014. Page 7 The World Bank Water Resources Development (P107350) Figure 1. Theory of Change Page 8 The World Bank Water Resources Development (P107350) Component B: Completion of the Corumana Dam: Civil and Hydro-Mechanical Works (Planned: US$42 million from IDA. Actual: US$39.21 million from IDA). 7. The objective of this component was to support completion of the Corumana Dam to increase bulk water for the GMMA. Specific outputs included the: (a) installation of spillway gates with repair works on concrete pillars/abutments; (b) construction of a saddle dam with an emergency spillway; (c) development of a monitoring system; (d) provision of safety equipment; and (e) adoption of dam safety measures. These outputs were to be achieved through the provision of: i) consulting services for detailed design and supervision and technical assistance (TA) within the ARA-Sul to assist with implementation; ii) works required for completion of the Corumana Dam; and iii) goods to support implementation. Component C: Completion of the Corumana Dam: Environmental and Social (Total: US$6.5 million: US$5.5 million from IDA and US$1.0 million from GoM. Actual: US$4.02 million from IDA). 8. The objective of this component was to ensure that completion of the Corumana Dam was compliant with environmental and social safeguards (ESS). The specific project safeguard instruments included: (a) the Environmental Management Plan; and (b) the Resettlement Action Plan (RAP), including a Community Livelihood Restoration Plan (LRP). These were to be implemented through the provision of: i) consulting services and TA; ii) works, including a water supply and sanitation program and community development measures in the vicinity of the dam; and, iii) goods, equipment and incremental operating costs to be provided to assist with monitoring, stakeholder consultations and implementation. The Government was to directly finance all cash compensation and land acquisition as part of the RAP. Component D: Water Resources Development (Total: US$35.75 million: US$35 million from IDA and US$750,000 from PHRD. Actual: US$28.28: US$27.53 from IDA and US$750,000 from PHRD). 9. The objective of this component was to help address Mozambique’s infrastructure deficit in the water sector. Specific outputs included: (a) feasibility studies, tender designs, strategic assessments, river basin plans, and the development of a pipeline of future water resource investments; and (b) environmental and social assessments for future potential water sector related investments. These outputs were to be delivered through the provision of: i) consulting services and TA for the detailed design of water resources infrastructure, water supply networks and distribution systems, along with the preparation of ESS instruments; and, ii) operating expenses associated with workshops, training and capacity enhancement initiatives. The Component was to be implemented by DNA with the specific studies and works coordinated through the relevant ARAs and beneficiary agencies. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION 10. The project was restructured three times. The first restructuring was approved on September 27, 2013, as part of an Additional Financing (AF) from IDA resources for SDR 21.2 million (US$32 million, equivalent), to finance the costs of emergency rehabilitation of civil works (e.g., dykes and levees) following the extreme flooding in southern Mozambique, along with longer-term interventions for improved flood management and mitigation.6 The restructuring included changes to the monitoring and evaluation (M&E) framework to reflect the introduction of new project activities. A second restructuring took place on December 8, 2017, primarily to extend the closing date by 23 months, 6 National Water Resources Development Flood Response Project Additional Financing (P146098, Report No. PAD702), September 2014. Page 9 The World Bank Water Resources Development (P107350) from December 15, 2017 to November 14, 2019, and to adjust an outcome-level target and several intermediate indicators.7 The third restructuring, approved on November 12, 2019, further extended the closing date by 7.5 months to June 30, 2020, and adjusted the target of another PDO indicator. All told, the closing date was extended twice, with a cumulative extension of 30.5 months, extending the implementation period from six to 8.5 years. B.1. Revised PDOs, Outcome Indicators and Targets 11. There were no changes to the PDO, and the three outcome-level indicators remained the same, though target values were adjusted for two of them. The target value for “river basins covered with management plans” decreased from 19 to 12 to better align with the project budget, while the target value for “incremental reservoir yield available from Corumana Dam for GMMA” was reduced from 16.8 million cubic meters per year (Mm3/year) to 11.6 Mm3/year, to reflect temporary restrictions on the dam’s FSL (see paragraphs 16-17). B.2. Revised Components 12. The 2013 AF for US$32 million increased the scope of the project. Specifically, Components A and D included the following new activities: • Component A - Water Resources Management (US$4 million). New activities included integrated flood management and mitigation studies for the Limpopo River basin in Mozambique to inform longer-term infrastructure investments and planning. • Component D - Water Resources Development (US$28 million). New activities included: (i) emergency civil works to rehabilitate and reconstruct prioritized dykes and levees; (ii) rehabilitation of the Macarretane Barrage on the Limpopo River, including hydro-mechanical works, gate control system, reinforcement of foundations, and the rehabilitation of office buildings; (iii) TA and supervision associated with the emergency civil works; and, (iv) other improvements to the existing infrastructure of dykes and levees. B.3. Other Changes 13. The restructurings included other changes to the M&E framework. The 2013 AF introduced seven new intermediate indicators linked to the emergency reconstruction in the Limpopo valley after the floods. The 2017 restructuring eliminated the intermediate indicator “Improved cost recovery of ARAs”, noting that the result “would require substantially more support, including legal and institutional reform, investments and capacity building for water licensing and monitoring systems that are beyond the project scope.”8 All told, the number of intermediate indicators rose from 11 to 17. Details of M&E changes can be found in Annex 6. 14. There were de facto changes to the project that were not recorded in project restructurings. The original plan for the PPCR to co-finance the NWRDP by US$10 million to upgrade the country’s hydro-climatic system was superseded by a standalone project, namely the US$15 million Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049, hereafter referred to as the “PPCR project”).9 The PPCR project financed the modernization 7 Restructuring Paper (Report No. RES29775) for the National Water Resources Development Project, December 8, 2017. 8 Restructuring Paper (Report No. RES29775) for the National Water Resources Development Project, December 8, 2017. 9 The Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049) closed on December 31, 2019. The project’s ICR assessed Page 10 The World Bank Water Resources Development (P107350) of hydro-climatic observation, monitoring, data transmission, modeling, and dissemination capacity, which were originally part of Component A of the NWRDP. This decision reflected the complexity of the works, activities, and institutional arrangements, all of which were thought to be more easily implementable through a standalone project. 15. There was a difference of about US$28 million in planned vs. actual costs. Part of the difference stems from the forgoing of PPCR financing and associated GoM counterpart funding for the hydro-meteorological work, as discussed above. Counterpart funding for RAP implementation also did not materialize; by the time the ESIA and RAP were updated and ready for implementation, the country was in the midst of a severe fiscal crisis. Furthermore, the November 2019 restructuring cancelled US$5 million, which was reallocated to another project to support the government’s response to Cyclones Idai and Kenneth. 10 Lastly, and as discussed in paragraph 69, exchange rate depreciation led to a significant reduction in IDA funding. Table 1 shows planned and actual costs by component. Table 1. Planned vs. Actual Costs IDA PPCR GOM PHRD Total Planned Component A 15.00 10.0 2.0 27.00 Component B 42.00 42.00 Component C 5.50 1.0 6.50 Component D 35.00 0.75 35.75 Contingencies 4.50 4.5 Total 102.0 10.0 3.0 0.75 115.75 Actual Component A 16.15 16.15 Component B 39.21 39.21 Component C 4.02 4.02 Component D 27.53 0.75 28.28 Total 86.91 0.0 0.0 0.75 87.66 B.4. Rationale for Changes and Their Implication on the Original Theory of Change 16. The change to the target value of the outcome indicator “incremental reservoir yield from Corumana Dam” reflected the changes to the technical design of the dam. The reservoir yield target contained in the PAD was based on the original design of the dam from the late-1970s, which along with the installation of the radial gates, included the construction of a saddle dam with an earth fuse plug to act as an emergency spillway. However, detailed designs and hydrological studies undertaken during implementation made it clear that the fuse-plug could no longer play the role of an emergency spillway as studies showed that the probable maximum flood estimation was higher than originally projected when the dam was built.11 Moreover, the downstream area of this saddle dam, free of people during project preparation, became inhabited by hundreds of people who would be in harm’s way if the fuse-plug were activated. Based on advice from the supervising contractor and the PoE, the GoM decided in February 2016 to do away with the fuse plug element and to close the saddle dam on the right side of the reservoir with saddle embankments.12 the project with a Moderately Satisfactory Outcome rating. 10 Restructuring Paper (Report No. RES38756) for the National Water Resources Development Project, November 12, 2019. 11 De Almeida, Luis (Project Coordinator, DNGRH). Phone interview by John Factora. September 29, 2020. 12 Restructuring Paper (Report No. RES38756) for the National Water Resources Development Project, November 14, 2019. Page 11 The World Bank Water Resources Development (P107350) 17. In addition, new Mozambican dam safety regulations stipulated that the probable maximum flood needed to be accommodated by the main spillway and, if needed, a permanent auxiliary spillway. This would ensure adequate flood routing without risk of failure. As such, and given the limited capacity of the existing spillway, an auxiliary spillway, including a permanent uncontrolled overflow structure, chute, and stilling basin in the left embankment of the dam was designed with a preliminary cost of around US$40 million. The World Bank and GoM agreed that the auxiliary spillway was beyond the scope and timeframe of the NWRDP and that other sources of financing would need to be found. Until the auxiliary spillway is constructed, the FSL would be restricted to 114 meters above sea level (mASL) (instead of 117 mASL, as original planned), which corresponds to a reservoir yield of 960 Mm3/year, and an incremental yield of 11.6 Mm3/year. Although this change impacted economic efficiency (see paragraph 50), the Theory of Change would not be impacted as the project would still increase the dam’s reservoir capacity and augment bulk water supply. 18. The implications of the other project changes on the theory of change were mixed. The AF bolstered the project’s objective of strengthening water resources management, improving water security, and improving resilience to climate change by rehabilitating hydraulic infrastructure destroyed in the 2013 floods and by improving longer-term flood management and mitigation. In contrast, the preparation of the PPCR project as a standalone operation meant the NWRDP had very little contribution to improving the country’s hydro-climatic monitoring system and its associated long- term goals of improved economic planning and climate change preparedness. Dropping the national water allocation and revenue management system also reduced the project’s scope to affect sustainable management of water resources. II. OUTCOME A. RELEVANCE OF PDO 19. The project built on a history of Bank engagement in the water sector in Mozambique since 1998 and its global experience with complex water resources development projects. The same ministry and directorate had been involved in a series of Bank-financed projects, including the National Water Development Project I and II (PNDA – Programa Nacional de Desenvolvimento de Águas).13 The project reflected lessons learned from these earlier interventions (see paragraph 53). The PDO was sufficiently challenging but was not overly ambitious, reflecting the GoM’s priorities and capacity, and avoiding supply-driven or overly complex, politically challenging activities that would have exceeded the GoM’s willingness or ability to implement. The project objectives gained even more urgency during the course of implementation, given concerns that Maputo was approaching a “day zero” scenario, similar to Cape Town, South Africa’s drought-induced water crisis in 2017-2018 (see Box 1). 20. The project was well aligned with the Bank’s regional and country strategies. The Bank strategy for Africa14 sought to close Africa's infrastructure gap to drive productive development of urban growth poles and to build resilience to climate change and natural disasters. In particular, the project was to contribute to the regional strategy's two pillars: 1) Competitiveness and Employment, and 2) Vulnerability and Resilience. The project anchored a growing water portfolio in the Bank’s Country Partnership Strategy (CPS), approved by the Board in 2007.15 In particular, the project supported Pillar III “Sustainable and Broad-based Growth” of the CPS, and was to contribute directly to Result Area #5 “Strengthened economic growth potential”, specifically Outcome 17 “Improved sustainable management of water resources” and 13Mozambique National Water Project 1 (P039015) approved February 1998 and National Water Project 2 (P052240), approved June 1999. 14Africa's Future and the World Bank's Support to It: Africa Regional Strategy (59761), The World Bank, February 2011. 15 Country Partnership Strategy for the Republic of Mozambique (Report No. 39395-MZ), April 24, 2007. Page 12 The World Bank Water Resources Development (P107350) Outcome 18 “Enhanced capacity to respond to disasters”. It also contributed to Pillar II “Equitable access to key services” through Result Area #3 “Improved government effectiveness in the provision of services” via Outcome 8 “Increased access to potable water” by providing upstream infrastructure for bulk water supply. The CPS, in turn, was strongly aligned with Mozambique’s Poverty Reduction Action Plans (PARP),16 the GoM’s strategy for growth and development, which acknowledged the need to improve the country’s water resources management to minimize the negative impacts of climate variability, ensure the sustainability of water supply services and the availability of water for agricultural production as an important factor in achieving development and poverty reduction goals. 21. The project remained relevant to the World Bank Group Country Partnership Strategy (CPS) for FY17-21.17 The CPS highlights the key ongoing challenges of developing and managing the country’s water resources; addressing the country’s infrastructure deficit; improving basic services such as access to safe water, sanitation, and hygiene; and reducing the country's vulnerability to weather-related shocks, all key elements of the NWRDP. The project directly supported Pillar III, “Enhancing Sustainability and Resilience” of the CPS which identifies adequate access to infrastructure resources and services as a key driver of growth and poverty alleviation. It also cut across the other two Focus Areas of “Promoting Diversified Growth and Enhanced Productivity” and “Investment in Human Capital”, given the role water resources play in productivity and human development. The PDO aligned with the CPS objectives of improving management of climate risks and natural resources, improving access to water and sanitation, promoting inclusive urbanization and decentralization. Lastly, the project addressed the emerging priority of responding to natural disasters, as discussed (in the context of the 2019 hurricanes) in the April 2020 Performance and Learning Review of the CPS.18 22. The project also remains consistent with the GoM’s national and sectoral priorities. The project supported Mozambique's 2015-2035 National Development Strategy19 and particularly its pillars on infrastructure development to support the sustainable management of water resources, and innovation and technological development to support the management of natural resources. The project contributes to the government’s Five-Year Program for 2020-202420 and specifically its priorities of inclusive and sustainable growth, social and economic stability, economic productivity and competitiveness, and climate change. In particular, the project directly supports the program’s strategic objectives of infrastructure development, which specifically references the rehabilitation and development of dams and other hydraulic infrastructure, the strengthening of natural resources management, and the need to reduce climate-related risks. The project remains consistent with various sectoral instruments, including the 2013-2025 National Strategy for Adaptation and Mitigation to Climate Change,21 which proposes priority actions to reduce climate risk and strengthen early warning systems; and (ii) the Master Plan for Water Resources Management in Mozambique,22 which analyzed national water resources and proposed the rehabilitation and development of hydraulic infrastructure, ground water resources, and drainage of inland rainwater. The latter specifically cites the Corumana Dam for priority rehabilitation. 16 República de Moçambique, Conselho de Ministros, Plano de Acção para Redução da Pobreza (PARP) 2011-2014, May 3, 2011. 17 Country Partnership Strategy for The Republic of Mozambique, FY17-21 (Report No. 104733-MZ), March 30, 2017. 18 Performance and Learning Review of the CPS for Mozambique, FY17-21 (Report No. 144024-MZ), April 3, 2020. 19 República de Moçambique, Estratégia Nacional de Desenvolvimento (2015-2035), Maputo, July 2014. Accessible through www.cabri- sbo.org/en/documents/national-development-strategy-2015-2035 (accessed on October 8, 2020). 20 República de Moçambique, Conselho de Ministros, Proposta do Programa Quinquenal do Governo: 2020-2024, Maputo, 2020. 21 República de Moçambique, Ministério para a Coordenação da Acção Ambiental, Estratégia Nacional de Adaptação e Mitigação de Mudança Climáticas, 2013-2025. 22 Republic of Mozambique, National Directorate of Water Resources Management, Development of Master Plan for Water Resources Management in Mozambique, February 2018. Page 13 The World Bank Water Resources Development (P107350) 23. The relevance of the PDO is rated "High." The Bank has been one of Mozambique’ most important development partners. This is reflected in the large country assistance program. The water sector portfolio alone amounts to more than $700 million. The project builds on a long history of support in water resources management and development and is well coordinated with development partner programs. The project objectives were sufficiently challenging, but not overly ambitious, and reflects the GoM’s priorities as elaborated in its poverty reduction strategies and plans and policies. Lastly, the project objectives were strongly aligned with the Bank CPFs, both at appraisal and completion, and addressed the priorities of water security and natural disaster response that gained heightened relevance during implementation. B. ACHIEVEMENT OF PDOs (EFFICACY) B1. Assessment of Achievement of Each Objective Objective 1. To strengthen the management of national water resources (Rating: Substantial) 24. The project substantially achieved the objective of strengthening national water resources management. As discussed in paragraph 14, formal restructurings and de facto changes reduced the scope of this objective. Chief among them was the decision for the PPCR not co-finance the project to modernize the country’s hydro-meteorological system. The work was instead carried out through a standalone project, which closed in December 2019 and was rated Moderately Satisfactory (MS).23 None of the outcomes stemming from the work on the hydro-meteorological system are attributed to the NWRDP. No formal restructuring was done to drop the associated outcome indicator “percent of hydro- climatic stations operating and reporting”. Moreover, the indicator does not provide a complete account of all the outcomes promised under this objective. As such, other observed outcomes were used to assess achievement. These included: (a) improved capacity to negotiate and comply with international waters instruments; (b) informed sustainable and equitable management of priority river basins through strategies and studies; (c) informed long-term infrastructure investment planning in the Limpopo River basin through an integrated flood management and mitigation study; (d) informed flood management preparedness through a decision support system (DSS); and (e) improved capacity through extensive capacity building. Achievement of these results on aggregate have strengthened water resources management. 25. On international waterways, the project supported the GoM on several fronts. They included collaboration with other riparian countries to review and update two agreements on international waters, namely the Interim IncoMaputo Agreement between Mozambique, Eswatini and South Africa, and the Púngue and Búzios Watercourses Agreement between Mozambique and Zimbabwe. Both agreements were signed by all riparians, which exceeded the target of reviewing one agreement and updating another.24 Given Mozambique’s position as the most downstream country in eight out of the nine international rivers it shares with other Southern African Development Community (SADC) countries, the aim of reviewing the agreements was not only to ensure transboundary inflows to meet demand, but to explore sustainable and equitable ways of developing and managing water for shared benefits, in keeping with SADC Regional Strategic Action Plan on Integrated Water Resources Development and Management. The review and negotiations were supported by capacity building activities for key members of the negotiating team from the Department of International 23 The Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049) closed on December 31, 2019. See the project’s ICR (Report ICR5203) and IEG’s review (ICRR0022203). 24 News articles cover the signing of these agreements. See for example: “Mozambique and Zimbabwe sign agreement to enhance wat er cooperation in the Busi Watercourse.” Smart Water Magazine, September 9, 2019. https://smartwatermagazine.com/news/global-water- partnership/mozambique-and-zimbabwe-sign-agreement-enhance-water-cooperation-buzi Page 14 The World Bank Water Resources Development (P107350) Rivers (DRI) of the Department of National Water Resources Management (DNGRH), including financing masters-level studies on international water law and shorter learning modules on water conflict management, mediation, and negotiation processes. News articles indicate implementation of the agreements at the local and national levels.25 26. The project financed the development of the National Strategic Action Plan to Monitor the Compliance of International Shared Watercourses Agreements.26 The action plan gives the GoM a much-needed road map to monitor, negotiate, and enforce international river agreements. The preparation entailed broad participation by DNGRH, DRI, the ARAS, the Ministry of Transport and Communications, and the Ministry of the Sea, Inland Waters and Fisheries. The implementation of the action plan is being supported by the state budget under the coordination of the DNGRH. One of the actions taken was strengthening Mozambique’s participation in joint water commissions with riparian countries, including the Zambezi Water Resources Commission (ZAMCOM), for which the project financed the Mozambican delegation’s membership on an interim basis. The GoM reported that the action plan has had the effect of improving internal coordination across government departments and with non-governmental stakeholders on transboundary watercourses, in addition to strengthening collaboration with riparian countries. As of this writing, negotiations are underway to establish the Joint Commission for the Incomáti and Maputo Rivers. 27. The project financed river basin plans to improve sustainable, equitable and participatory management of water resources in the Zambezi, Lúrio, and Limpopo river basins. These plans also sought to identify future multi- purpose development potential in the three river basins, an aspect that will be discussed in the next section on water resources development. These plans contained a comprehensive monograph of land use, water availability, current and planned water demand, socio-economic conditions, environmental and water quality, and legal and institutional framework. Moreover, the plans analyzed the mandate, core functions, technical and management capacity to meet institutions’ goals, mission, and functions of river basin authorities. Additionally, DNGRH prepared synopses of each basin plan; each was approved by the Council of Ministers as binding documents for spatial planning and development. 28. The plans were tailored to the specific needs of the corresponding ARAs and prepared in a participatory manner involving key users of the basin. In the case of the Zambezi and Limpopo, consultations with water resources management entities in the riparian countries were also undertaken. The Zambezi monograph and strategy is assisting ARA-Zambeze to position itself with respect to the basin wide strategic plan envisaged under the Zambezi Watercourse Commission (ZAMCOM) Agreement. The plan highlighted the benefits of cooperation among the riparian countries in the Zambezi River Basin through a multi-sectoral economic evaluation of water resources development, management options and scenarios from both national and basin-wide perspectives. The Lúrio River Basin plan paid particular attention to institutional development for local authorities responsible for developing water resources. 29. The plan for the Limpopo includes the integrated flood management plan which looks at structural and non- structural measures to help prevent and manage flooding in one of the most flood-prone areas in the country. The plan for the Limpopo was part of a broader effort after the 2013 floods to not only support emergency rehabilitation, but 25 See for example: “Water department to engage stakeholders on interim IncoMaputo agreement.” Engineering News. September 17, 2019. www.engineeringnews.co.za /article/water-dept-to-engage-stakeholders-on-interim-incomaputo-agreement-2019-09-17 26 The strategy includes an evaluation of the water resources management framework of the countries that share river basins with Mozambique; an evaluation of the existing agreements on shared watercourses; updated SADC protocols on shared watercourses; categorization of agreements according to scope and objectives; description of shared basins; legal and institutional aspects of join water resources management; verification and monitoring of compliance; dispute resolutions. Page 15 The World Bank Water Resources Development (P107350) to also “build back better” in the long term. Specifically, the study sets out structural and non-structural measures on flood management and outlines the role of state and non-state actors across several sectors at the local, provincial, and national levels. The set of actions include the strengthening and management of flood early warning system and the information and communication management systems for national emergencies, and introduction of drought-resistant crops. The development of the plan benefited from and made extensive use of light detection and ranging (LiDAR) data generated by aerial surveys conducted under another Bank project.27 The implementation of the plan has a 10-year horizon (2020-2030), and preparatory activities have already begun. The recommendations of the study provide local authorities with guidelines for planning and structural development processes in the basin. With this study, the NWRDP makes a valuable contribution to better flood prevention and management. 30. Related to the above, the project supported the development of an integrated decision support system (DSS) for flood management in the Limpopo. The DSS helps integrate multiple criteria decision making, a collection of methodologies, such as remote sensing, geographic information system, hydrologic models, and real-time flood information systems; compare, select, or rank multiple alternatives to elicit and model flood preferences of stakeholders; and improves the coordination across agencies, organizations and affected citizens. The DSS is managed by DNGRH, but as a web-based system, access has been shared with ARA-Sul and other partners for the full use of the system. Since it integrates different types of information from the different sectors directly involved in territorial planning, it also assists in infrastructure development in the river basin. The DSS is operating effectively and informing decision making. 31. The project entailed a comprehensive institutional strengthening program that cut across all project components. Formal training benefited 125 government officials from DNGRH and all ARAs. All told, twenty-six courses were designed and carried out; most were short-term and held in-country and covered both cross-cutting and technical subjects. The former included financial management, procurement, results-based monitoring and public relations. Specialized and technical courses included international law on transboundary water, regional regulatory frameworks for environmental planning and management, geophysical surveys, hydromechanics, water conflict management, mediation and negotiation processes, hydrologic modelling, water resources planning model, institutional and legal framework of the management of tailings dams, software for hydraulic engineering, hydromechanics (protection of hydromechanical facilities), and socio-environmental management of hydraulic infrastructure. The training had a tangible impact on technical and managerial capacity, with ARA staff demonstrating the ability to process and analyze data of, inter alia, hydrological reports, as opposed to just collecting and storing data. As a result, water quality and drought bulletins were systematically produced by ARA-Centro, ARA-Centro-Norte and ARA-Norte. ARA-Sul and ARA-Zambeze, having routinely issued these bulletins, began to calculate water quality index.28 The project improved the financial management systems of DNGRH and ARAs, through training in international good practices and in managing off-CUT (Single Treasury Account) funds. Lastly, the project financed post-graduate studies in international environmental law, international water law, water resources management, and human resources and labor law. Objective 2. To strengthen the development of national water resources (Rating: Modest/Substantial) 32. The objective was partially achieved before the December 2017 restructuring and substantially achieved afterward. The aim was to help address Mozambique’s infrastructure deficit in the water sector. The project prepared 27 Enhancing Spatial Data for Flood Risk Management Project (P149629), approved June 2014. 28 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. Page 16 The World Bank Water Resources Development (P107350) river basin plans to help sustainable and equitable development of water resources in three priority river basins. Though the project fell short of the number of basin plans targeted (see below), other observed outcomes had a more immediate and meaningful impact on beneficiaries and, on aggregate, constitute achievement of the objective. The results include detailed technical design of three dams, including their water supply networks and distribution systems, and the preparation of the associated ESS instruments that allowed the dams to go to tender; and the timely rehabilitation of critical hydraulic infrastructure destroyed during the 2013 Limpopo River Basin floods, including the Macarretane dam. 33. Per the results framework, the objective was to be assessed through the outcome indicator measuring the number of river basins plans. Of the 12 river basin plans targeted, five constituted the baseline, leaving seven to be delivered by the project. Due to budget constraints, the project only financed three, including those for the Limpopo, Zambezi, and Lúrio.29 However, the project leveraged financing from the GoM and other development partners for the remaining plans, notably for the Mutamba and Guiúa River basins (by the GoM) and the Maputo and Incomáti River basins (by the Netherlands). In addition, the African Development Bank (AfDB) financed river basin plans for the Buzí, Rovuma, and Save river basins. 30 The plans undertaken by other financing sources were completed outside the framework of the project but reflect the role the Bank and project played in “crowding-in” external support to achieve the project’s objectives.31 All plans were based off similar terms of references and followed the same template with similar content. Lastly, an additional 10 development plans (which among other topics, looked to improve rural infrastructure, including irrigation schemes) were prepared under the Bank-financed SUSTENTA program 32 for the following river basins: Monapo, Mecuburi, Meluli, Motomonho, Ligonha, Melela, Molócue, Nipiode, Raraga, Moniga.33 34. The three river basin plans prepared by the project contained river basin development scenarios and strategic investment plans. The former entailed three optional scenarios through 2030 from which the GoM could select the most appropriate option based on the assessment criterion: economically beneficial, socially just, and environmentally sound. It analyzed present and future water demand in key sectors, including water supply and sanitation, energy, industry and mining, agriculture and irrigation, fisheries, tourism, and ecology. It looked at alternative surface and groundwater supply sources, identified potential investments to augment water supply and flood protection, and potential dam sites. Alternative development schemes were ranked from a multi-purpose perspective and a simplified ESS screening and costing were carried out, along with economic and financial assessments. A strategic investment plan for water resources development identified future multi-purpose infrastructure in the three basins, based on existing project concepts, and outlined implementation plans, including necessary preparatory studies for the projects. Lastly, cost estimates were provided, including unit reference values for developed water and proposed funding mechanisms. The plans entail critical investments in the order of US$2 billion for the Limpopo River Basin, US$1.8 billion for the Zambezi River Basin, and US$570 million for the Lúrio River Basin.34 Implementation is long-term and expected to take 25 years (2020-2045). 35. The project also financed the detailed technical designs, including for water supply networks, distribution systems, and environmental and social impact assessments for the Gorongosa, Metuchira and Ressano Garcia dams. 29 The PAD originally envisioned 19 river basin development plans. Due to budget constraints, the target was reduced to 12 river basins. A further restructuring to reduce the target to three was not done. 30 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. 31 McMillan, Shelley (Senior Water Resources Management Specialists, World Bank). Phone interview by John Factora. September 3, 2020. 32 Agriculture and Natural Resources Landscapes Management Project (SUSTENTA in Portuguese, P168940), approved June 30, 2016. 33 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. 34 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. Page 17 The World Bank Water Resources Development (P107350) The achievement exceeded expectations as the original plan was to only undertake feasibility studies. Based on these outputs, the three dams were immediately ready for tendering. The Gorongosa Dam was designed to provide drinking water for the Gorongosa municipality, with the possibility of also providing for irrigation and to generate small amounts of power. The Metuchira Dam was designed to provide drinking and irrigation water for local villages, along with a small amount of power for local users downstream. Finally, the Ressano Garcia Weir was designed for the Incomáti River to measure compliance with the Interim IncoMaputo Agreement and the provision of bulk water supply to the border town of the same name. The project worked in close collaboration with the AfDB, which separately financed the feasibility study for the Mapai dam on the Limpopo River. As a result of these studies, the Gorongosa dam is currently under construction under GoM financing. With the construction of the Moamba Major dam,35 designed and financed by the Government of Brazil, the Ressano Garcia weir was deemed no longer necessary. The GoM is currently raising funds for the construction of the Metuchira and Mapai dams.36 36. In response to the 2013 floods that devastated parts of the Limpopo River basin, the Bank responded in part by channeling the AF through the project to rehabilitate flood protection dikes and other hydraulic infrastructure. The AF funded civil works, TA, and supervision to reconstruct dikes, weirs, and levees in communities along the Limpopo River, in one of the country’s most significant agricultural areas. The works consisted of repairing damaged sections of the infrastructure, widening the base of the dikes, reinforcing slopes and foundations. Two irrigations canals were also constructed in two districts, supplying water to two important irrigation schemes producing rice, maize and horticulture. A manual for operations and maintenance (O&M) was issued to help define responsibilities for proper upkeep. All in all, 50 kilometers (km) of dyke system were rehabilitated under the project, protecting a population of about 287,100 people and 45,000 hectares (ha) of agricultural fields. The rehabilitation works were carried out in a timely manner, restoring flood protection and all-season irrigation and water supply ahead of the 2014 rainy season.37 However, the works did not achieve the original ambition of rehabilitating the dikes to sustain a one in 50-year flood.38 To achieve this, the dike system would have had to be rebuilt as a much larger system, which was beyond the scope, budget, and timeframe of the project.39 Instead, rehabilitation works restored the dikes to their pre-flood capacity, which the GoM assessed as adequate for a one in 20-year flood event. Adjustments to the results framework should have been made to reflect this. 37. The project also rehabilitated the Macarretane dam, part of a multifunctional infrastructure system providing flood protection, all-season water supply, irrigation, and transportation. The rehabilitation works focused on three main activities: the hydromechanical works, the gate control system and reinforcement of the foundation. The hydromechanical works consisted of protecting the mechanical structures that had undergone progressive corrosion. Improvements were also carried out on the gantry and other metallic structures associated with the hydromechanical system. The gate control system activities included the replacement of the damaged electrical panels and water level sensors, as well as the restoration of the automated gate-control system, critical to managing the operation of the dam’s 39 gates in a safe and efficient manner. Reinforcing the foundation entailed the rehabilitation of the riverbank 35 The dam will have capacity to store 760 million cubic meters of water, control the flow of the Incomáti river, increase water supply to the cities of Maputo, Matola and Ressano Garcia and have a power plant to produce 15 megawatts of electricity. 36 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. 37 The Aide Memoire of the Implementation Support Mission of December 2014 reported that rehabilitation of prioritized dykes had already been completed for four of the five lots, with the last being 75 percent completed. The assessment was that the works were timely in restoring flood protection ahead of the 2013-14 rainy season. 38 The intermediate indicator was introduced as part of the additional financing. See Project Paper (Report No. PAD702). 39 De Almeida, Luis (Project Coordinator, DNGRH). Phone interview by John Factora. September 29, 2020. Page 18 The World Bank Water Resources Development (P107350) containment walls, construction of two sheet pile walls and reno-mattresses to hold the soils in place, and the construction of the energy dissipation device. The latter was undertaken vis-à-vis the Bank-financed Emergency Resilient Recovery Project (ERRP).40 The NWRDP also financed the design for the rehabilitation of the downstream apron, with the works carried out under the ERRP. Completion of these works has had an important economic impact, given the role the Chokwe irrigation system plays in one of the main irrigated areas in the country, which the Government recently declared as a Special Economic Zone. The inauguration of the rehabilitated dam took place in September 2020 and the dam is once again supplying water to about 12,000 producers across 35,000 hectares in Chokwe. Objective 3. To increase the yield of the Corumana Dam to augment water supply for GMMA (Rating: Modest / Substantial) 38. The objective was partially achieved before the November 2019 restructuring and substantially achieved afterward. The project sought to complete the Corumana Dam, which was left without spillway gates and other features when constructed in 1989, thereby resulting in increased reservoir capacity, dam yield, bulk water supply, and climate resilience (see Box 1). The project financed the design, supervision, and the hydromechanical works, the completion of which has allowed the dam’s FSL to rise from 111 mASL to 114 mASL and to increase the reservoir storage capacity from 720 million cubic meters (Mm3) to 960 Mm3. These changes, in turn, increase the dam’s yield from 299.8 Mm3/year (yr) to 311.4 Mm3/yr, resulting in an incremental yield of 11.6 Mm3/yr at a 98 percent confidence interval. As a result, bulk water supply available from Corumana to the GMMA is calculated to increase from 69 Mm3/yr to 81 Mm3/yr (see Table 2). More water is also available for irrigation. As depicted in the Theory of Change (Section I.A.2), a separate Bank- financed project, the GMWSEP, will allow for the actual abstraction, treatment, conveyance, and distribution of water to households in the GMMA. The GMWSEP is set to close in September 2021 with the water treatment plant already commissioned and with the project on track to distributing 60,000 m3/day to the GMMA. 39. With the completion and testing of major works at the Corumana Dam, the formal hand-over to the GoM was completed. The saddle dam and the main spillway have been completed, with the six spillway gates having been installed in April-May 2020, after extensive delays in their design, manufacturing, and delivery. Other critical elements of the spillway works included repairs to the concrete pillars and abutments; the installation of the hydromechanical and control equipment; widening of the dam embankment; the installation of the monorail crane, stoplogs, and maneuvering gantry; and the excavation of a new toe drain. The rehabilitation of the intake tower hydromechanical equipment and construction of the control room were also completed. A few finishing works were undertaken after the project closed, including the installation of new rubber seals, alignment of the gates within the bay, repair work on the crawler of the monorail crane, and adjustments and updates of the supervisory control and data acquisition (SCADA) and interfaces, as required for the central control system.41 The dam monitoring system was installed, with the pressure relief wells and v- notch outlet structures completed to measure seepage from the toe drain. The monitoring system also includes a telemetric connection to hydrometric stations located upstream (in South Africa) and downstream, the information from which will be processed by the SCADA system. The installation of the piezometers used for monitoring flow has been delayed due to the COVID-19-related restrictions preventing delivery of the equipment from South Africa. The equipment 40 Mozambique Emergency Resilient Recovery Project (P156559), approved September 29, 2015. 41 See the 10th Report by the Panel of Experts for the Mission of November 25-December 7, 2020. Page 19 The World Bank Water Resources Development (P107350) is now on site for installation.42 Once the hand-over certificate is issued, the hydro-mechanical works undergo a three- year defects notification period for the contractor to address any defects before a final performance certificate is issued. Table 2. Corumana Dam: Key Performance Indicators Before the Modified Project Original Project Project Design Design Full supply level (FSL) 111 mASL 114 mASL 117 mASL Storage capacity 720 Mm3 960 Mm3 1,240 Mm3 Yield43 299.8 Mm3/yr 311.4 Mm3/yr 322.1 Mm3/yr Incremental yield44 n.a. 11.6 Mm3/yr 22.3 Mm3/yr Water supply for GMMA 69 Mm3/yr 81 Mm3/yr 86 Mm3/yr 40. The three-year defects notification period also allows time for the reservoir to fill. Simulated wet tests were conducted in the Fall of 2020 and in in the Spring of 2021, during which stoplogs were lowered into each gate and the space between the stoplog and the gate filled with water. The tests indicated that the gates and seals were safe and functioning as specified. However, issuance of the performance certificate is dependent on wet tests conducted under “normal operating conditions”, requiring the gates to be submerged and the reservoir filled. This could take an additional year for enough water to fill the reservoir, despite the 2021 La Niña weather phenomenon producing heavier than normal precipitation.45 As of early-May 2021, the water level of the reservoir was 109.8 m, just short of reaching the gates.46 41. Other dam safety measures are proceeding well. The O&M manual, including the instrumentation manual, has been reviewed and commented on by the Bank and PoE and needs to integrate details of the equipment and instruments before final clearance. The final O&M manual is expected to be cleared by June 30, 2021 (see handover certificate in Annex 8). The Emergency Preparedness Plan (EPP) has been cleared by the Bank and PoE but needs to be operationalized. This process involves dissemination and coordination with key stakeholders, including local authorities, disaster management authorities, community leaders, basin management committee members, police and other personnel as defined in the EPP. This process has been hampered by COVID-19 related restrictions on mobility. Future dam safety measures include semi-annual routine dam inspection and dam performance reports carried out by the dam operators, as well as biennial dam inspection and dam performance report by the regional dam safety entity. 42. The works on Corumana Dam was accompanied by a training program for ARA-Sul dam operators on dam O&M, instrumentation, and dam safety measures. The program included some training in China provided by the manufacturer of the radial gates and training on-site. The trainees expressed satisfaction with the training and confidence in their ability to operate the dam. However, there was discussion across ARA-Sul over the missed unique opportunity for dam personnel to get more hands-on professional and technical training during the civil works and installation of equipment. The GoM lamented that more Mozambican engineers could have been seconded to work with the resident supervising 42 De Almeida, Luis (Project Coordinator, DNGRH). Conference call with project team. March 2, 2021. 43 Yield is defined as the volume of water that can be supplied from a reservoir with a confidence interval of 98 percent. Incremental yield is the additional volume of water that results from an increase in the FSL of the reservoir. 44 The outcome indicator for Objective 3 was originally set at 16.8 Mm3/yr assuming FSL at 117 mASL. Subsequent hydrology analysis indicates that actual incremental yield is 22.3 Mm3/yr. 45 De Almeida, Luis (Project Coordinator, DNGRH). Phone interview with John Factora. September 29, 2020. After several years of El Niño-induced drought, the southern Africa region is expected to undergo a La Niña weather phenomenon in 2021, with a high likelihood of excessive rainfall. 46 Based on weekly reports, the water level in the reservoir rose by more than 11 meters, from 98 m to 109.8 m between January 25, 2021 and May 10, 2021. The reservoir is now about 60 percent filled. Page 20 The World Bank Water Resources Development (P107350) engineers at the dam for on-the-job training, and that outreach and exchange programs with local universities could have been planned for student engineers and natural resource managers.47 Box 1. Augmenting Water Supply to the GMMA48 The drought that affected much of southern Africa in 2017-18 and which led to the “Day Zero” water crisis in Cape Town, South Africa also triggered strict water rationing across the GMMA. In many parts of the city, taps were turned off every other day and irrigation was banned. Years of progress on household water connections were reversed, as residents returned to the daily burden of filling containers with water from community taps. The Pequenos Libombos Dam, the main source of water supply to GMMA, had dropped to about 15 percent of capacity, as a result of five years of below average rainfall in the Umbeluzi River Basin. This forced the Umbelúzi WTP to reduce treatment and distribution to about 120,000 m3/day, roughly 60 percent of installed capacity. As an emergency measure, the GoM built a compact WTP at the base of the Corumana Dam in 2018-19 to supply 30,000 m3/day of treated water for the GMMA. The GoM also opened 22 small water supply systems throughout the city. With the completion of the hydromechanical works, Corumana Dam has significantly increased bulk water capacity. With the main WTP coming online, an additional 60,000 m3/day of treated water is set to benefit 560,000 residents. A planned Phase 2 of the treatment plant is expected to double the capacity to 120,000 m3/day over the medium term, benefiting more than a million people. Long-term projections are for Maputo to double in size, from 1.7 million people to 3.5 million by 2035, requiring production capacity to also double. Discussions are already underway to construct the auxiliary spillway at the Corumana Dam, which would further increase water production and help ensure long-term water security for the GMMA. 43. In addition to on-going dam safety measures, there are some pending ESS activities. The environmental management plan (EMP), the RAP, and the LRP recorded good progress, particularly in the last year of implementation, but they are all at various stages of completion. The project financed the construction of 133 new houses in four different sites, along with access roads, water supply system, electricity supply, and rehabilitation works at the local school and health center, all of which benefited both project-affected persons (PAPs) and host communities. All told, 660 PAPs were resettled. Each household was compensated with land (around two hectares per family) prepared for agriculture purposes and in cash for fruit trees and moving allowances. The closing of the project amidst the COVID-19 pandemic led to logistical and budget-related delays in carrying out remaining activities. The Bank has followed up closely with GoM at the technical and ministerial levels. The MOPHRH engaged with local administrative offices to accelerate issuance of land and property titles and to provide agricultural extension services to PAPS. Resettlement kits containing farming and household tools were sourced from other government agencies and have been distributed to families, as were seeds (maize and beans). Arrangements have also been made to mobilize technical support from a government financial institution to assist with the training of communities on financial management. And after some delay, the MOPHRH was able to mobilize internal budget resources to complete outstanding activities related to the RAP (see Annex 8). These include construction of auxiliary structures (e.g., external kitchens, barns), some compensation for “social structures” (e.g., commercial stalls, garages, warehouses), and relocation of existing gravesites. As for the EMP, several activities are pending, including the development of a risk assessment for the operational phase of the dam, including the specific 47 Independent Project Completion and Evaluation for the National Water Resources Development Project, Ministry of Public Works, Housing and Water Resources, National Directorate for Water Resources Management, Government of Mozambique, June 2020. 48 Based on: “Mozambique prays for rain as water shortages hit country’s poor.” The Guardian. April 22, 2018, www.theguardian.com/environment /2018/apr/22/mozambique-prays-for-rain-water-shortages-hit-poor-maputo-southern-africa-drought; “Drinking water from Corumana dam available from July.” Club of Mozambique. February 19, 2019, https://clubofmozambique.com/news/drinking-water-from-corumana-dam- available-from-july/. Page 21 The World Bank Water Resources Development (P107350) requirements for oil spill and fire management. ESS audits will be conducted once all activities are completed, which the GoM now projects to be April 2022. 44. The success in completing the Corumana Dam and augmenting bulk water supply to GMMA is tempered by the late restructuring of the outcome indicator target. As described in paragraphs 16-17, the incremental yield of 16.8 Mm3/yr was reduced to 11.6 Mm3/yr due to hydrological studies and new regulations, which necessitated a redesign of the dam, including a new auxiliary spillway in the left embankment (to be completed under a separate project) to ensure safe discharge during maximum flood events. Once the auxiliary spillway is constructed, the FSL, reservoir capacity, and yield level would increase to their original design (see Table 2).49 The decision on the redesign was undertaken in 2017 but it was not until November 2019 that the restructuring was completed. While the objective was substantially achieved after the restructuring took place, it was partially achieved for most of the implementation period. B2. Justification of Overall Efficacy Rating Table 3. Summary of Key Outcomes Achieved by Project Objective Outcomes Baseline Achieved Comments Objective 1. Strengthen the management of national water resources This outcome exceeded expectations. The project originally International waterways agreements 0 2 sought to review and update one agreement. Instead, two were signed and under implementation reviewed, updated, signed, and are under implementation. Flood management preparedness in the This outcome is part of a broader effort to improve flood Limpopo is informed by a decision No Yes management in the Limpopo and reflects a novel approach of support system (DSS) under operation integrating various strategies into one viable system. Objective 2. Strengthen the Development of National Water Resources River basins covered by basin The project financed 3 river basin plans and leveraged financing 5 8 development plans (PDO indicator) from GoM and development partner to finance an additional 4. Macarretane Dam rehabilitated and Following the 2013 Limpopo floods, the project rehabilitated the supplying water to Chokwe irrigation No Yes dam, which is back to supplying water to about 12,000 farmers scheme across 35,000 ha in the country’s largest irrigation scheme. This outcome exceeded expectations. The target was to complete 3 feasibility studies for future dam projects. The project Pipeline of viable water storage projects 0 3 completed feasibility studies, ESIAs, and detailed technical ready for tender designs, one of which is already under construction. The project also leveraged AfDB financing for one other dam. Objective 3. Increase the yield of the Corumana Dam to augment water supply for the GMMA Incremental reservoir yield available 11.6 Completion of works allows reservoir storage capacity to from Corumana for GMMA (PDO 0 Mm3/yr increase, resulting in increased dam yield. indicator) 69 81 Increased dam yield results in increased bulk water supply for Bulk water supply for GMMA Mm3/yr Mm3/yr GMMA. 49According to the Aide Memoire for the supervision mission of November 25-December 7, 2020, the supervising engineering firm (AECOM) designed the new spillway on the left abutment to allow maximizing flood attenuation. In the proposed concept the new spillway would act as the main spillway and the gates of the existing spillway would be used only to discharge severe and extreme floods downstream. This differs from original concept. Discussions between the POE, GoM, and AECOM continue as of this writing on designs options. Page 22 The World Bank Water Resources Development (P107350) 45. Overall efficacy rating is Modest prior to the December 2017 restructuring and Substantial afterward. Objective 1, Strengthening the Management of Water Resources, was substantially achieved. Objective 2, Strengthening the Development of Water Resources, and Objective 3, Increasing the Incremental Yield to Augment Water Supply to the GMMA, were both substantially achieved after the restructuring of their respective targets. The changes require a split rating, as presented in Section II.D. Despite the Bank’s failure to ensure timely consistency between outcome indicators in the results framework and project scope, the project achieved many of the envisioned results, including the outcomes (both those in the results framework and those inferred from the PAD) related to each of the three PDOs, as shown in Table 3. At the output level, 14 of the 17 intermediate indicators were achieved (see Annex 1). C. EFFICIENCY C1. Assessment of Efficiency 46. The analysis for Efficiency examines the economic viability of the project compared to results envisioned at appraisal. The focus of the economic analysis at appraisal was on the benefits to be accrued from increased water supply for households in the GMMA. As the scope of the NWRDP did not extend beyond completion of the dam, including dam impoundment, the analysis takes into account the auxiliary infrastructure, such as transmission pipeline from the dam to urban area, WTP and distribution pipe network, all of which are being carried out under the Bank-financed GMWSEP. The economic analysis considers costs and benefits arising from components of both projects. Table 4. Investment Costs at Appraisal vs. Actual Costs (in millions) Actual PAD Investment Nominal Prices 2011 Prices US$ MZN US$ MZN US$ MZN Corumana Dam completion Civil work 40 1,162 37 2,315 23 1,475 Detailed design and 2 58 5 244 3 156 construction supervision Greater Maputo Water Supply Expansion Project50 Water treatment plant 58 1,698 34 2,154 22 1,372 Transmission 74 2,168 77 3,066 67 2,696 Distribution network 18 522 18 1126 12 716 Subtotal without auxiliary 192 5,608 171 8,905 127 6,415 spillway Auxiliary spillway (planned) n/a n/a 40 2,543 23 1,418 TOTAL n/a n/a 211 11,448 150 7,833 47. Several factors impacted project efficiency. One of them was the delay in implementation. The original plan was to complete substantial civil works by 2016 and gain project benefits afterwards. But actual implementation of the 50The costs under Greater Maputo Water Supply Expansion Project shown in this table are the full contract amount to achieve the project target of 60,000 m3/day supply. In the economic analysis, these values were deducted using the six-tenths-factor rule to estimate the cost portion that matches the attainment of the objective of this project (46,027 m3/day). Page 23 The World Bank Water Resources Development (P107350) major civil works contract did not begin until 2017. Another major change was the cancellation of the fuse plug at the saddle dam and the decision to add an auxiliary spillway to the design. As noted in earlier sections, until the auxiliary spillway is built and operational, the Corumana dam reservoir needs to operate at a FSL of 114 mASL, compared to the original FSL of 117 mASL, resulting in an incremental water yield of 11.6 Mm3/yr, compared to the original target of 16.8 Mm3/yr. According to the current design of the auxiliary spillway, construction costs is estimated at US$40 million and will take almost two years to build. To maximize the incremental water yield until the completion of the auxiliary spillway, the GoM proposes to allow the FSL to rise to 117 mASL toward the end of the rainy season to harness as much water as possible at a time when the probability of flood occurrence / disaster is low. 48. The costs and benefits were converted to 2011 prices in order to compare them against what was projected at appraisal (see Table 4). Taxes and duties included in financial costs of the project were deducted to convert to economic costs. A 45-year project life period and O&M cost at 0.3 percent per year against capital expenditure were assumed. A revised discount rate of 6 percent was used, instead of 12 percent used at appraisal, following the Bank guidance issued in 2016. In addition to annual O&M cost, the analysis also followed the assumption at the appraisal stage that major facility rehabilitation fee is included every 15 years. 49. Benefits accrued from the Corumana dam completion were estimated using the same approach as that done at appraisal. Specifically, assumptions included i) household benefits by increased water supply (valued in the form of willingness to pay, W2P); ii) reduced flood damage; iii) increased hydropower generation and iv) increased catch of local fisheries. The W2P is considers the positive health benefits accrued from improved water supply, hence no separate valuation of health benefits is considered either at appraisal or the implementation completion and results report (ICR). The benefit from increased water supply comprised the majority (90+ percent) of total benefits. At appraisal, the W2P used was US$1.13/m3 based on the average water tariff of MZN 32.77/m3 adopted by the water utility of Maputo. The analysis used the revised W2P of US$1.06/m3, considering the decreased GDP per capita over the project period. Table 5. Results of the Economic Analysis NPV EIRR (US$ M) PAD 2.3 12.3% Actual Scenario 1: Dam operates up to 114 masl until the completion of auxiliary spillway. After the 24.0 9.9% completion of auxiliary spillway, the dam operates up to 117 masl Scenario 2: Dam operates up to 114 masl during the rainy season but 117 masl during the dry 26.4 10.4% season until the completion of the auxiliary spillway Scenario 3: Auxiliary spillway is not constructed, and dam operates up to 114 masl permanently 15.2 9.0% 50. Economic efficiency is rated Substantial. As Table 5 shows, results of the EIRR did not reach the level of appraisal, but all the scenarios exceeded the revised breakpoint of 6 percent economic internal rate of return (EIRR), with positive net present values (NPV). The largest factor of this downturn is attributed to increased cost due to the cancellation of the fuse-plug function of the saddle dam and the need to construct the auxiliary spillway instead. The unforeseen design change also caused a delay in benefit yield, since the Corumana dam reservoir needs to be operated at 114 mASL, instead of the original target of 117 meters, until the completion of auxiliary spillway. The scenario analysis as well as sensitivity analysis indicates that the most economically viable scenario is the second scenario (construction of auxiliary spillway), whereas the least viable scenario is cancellation of auxiliary spillway. Annex 4 includes a full analysis on efficiency. Page 24 The World Bank Water Resources Development (P107350) D. JUSTIFICATION OF OVERALL OUTCOME RATING 51. Overall Outcome is rated Moderately Satisfactory (MS). The rating reflects High Relevance; a split Modest- Substantial Efficacy; and Substantial Efficiency. A split assessment was undertaken due to the two project restructurings that reduced the target values of two outcome indicators. This led to split outcome ratings of MU, MS, and MS, which because of the disbursement-based weightage assigned to each, resulted in an Overall Outcome of MS (see Table 6). Table 6. Split Rating Analysis Without December 2017 November 2019 Restructuring Restructuring Restructuring Relevance High Efficacy (PDO) Modest Substantial Substantial Strengthen the management of national water resources Substantial Strengthen the development of national water resources Modest Substantial Increase the yield of the Corumana Dam to augment water supply Modest Substantial Efficiency Substantial Outcome ratings MU MS MS Numerical value of the outcome rating a 3 4 4 Disbursement US$40.09 million US$28.3 million US$18.52 million Share of Disbursement b 46% 32.5% 21.3% Weighted value of the outcome rating 1.38 1.3 .852 Final Outcome rating MS (1.38 + 1.3 + 0.852 = 3.532) a Note: Highly Unsatisfactory (1); Unsatisfactory (2); Moderately Unsatisfactory (3); Moderately Satisfactory (4); Satisfactory (5); Highly Satisfactory (6) b Out of $86.91 million disbursed III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 52. The project was prepared with solid analytic underpinnings, strong legal and policy framework, and extensive national processes to address the development of the water sector. The Bank prepared the Country Water Resources Assistance Strategy (CWRAS) for Mozambique51 that provided the broad analytical underpinning to help the GoM and its development partners prioritize water resources interventions to maximize long-term poverty reduction and economic growth, reduce vulnerability to water shocks, and to narrow the country’s infrastructure deficit.52 A tripartite study by Mozambique, South Africa and Swaziland on the “Augmentation of Water Supply to the City of Maputo and its Metropolitan Area” in October 2009 and the 2011 “Master Plan for the Greater Maputo Water Supply System” helped identify water sources and priority investments to improve bulk water supply for the GMMA. Moreover, the broader National Water Resources Development Program, to which the project belonged, had a national scope with a policy and institutional reform agenda to implement the country’s Water Policy, the National Water Resources Management 51 Making Water Work for Sustainable Growth and Poverty Reduction, Country Water Resources Assistance Strategy for Mozambique, 2008-11 (Report No. 56855), World Bank, August 2007. 52 Country Partnership Strategy for the Republic of Mozambique, FY08-11 (Report No. 39395-MZ), April 24, 2007 Page 25 The World Bank Water Resources Development (P107350) Strategy, and the Regulations for Licenses and Concessions. The program also entailed a series of strategic investments to address climatic variability while increasing multi-sectoral coordination. 53. Lessons learned informed the design of the project. The project was informed by lessons learned from the Bank’s engagement in the water sector in Mozambique since the end of the civil war. The World Bank’s 2003 Water Resources Sector Strategy (WRSS) provided the guiding principles for the CWRAS and the project preparation process drew on lessons derived from the implementation of the 2003 WRSS as well as the recommendations of the IEG report on water and development. These lessons highlight the importance of properly aligning the implementation of environmental and social measures during the early stages in design of complex water resources development projects. They also highlight the need to consider interventions within an integrated basin-wide management framework that explicitly considers the cumulative impacts of all developments, not only of the specific project. Moreover, the DNA had implemented the Bank- financed PNDA, the ICR for which raised concerns over sustainability, the need for a comprehensive M&E system, the need to ensure strategic TA in place early to strengthen management and leadership during the initial phase, and the need to strengthen contract management during implementation through early mobilization of the project administration and management team (PAMT) and establishment of robust systems for managing procurement.53 54. Preliminary designs were completed, but final, detailed designs were deferred to implementation. A feasibility study for the “Raising of the full supply level of Corumana Dam” was completed in 2003 under the Bank-financed PNDA. During project preparation, a Technical Services and Dam Safety (TSDS) consultancy completed basic designs and updated cost estimates for the six crest spillway gates and repair works of concrete pillars, saddle dam (earth fuse plug) construction, monitoring system and equipment along with other safety measures based on hydrological, topographical and geotechnical studies. The Policy and Human Resources Development Fund (PHRD) financed other preparatory technical activities. A project preparation advance (PPA) recruited the specialists in dam design, engineering geology, hydrology, and ESS, together comprising the dam PoE to advise on the preliminary design work. Provisions were made to ensure their continued engagement during implementation. Despite strong preparatory work, detailed designs were not done until after project approval, which severely affected implementation timeline, project management, and M&E. 55. The M&E framework reflected elements of the broader program. The NWRDP anchored the broader program, which included support for institutional and policy reforms and strategic investments in the water sector supported by various development partners. The Bank alone financed four operations as part of the program.54 While a program approach allows for clear efficiencies in project preparation, donor coordination, implementation, and administration, each project needs to be bound by the project’s scope. However, the M&E framework of the NWRDP reflected results to be achieved beyond the project. As such, and as discussed in paragraph 22, some intermediate and outcome indicators could not be effectively used to monitor and evaluate achievement of the objectives (see also Section IV on M&E quality). 56. There were shortcomings in the identification and assessment of risk and mitigation measures. As was common among Bank projects prepared at the time, the risks identified in the project’s Operational Risk Assessment Framework (ORAF) contained a variety of risks, including many at the sectoral or country levels, that did not necessarily reflect the 53Project Appraisal Document for the National Water Resources Development Project, Report No. 63794-MZ, August 29, 2011. 54This included the additional financing to the NWRDP. The other two projects implemented concurrently by the NDA included the Climate Resilience: Transforming Hydro-Meteorological Services Project (P131049) approved April 2013; the Enhancing Spatial Data for Flood Management Project Risk Management Project (P149629) approved in June 2014. The same task team prepared the GMWSEP which was implemented by another agency. The PAMT also co-implemented the Emergency Resilient Recovery Project (P156559) approved in September 2015. Page 26 The World Bank Water Resources Development (P107350) direct, inherent risks to achieving the project’s development objectives. Other risks, such as institutional capacity (including fiduciary capacity), were assessed as Medium, reflecting confidence over the dedicated centralized PAMT embedded within DNA, and taking into account capacity building measures in DNA during a previous Bank project.55 While delivery quality risk was assessed as High, with concerns raised over the quality of consultants during project preparation, overall risk rating for implementation was rated Medium, which, in hindsight was overly optimistic given the complexity of the project, the large works contracts involved, ESS issues, and institutional capacity. B. KEY FACTORS DURING IMPLEMENTATION B1. Factors Within Government Control 57. Implementation arrangements did not materialize as envisioned. Project implementation centered on the PAMT, conceived as a centralized administrative team comprised of government officers and individual consultants embedded within DNGRH.56 The PAMT was to be supported by a strategic, multidisciplinary TA program provided by a private firm to act as a sounding board on technical, administrative, and managerial matters. 57 This vision did not materialize. Civil servants could not be enlisted to work on the project, given what was perceived to be (a) formidable differences between government and Bank procedures (e.g., procurement, ESS); (b) additional workload; and (c) lack of incentives. On the latter, new GoM regulations prohibited external project funding to supplement civil servant salaries. These interrelated issues resulted not only in a PAMT comprised of consultants, but in the inability of the PAMT to draw in technical civil servants to provide inputs when needed (e.g., on terms of references). The strategic TA similarly did not provide the envisioned support to the PAMT. Procurement delays meant the contract began two years into implementation; more importantly, it was widely viewed as costly and inefficient, with foreign consultants brought in to advise on issues (e.g., legal) without proper understanding of the Mozambican context.58 It was also seen as a “parallel structure” and not fully embedded in the PAMT, which made it difficult to engender trust and strong collaboration.59 As such, the Bank team and the PoE played an oversized role in assisting the client with its technical functions. After four years, the contract for the TA ended without an extension. 58. PAMT capacity did not keep pace with the growth of the DNGRH portfolio. The PAMT was configured to manage and administer the NWRDP. However, in quick succession and within a three-year period, PAMT staff were deployed to work on the 2013 Limpopo floods and the associated AF, the PPCR project, the Spatial Data (LiDAR) project, and the ERRP project – all Bank financed – without a concomitant increase in staffing. As a result, the PAMT was overstretched, affecting the level of involvement, quality, and timeliness of project activities. Despite the clear need to reinforce PAMT capacity fairly early on, it was not until 2014 that one additional professional was recruited and then in 2015 a cadre of specialists was brought on board to help implement all Bank-financed projects in the DNGRG portfolio. 55 See Mozambique National Water Development Project I (MZ-PA-52240), approved May 1999. 56 The PAMT was to be in keeping with the principles agreed to among the Group of 19 donors in Mozambique, the lessons learned from previous programs implemented by the DNA and reflected the need for cost containment and dedicated implementation capacity for all DNA projects. 57 The PAMT was responsible for supporting DNA in implementation, coordination and management of all project activities related to the following: i) procurement processing; ii) disbursement of project funds; iii) replenishment of the Designated Account; iv) processing of withdrawal applications; v) maintenance of project documentation, including Designated Account reconciliations and M&E reports; vi) technical support, including engineering management and oversight of environmental and social aspects; and, vii) communications, M&E functions. 58 De Almeida, Luis (Project Coordinator, DNGRH). Phone interview by John Factora. May 16, 2021. 59 Muximpua, Odete (Water Supply and Sanitation Specialist, World Bank). Personal communication, February 1, 2021. Page 27 The World Bank Water Resources Development (P107350) 59. Procurement processing proved to be a bottleneck. Contracts related to the Corumana Dam – for design and supervision, civil and hydro-mechanical works, and the updating of the Environmental and Social Impact Assessment (ESIA) and RAP – with a combined 70 percent of the original project financing, suffered protracted delays. The lack of technical rigor and poor quality of procurement packages led to extended exchanges between the Bank and GoM. Separate internal reviews and approvals of the contracts by the country’s administrative tribunal and the External Economic Relations Commission also caused delays. All told, the contract for the ESIA and RAP updates took two and a half years to process and was signed in mid-2014; that for the design and supervision contract took three years and was signed in April 2015. The time needed to finalize designs meant that the tendering for the works contract was not signed until May 2017, seven months before the original closing date (Figure 2). Moreover, delays in procurement for the Corumana Dam had a cascading effect on non-dam related contracts. Given the volatility of construction costs, GoM (at Bank’s advice) prioritized the Corumana Dam contracts to ensure financing availability. Only after those contracts were finalized was the financing envelope for non-dam related activities clear. While necessary, the sequencing and prioritization delayed progress on other components. The delays also risked undermining the synchronization of the Corumana Dam completion with the construction of the WTP under the GMWSEP. 60. Delayed procurement was followed by delays in ESIA and RAP updates and implementation. The updating of the environmental and social impact assessments (ESIA) and RAP took about four years to complete and was finally approved by GoM in April 2018, seven and a half years after project effectiveness. The updating was marred by several issues. New, more demanding governmental regulations and guidelines issued on resettlement forced major revisions to both the ESIA and RAP. The changes to the design of the dam (discussed in paragraphs 16-17) also required major revisions to both instruments. The delays, in turn, made it challenging to finalize the numbers of PAPs, as periodic updates to the census and socioeconomic surveys recorded new families settling into the dam surroundings.60 Finally, GoM also raised concerns about the quality and timeliness of the consultant’s work.61 Once the RAP was approved, land acquisition and project financing of the resettlement housing advanced relatively quickly. However, other elements of the RAP under the responsibility of the GoM proved slow, largely due to limited capacity and budget. Aide Memoires called for the GoM to hire a dedicated service provider and additional capacity to support implementation. A consultant and service provider were finally recruited in December 2019 and February 2020, respectively, five months before the project was set to close and just prior to the onset of COVID-19 restrictions. In the meantime, many of the newly constructed resettlement houses remained empty and were subject to vandalism and stolen fixtures (doors, locks, windows), requiring on-site security.62 Resettlement accelerated once the service provider and personnel were on board (including those from the supervising firm) with almost daily site visits, including by the Bank team. But by then, time simply ran out, and as documented in paragraphs 40-43, a number of activities were left pending at project closure. 61. Inadequate project and GoM financing hampered RAP implementation. The project budget allocated to ESS implementation and compliance amounted to US$6.5 million. This proved to be inadequate to cover the costs of various service providers, as well as the development of the social infrastructure recommended under the RAP and LRP. The cost of the resettlement houses, access roads, and boreholes for the interim water system alone neared US$10 million. The electricity supply system, integrated water supply system, and works at the school and health center cost an additional 60 The Aide Memoire of the Implementation Support Mission of October 2015 noted the importance of establishing a “cut -off-date” to establish, define, and communicate RAP beneficiaries at the community and district level. 61 The Minutes of the Technical Discussions in August 2017 noted the delays by the contractor and the negative impacts on the project. 62 Omar, Salma (Senior Social Development Specialist, World Bank), Personal communication, February 4, 2021. Page 28 The World Bank Water Resources Development (P107350) US$7.6 million. In addition, consultancy services related to the updates and review of the ESIA, RAP, and EMP, along with the RAP / LRP service provision, the transporting of PAPs, the ESS auditors, and the preparation of agricultural land amounted to US$5.9 million. Much of the budget shortfall was addressed through funding reallocations from activities related to water resources development and through contingencies. The US$5 million that was cancelled and reallocated to another Bank project to support the response to Cyclones Idai and Kenneth in 2019 compounded the situation, as did exchange rate fluctuations (see below). The GoM, for its part, was responsible for land acquisition and paying cash allowances and compensation to the PAPs. Aide Memoires from the early phases of the project noted the GoM’s allocation from the state budget for such purposes. However, by the time RAP implementation began, the country had entered into fiscal crisis, severely limiting funds for RAP implementation.63 The option of having the project fund cash compensation was considered in 2017 but was not followed through.64 Figure 2. Timeline of Corumana Dam-Related Contracts65 62. With only a two-year window to complete the works at the Corumana Dam, the timeline for civil and hydro- mechanical works needed to be strictly adhered to. However, implementation was marred by a number of challenges. While substantial progress on resettlement needed to be made before works could begin in earnest,66 even mobilization and preparatory activities were slow to commence. Project documents and communications note the lack of experienced and skilled personnel on-site; the slow pace and uneven quality of the works; and quality issues in the manufacturing of the radial gates in China and the delays in transporting them on-site.67 In addition, there were several instances where work stopped altogether, including: a five-week suspension in 2019 due to non-compliance with health and safety requirements;68 intermittent labor stoppages and tensions in the local community;69 and a two-month hold on works on the saddle dam to allow for demining of the area upon the discovery of unexploded ordinances. The design of the 63 The Restructuring Paper of December 2017 (Report No. RES29775) references a letter from GoM in October 2016 committing to these costs. 64 Muximpua, Odete (Water Supply and Sanitation Specialist, World Bank). Personal communication, February 12, 2021. 65 Based loosely on the Independent Project Completion and Evaluation for the National Water Resources Development Project, Ministry of Public Works, Housing and Water Resources, National Directorate for Water Resources Management, Government of Mozambique, June 2020. 66 Aide Memoires of the Implementation Support Missions in March and June 2018 called for caution in beginning civil works as most of the households to be resettled were located in the works area. 67 The radial gates arrived on-site in the Fall of 2019. 68 Aide Memoire of the Implementation Support Mission of July 17-23, 2019. 69 Aide Memoire of the Implementation Support Mission of July 17-23, 2019. Page 29 The World Bank Water Resources Development (P107350) contract exacerbated delays as all activities were lumped under a single deadline. Hence the demining forced an extension of the timeline for all activities, not just those directly impacted by the demining work. The poor communications and collaboration between the works contractor and the supervising contractor also affected progress on the ground. Each of these issues was addressed. For instance, employees of the supervision contractor were posted in China to supervise the manufacturing of the gates;70 DNGRH officials spent considerable time mediating with labor and the community; a competent full-time health and safety officer was appointed to manage compliance and train personnel; and DNGRH more forcefully intervened in contract management by registering threats of sanctions. 63. Despite the myriad issues affecting implementation, GoM maintained strong ownership and a focus on results. While allocating resources (financial, technical, administrative) proved challenging throughout the life of the project, DNGRH, the ARAs, and the PAMT generally worked in a coordinated manner among themselves and with the Bank, the PoE, and other stakeholders. There was a demonstrated openness to accepting and providing differing viewpoints and to use the project as a catalyst for donor support, as exemplified in the mobilization of funds following the 2013 Limpopo floods and the completion of several river basin management plans. The Deputy National Director responsible for coordinating the NWRDP was later appointed as ARA-Sul Director which contributed positively to project coordination. All told, there were three National Directors and four Directors of ARA-Sul who were involved in the preparation and implementation of the project. Along with institutional reforms within DNRGH, these changes caused some disruption to project coordination. PAMT officials noted that the Steering and Monitoring Committees – parts of the original implementation arrangement that never materialized -- would have helped smooth over such disruptions. B2. Factors Within the Bank’s Control 64. Having a task team decentralized to the field was critical, given the level of technical and design complexity, and a challenging operating environment. All three of the project’s task team leaders were all highly experienced senior water resources managers who were field-based (one in Pretoria and two in Maputo) and who were assisted by largely field-based international and Mozambican team members. The task team was comprised of other water resources specialists, engineers, dam safety experts, generalists, and specialists in procurement, financial management, social development, environmental specialists, and communications. The location of the task team allowed for closer supervision and implementation support. This proved beneficial, given the myriad issues that arose during implementation, including the timely response following the 2013 floods and during the critical last years of implementation at the Corumana Dam, with frequent site visits and meetings. 65. The Bank responded to the January 2013 Limpopo floods with speed and agility. Within days of the flood, team members were deployed to the affected areas of the Limpopo valley to liaise with national, regional and local authorities, local and international partners, and other stakeholders to help with a preliminary needs assessment. The work provided the basis for a program of support for emergency rehabilitation of civil works in the Limpopo River basin and longer-term interventions informed through integrated flood management and mitigation studies. The rehabilitation of priority infrastructure was needed to ensure minimum flood protection ahead of the following flood season (October-March). The Bank played a convening role on the emergency response. The AF for the NWRDP for US$32 million (SDR 21.2 million) was prepared under the provisions of paragraph 11 of OP/BP 10.00 Investment Project Financing for “situations of urgent 70 Discussions with personnel from the supervising contractor during the January 2020 mission. Page 30 The World Bank Water Resources Development (P107350) need of assistance”, which allowed for retroactive financing of up to 40 percent of the AF. The AF was prepared within eight months of the flood, ahead of the following rainy season, and with emergency works completed on time. 66. However, Bank procedures contributed to delays. Discussions with DNGRH indicated that even though the project fell under safeguards arrangements, the roll out of the Bank’s Environmental and Social Framework (ESF) and the risk-based approach had a stifling effect on decision-making and pragmatic, flexible project management. DNGRH officials pointed to the additional six months it took for the Bank to approve the RAP after GoM’s approval in April 2018, during which the Bank asked for additional requirements before the civil works contractor could even mobilize and begin preparatory work. Officials noted that the additional requirements were not binding as the GoM did not have to approve the RAP for a second time. 67. Formal project restructurings were not undertaken to more fully reflect changes in project scope and design. As discussed in paragraph 55, two of the three outcome indicators were designed to measure results that could not be fully attributable to the project. During the mid-term review (MTR), the Bank and GoM seemed to be in agreement that the outcome and intermediate indicators on hydro-meteorological stations should be dropped, since these would be financed by a separate project. However, these discussions did not lead to a formal restructuring, which would have included a new outcome indicator to measure strengthened water resources management.71 Furthermore, the outcome indicator to measure strengthened water resources development was restructured in December 2017 to reduce the target from 19 to 12 river basin plans. However, the Bank team mistakenly assumed that river basin plans financed by other sources would count toward the target. Lastly, while changes to the dam’s design were undertaken in 2017, the restructuring to lower the target yield did not take place until November 2019, less than a year before the project closed. Failure to ensure consistency between outcome indicators / targets and project scope made it difficult to assess Efficacy. 68. All told, it took six years, nearly the entire original timeframe of the project, to process the main contracts and begin works on the Corumana Dam. The Bank granted a 30-month extension through two restructurings, which was adequate to complete the major works of the dam. However, critical activities, including some finishing woks, dam safety, and ESS measures remain pending – all of which have been severely affected by the COVID-19 pandemic and related restrictions. The project closed with an agreement that the Bank would continue to provide supervision support to any remaining works and activities through the end of dam impoundment and commissioning, including the full implementation of the RAP and the community livelihoods program, but that the GoM would be responsible for financing these pending activities. There was also agreement that the GMWSEP would continue to finance dam safety activities, including the work of the dam PoE, and the supervising engineering firm until the project’s closure scheduled for September 2021, while the ERRP would continue financing the core PAMT members until its closure in May 2021. The Bank team has held two virtual missions since the project closed and continues to follow up with GoM counterparts at the technical and ministerial levels to ensure completion of remaining activities. B3. Other Factors that Affected Implementation 71Project documents as early as 2012 acknowledged that hydro-climatic activities would be financed through a separate project. The Aide Memoire for the Implementation Support and Mid-Term Review undertaken in December 19, 2014 discussed the need to drop the indicator to avoid duplication with the PPCR project. However, no formal restructuring was undertaken to this end and the indicator was kept for the remainder of implementation. Page 31 The World Bank Water Resources Development (P107350) 69. Exchange rate depreciation affected project scope. The project’s initial budget was Special Drawing Rights (SDR) 43.8 million, equivalent to US$70 million, while the AF was SDR 21.2 million, equivalent to US$32 million, for a total of US$102 million. With the devaluation of SDR against the US dollar, the project budget was reduced to US$89 million. In the November 2019 restructuring, there was a partial cancellation of US$5 million requested by the Government to be reallocated to mitigate the impacts of Idai and Kenneth cyclones. Therefore, in total, the project had its final budget reduced by about US$18 million, corresponding to a 17.6 percent reduction. This amount would have been enough to fully carry out the RAP and LRP, along with a number of other activities and outputs that were proposed, such as additional feasibility studies for priority water resources infrastructure and river basin management plans. IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. QUALITY OF MONITORING AND EVALUATION (M&E) A1. M&E Design 70. As discussed in paragraph 55, the M&E framework included outcome and intermediate indicators to measure results that could not be fully attributable to the project. As the project is the unit of account, the inclusion of outcome indicators measuring broader program performance impedes project monitoring and evaluation. The indicators in question included the outcome indicators related to the hydro-climatic system (implemented through a separate project) and the river basin management plans (target relied on outputs from other sources). Intermediate indicators that were program-oriented included national water storage capacity and the upgrading of hydro-met stations in the Limpopo. Moreover, for objectives as broad as “strengthening water resources management and development”, multiple outcome indicators to cover various aspects of each objective would have facilitated the evaluation of efficacy. Instead, assessment of objectives relied on a set of outcomes not originally designated as PDO-level indicators. A2. M&E Implementation 71. Execution of the M&E functions were mixed. An M&E officer within DNGRH was appointed to the PAMT to ensure integration of information from the relevant implementing and beneficiary agencies within the national monitoring framework for the sector. Standard reporting formats were developed, and semi-annual reports submitted to the Bank, reflecting progress on project-specific indicators. The MTR was conducted as planned while the GoM’s completion report was undertaken shortly before the project closed with a strong focus on the achievement of indicators. Three project restructurings included changes to the M&E framework to better reflect the financing envelope and resource envelope, including the dropping of one intermediate indicator and the addition of six new indicators to monitor outputs related to the AF. However, the lack of proactivity on the part of the Bank to ensure consistency between the M&E framework and the project scope severely hampered the final evaluation of the project’s objectives. A3. M&E Utilization 72. The M&E framework helped inform project management and decision-making. Despite implementation issues that caused severe delays, the implementing agencies and the Bank remained focused on achieving the results set out in the project M&E framework. Supervision missions began and ended with a review of, and discussion on, each of the project indicators. These discussions centered on the underlying implementation bottlenecks affecting the achievement Page 32 The World Bank Water Resources Development (P107350) of the results. In the case of the outcome indicator on the incremental yield from Corumana Dam, the change in the target engendered a protracted discussion (several years long) on the design and economic efficiency of the civil works, as well as the broader goal of meeting the demand for water in the GMMA. Bank Implementation Status and Results (ISR) Reports assessed ratings for Development Objective and Implementation Progress, as well as other associated ratings, based on mission discussions on progress in meeting the indicators. A4. Justification of Overall Rating of Quality of M&E 73. On balance, the quality of M&E is rated as Modest. The critical design flaw and the lack of proactivity on the Bank’s side to ensure congruence between project scope and outcome indicators represent significant shortcomings in the system’s design and implementation. The shortcomings made it difficult to assess achievement of the stated objectives, despite the extensive set of intermediate indicators and a clear Theory of Change. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE B1. Environmental and Social Safeguards 74. The project was classified as an Environmental Assessment (EA) Category A, requiring full assessment. The EA reflected the expected impacts of large-scale civil and hydro-mechanical works to increase the capacity of the Corumana Dam. Consequently, the following ESS policies were triggered: OP/BP 4.01 Environmental Assessment, OP/BP 4.04 Natural Habitats, OP/BP 4.11 Physical Cultural Resources, OP/BP 4.37 Safety of Dams, OP/BP 4.12 Involuntary Resettlement, and OP/BP 7.50 International Waterways. To comply with provisions under these policies the GoM prepared and applied an ESIA and RAP. The former included a suite of Environmental and Social Management Plans, including (i) a Construction and Workers Camp Management Plan; (ii) an Inundation Preparation Plan; (iii) a Reservoir Management Program; (iv) an Environmental Water Releases Program; (v) an Environmental Monitoring Plan. The RAP also included a community LRP. 75. Given the nature, magnitude and sensitivity of the proposed investments, the project included a standalone component to ensure compliance with Bank safeguards requirements as well as applicable national regulations. An Environmental and Social Service Provider (ESSP) was hired to review and update the original ESIA and RAP (completed in 2011) to take into account new, more demanding GoM regulations on resettlement and economic impact assessments. As discussed in Sections III.B.I and III.B.2, a number of challenges led to severe delays in finalizing and implementing the safeguards instruments. Project documents systematically reported safeguards performance, and during the four years it took to prepare the updated ESIA and RAP, performance was recurrently rated as MU (Moderately Unsatisfactory). 76. The 30-month cumulative extension of the closing date allowed good progress in implementing the RAP and EMP, despite the challenges discussed in Section III. The construction works caused limited and largely temporary adverse environmental impacts, with deviations from the ESMP, mainly concerning housekeeping and waste management, and initially poor OHS performance, which was corrected in later months through penalties, training, stringent monitoring and ongoing corrective action. On social safeguards, the project financed the construction of high- quality resettlement housing, including access roads, integrated water system, and electricity, as well as the resettlement of PAPs. But as detailed in paragraphs 40-43, the project closed with pending safeguards measures, including some compensation and allowances, several LRP activities, and social and environmental audits. These activities are on-going, Page 33 The World Bank Water Resources Development (P107350) to be paid by GoM with support from the PoE and the Bank. The project closed with overall environmental and social safeguards performance consistently rated MS during much of the project implementation phase and at project closure. See Annex 7 for a full discussion on compliance with relevant safeguards policies. B2. Financial Management 77. Financial management and audit throughout the project were consistently compliant with the Bank and the country's Administrative Tribunal requirements. For fiduciary monitoring, the PAMT maintained adequate accounting system to record project funds, expenditures, and resources following the procedures outlined in the Project Implementation Manual. The project produced acceptable quarterly interim financial reports and annual financial statements. Financial management supervision was also carried out by the Bank on a timely basis, overall sustaining regular reviews of interim financial and audit reports, all of which were unqualified. The PAMT demonstrated good practices in managing funds transfers, although training and capacity exchanges were consistently carried-out to sustain optimal FM practices. Payments were made through transactions from the designated account directly to the providers of goods, works, and services, upon submitted statements of expenditures by the implementing agencies. FM performance was consistently rated Satisfactory in Bank ISRs. B3. Procurement 78. As a whole, procurement was moderately satisfactory. There were no instances of non-compliance with Bank policies or Mozambican laws. Section III.B. describes the multiple factors affecting implementation, which was marred by severe delays owing to limited capacity within the PAMT to both process procurement packages and to ensure technical quality. These limitations affected the large contracts related to the Corumana Dam in particular. In addition to the mentoring, day-to-day support, and training provided by the Bank, the implementing agencies benefited from technical support from the PoE. Once the large contracts for the Corumana Dam were completed, the smaller packages were processed with relative efficiency. The recruitment of an additional procurement officer in the second half of the project also contributed significantly to improved procurement efficiency and performance. C. BANK PERFORMANCE C1. Quality of Entry 79. The quality of entry is rated Moderately Unsatisfactory. The project reflected the strategic priorities of both the Bank and the GoM and was viewed as a cornerstone of the government’s program to improve development and management of the country’s water resources, thus attracting parallel financing by other donors. The project was underpinned by robust analytic work and strong legal and policy foundations and furthered by PPA and trust funded technical design work and completion of the ESIA and RAP. The PoE specialists were also recruited during project preparation to provide a strong level of technical support on the design work. However, as discussed in paragraphs 55 and 70, some of the outcome indicators and their targets measured program-level results, posing attribution difficulties. And while the planned implementation arrangements reflected lessons learned, they did not account for bureaucratic constraints and lack of clear incentives. The strategic TA program, which was supposed to help bolster implementation capacity, should have been ready at project entry. Most critically, the decision to defer the detailed technical designs of the Corumana Dam to implementation had a cascading effect that resulted in severe implementation delays. Page 34 The World Bank Water Resources Development (P107350) C2. Quality of Supervision 80. The quality of supervision is rated Moderately Satisfactory. In addition to the 15 formal implementation support missions, there were frequent technical discussions and the task team provided just-in-time support, problem solved, visited project sites, implemented trainings, and participated in consultations. The task team’s agility was possible partly because most team members were field based. As discussed in Section III.B.2, this proved to be a notable advantage, particularly during key moments such as the response to the 2013 Limpopo floods and in providing intense supervision support during the two-year sprint to complete the civil works and safeguards before project closure. The former saw the Bank team use its convening power to leverage parallel donor funding while the latter entailed frequent site visits and follow up. In addition to the NWRDP and AF, the same team prepared and supervised four concurrent projects with DNGRH, which allowed for joint missions, meetings, and Aide Memoires. Given constrained implementation capacity, the Bank team also played an outsized role in providing technical and administrative support. 81. However, there were also actions, decisions, and processes that adversely affected implementation. As discussed in Section III.B.2, they included the protracted process of finalizing and approving the updated ESIA and RAP; the cancellation of US$5 million for the 2019 cyclone response which compounded the already constrained budget; and the closing of the project in June 2020 as scheduled, when remaining works and safeguards were stifled by COVID-19 restrictions and which could have benefited from another limited extension. The risks stemming from the latter is tempered somewhat by continued support on key activities by other Bank projects, though these projects are set to close in 2021. Lastly, project restructurings should have been better used to ensure alignment between project scope and the outcome indicators. The misalignment hampered project evaluation. 82. Project documents raised key implementation issues, proposed remedies, set out timebound actions, and were generally of good quality. Key recurring issues included capacity constraints, procurement delays, and the need to accelerate implementation, particularly on the works for the Corumana Dam and related safeguards. Four of the 15 ISRs rated the project Moderately Unsatisfactory, or more than one-quarter of the time, reflecting candor of reporting. Aide Memoires were generally well prepared and provided a good balance between implementation details and results- orientation. However, many of the ISR Reports tended to be sparse and not particularly useful as a management tool. C3. Justification of Overall Rating of Bank Performance 83. The rating for Bank Performance is Moderately Satisfactory. In keeping with ICR Guidelines, the Overall Outcome Rating helps to determine the Bank Performance rating in cases where Quality at Entry or Quality at Supervision is in the Unsatisfactory range and the other falls in the Satisfactory range. D. RISK TO DEVELOPMENT OUTCOME 84. The risk to the project’s development outcomes is Substantial. As discussed in Sections II.B.1 (paragraphs 39- 43), the sustainability of the objective related to the Corumana Dam is dependent on completion of remaining finishing works, dam safety measures, dam impoundment and commissioning, as well as pending RAP and LRP activities. The treatment and distribution of water from the dam to households in GMMA is also dependent on the completion of the GMWSEP. The Bank sought to mitigate the risks of closing the project with remaining activities by having the ERRP and GMWSEP finance the PoE, the supervising engineering firm, and the PAMT. However, these projects are set to close at Page 35 The World Bank Water Resources Development (P107350) the end of May and September 2021, respectively, and the contracts for both the PoE and the supervising firm may need to be extended for an additional year, depending on the timeframe for dam impoundment. Impoundment, in turn, depends on rainfall and inflows into the reservoir which will likely take another year. The RAP completion requires allocation from MOPHRH, which, as of this writing, has been secured. Lastly, while the GoM’s new regulation on dam safety reflects global best practice, there is concern over enforcement of the new regulations with clearly defined role and responsibility of the dam operator, documentation, and dam safety related activities to be carried out throughout the operational life of the dam. Lack of enforcement of dam safety regulations could pose risks to long-term dam safety. 85. There are risks stemming from technical and financial sustainability. The Corumana Dam and Macarretane Barrage are owned by the GoM and operated by the Incomáti and Limpopo watershed management units, respectively, within ARA-Sul. Both units are staffed by technical personnel with degrees and experience in hydromechanical, electronic and civil engineering. Regular dam safety inspections had been carried out prior to the project and the dam safety audit undertaken during project preparation concluded that O&M was satisfactory. Once the new O&M manual is completed, dam operators will need to be trained (delayed due to COVID-19 restrictions), building on some of the training already conducted, e.g., on gate operation. In terms of financial sustainability, on the basis of the autonomy granted to ARA Sul, capital investment is supported by the State Budget, while O&M expenses are borne by the revenue generated through its activities and eligible for subsidies of up to 44 percent by the State Budget. Thus, with the entry into operation of Corumana Dam and Macarretane barrage, ARA-Sul secured funds from the government budget (registered under codes MOP- 1983-0001 and MOP-2000-0002), to address the revenue shortfall. For 2020, funding amounted to MT 14.5 million, about US$240,000.72 ARA-Sul does not have an updated business plan that reflects and communicates the objectives, resources and management strategies for business development, and that integrates the emerging challenges of operating the investments made. The new functionality of the Corumana Dam and the broader Greater Maputo water supply expansion program will increase revenues, and ARA-Sul should improve its commercial focus by ensuring periodic tariffs adjustments with the state regulator, improving billing, improving collections and improving other sources of revenue. The ARAs should have a liquidity plan that demonstrates that it will achieve cash balance in the future and that the State's contribution to the operating expense will reduce over time.73 86. Other considerations must be taken into account in assessing longer-term risks. Mozambique is one of the most vulnerable countries in the world to the threats posed by climate change. The recurrent cycle of severe drought and torrential rains and floods will directly affect dam operations and yield to the GMMA. Regional climate change scenarios in the Sabíe River catchment indicate higher mean annual precipitation in the immediate future. Higher rainfall could lead to an increased inflow to the dam but less inflow at critical times. However, higher temperatures could also increase evaporation rates of the reservoir as well as an increased demand for downstream irrigation. The reservoir operation will require the balancing of flood control with the objective of maximizing yield. Water abstraction in South Africa from the Incomáti River also affects flows into the reservoir and available yield to the GMMA. For this reason, the review and 72 Independent Project Completion and Evaluation for the National Water Resources Development Project, Ministry of Public Works, Housing and Water Resources, National Directorate for Water Resources Management, Government of Mozambique, June 2020. 73 Beside allocations from state budget, ARA-Sul revenues come from raw water use fees charged for the following: (i) water abstraction, (ii) extraction of sediments from river beds and (iii) discharge of effluents, as established by the Regulation of Water Licenses and Concessions (Decree 43/2007 from the Council of Ministers). The mechanisms for collection of the water abstraction fee which is the sole responsibility of ARAs (different from other uses which are a shared responsibility with institutions under the umbrella of other ministries) are well established. In fact, it should be noted that ARA-Sul to maintain its financial autonomy needs to cover at least 2/3 of the institutions’ costs as required by Law 14/2020. Raw water abstraction fees from surface sources were updated by Decree 20/2016 and from groundwater sources fixed by Decree 18/2012 both from the Council of Ministers. Page 36 The World Bank Water Resources Development (P107350) update of the Interim IncoMaputo Agreement was a critical output of the project,74 as was the development of the National Strategic Action Plan to Monitor Compliance with International Shared Watercourses Agreements. The latter recommended strengthening the Department of International Rivers from three senior staff to 15 professionals to more effectively manage the country’s nine transboundary rivers and to ensure the GoM can represent the country’s interests on shared basins at the SADC level. The GoM has signaled its commitment to this priority as negotiations with South Africa are underway to establish the Joint Commission for the Incomáti and Maputo Rivers. 87. The implementation of the river basin plans and the smaller dams for which technical designs and ESS instruments were developed depends primarily on funding availability. As noted earlier, the Gorongosa dam is under construction. Fundraising is on-going for the other dams. The three river basin development plans were completed in 2019 and their implementation is at an early phase. Given the on-going internal financial crisis, there has been relatively few investments in large hydraulic infrastructure. More investments have been directed to small excavated reservoirs scattered throughout the country. Because the river basin development plans were prepared in close collaboration by the local basin authorities, they have a greater chance of phased implementation.75 V. LESSONS AND RECOMMENDATIONS 88. A dedicated and adequately funded project component focused on ESS aspects raises the visibility of ESS issues and helps ensure timely and quality implementation. The NWRDP was one of the first Bank projects with a component dedicated to ensuring ESS compliance and implementation. It was also an early test case for supporting livelihoods of PAPs, going beyond the traditional approach to “compensation”. Project funds were allocated to the construction of resettlement housing and access roads, integrated water system, and electricity – all of which proceeded relatively quickly (once the RAP was approved) and with high quality. On the other hand, allowances and compensation to the PAPs and the implementation of the community LRP, which were the responsibility of the GoM, suffered from lack of capacity and financing. The latter was hampered by the country’s debt and fiscal crisis that began in 2016. For large infrastructure projects with complex ESS issues, particularly in countries with limited fiscal space, project financing could help ensure the compliance and full implementation of ESS measures. 89. Contracts should be designed in a way that facilitates good contract management and builds in incentives for contractors to finish on time. For example, distinct and independent activities should fall under separate contract “sections” with distinct timelines and resources to allow for proper sequencing and timely completion. Instead, the civil and hydro-mechanical works contract for the Corumana Dam lumped all activities together under a single deadline. This meant that any external factors, such as the demining work, forced a deadline extension on all activities, even those that were not directly impacted by demining. This denied DNGRH the leverage to force on-time completion of activities through threats of sanctions for delays. Furthermore, consultancy contracts, which are not subject to penalties, should also be designed to incentivize timely completion. For instance, the ESIA and RAP contract should have been one lumpsum payment upon completion, as opposed to being time-based. 74 In October 2019, South Africa announced it would comply with the IncoMaputo Agreement for a minimum cross-border flow from the Crocodile and Komati rivers. See: https://www.engineeringnews.co.za/article/water-dept-to-engage-stakeholders-on-interim-incomaputo-agreement-2019- 09-17/rep_id:4136. 75 Inguane, Ronaldo (M&E Officer, DNGRH). Personal communication, August 28, 2020. Page 37 The World Bank Water Resources Development (P107350) 90. Having fewer, larger contracts may be more efficient administratively but could complicate implementation. For example, the civil and hydro-mechanical works contract for the Corumana Dam also included other activities that were completely distinct from the works on the dam, such as the integrated water supply system for the resettlement houses. The contractor’s inexperience with such a system meant diverting attention from the main works at the dam in order to sub-contract these activities to a design and engineering firm, and then to another works contractor. Given the delays at the dam, DNGRH decided to contract directly with a specialized contractor to perform this water supply activities, but this change in the contract resulted in several wasted months. 91. The Bank and the Borrower should ensure project design, including implementation support measures, fully takes into account any capacity constraints. While the realities on the ground did not allow the implementing arrangements to materialize as envisioned, the Bank and Borrower could have ensured adequate staffing in light of DNGRH’s growing portfolio. The Borrower must also ensure key positions are filled with capable staff. And while a program of technical assistance could potentially bolster implementation capacity, the provider should ideally be familiar with country and sector context in order to ensure advice is readily applied. 92. The Bank needs to ensure any changes to the project scope is documented in a formal project restructuring and accompanied by corresponding changes to the results framework. The project is the unit of account and the results framework should reflect results attributable to the project. The Bank and Borrower are both responsible for the quality, implementation, and utilization of the results framework. Not undertaking formal restructurings on a timely basis could negatively affect the proper monitoring of implementation and the evaluation of the project’s outcomes. 93. When deciding to close a project with pending activities, the Bank should seek firm and tangible government commitment to finance and complete each of the remaining activities. This is particularly important for sensitive activities with potentially high risks, including reputational risks to the Bank, such as resettlement compensation and dam safety. As discussed, while other Bank projects were to finance continuation of the PoE and the supervising engineering firm for a limited timeframe, other activities were left to the GoM to finance and complete. Despite an almost year-long delay, GoM funding has been secured by the MOPHRH, which should be sufficient to complete remaining activities. Short of project extension, the Bank could devise a special instrument/vehicle to continue providing minimal financing mainly for soft activities that would enable the Bank to support clients in fulfilling its obligations and mitigating risks to both the client and the Bank. . Page 38 The World Bank Water Resources Development (P107350) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Strengthen the management of national water resources. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of hydro-climatic stations Percentage 23.00 46.00 70.00 operations and reporting 15-Sep-2011 15-Sep-2011 30-Jun-2020 (AF/PPCR): % of hrdo- Percentage 23.00 46.00 70.00 climatic stations operational and reporting 15-Dec-2011 15-Sep-2011 30-Jun-2020 Comments (achievements against targets): The target was achieved but cannot be attributed to this project. The activities related to improving the country's hydro-climatic system were transferred to a separate project, making this outcome indicator obsolete for the purposes of evaluating the achievement of the objective of strengthening the management of national water resources. Objective/Outcome: Strengthen the development of national water resources. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Page 39 The World Bank Water Resources Development (P107350) River basins covered by water Number 5.00 19.00 12.00 8.00 resources management plans 15-Sep-2011 15-Sep-2011 08-Dec-2017 30-Jun-2020 Comments (achievements against targets): The target was partially achieved. The revised target of 12 (including a baseline of 5) assumed that river basin plans financed outside the framework of the project could be counted. The project financed 3 but leveraged financing from GoM and other development partners to undertake the remaining 4. All basin plans were based on the same TORs and followed similar template and content. Objective/Outcome: Increase the yield of the Corumana Dam to augment water supply for the GMMA Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Incremental reservoir yield Text 0.00 16.8 11.60 11.60 available from Corumana Dam for GMMA 15-Sep-2011 15-Sep-2011 12-Nov-2019 15-Feb-2021 Comments (achievements against targets): The outcome was achieved. Major works were completed and all spillway gates were tested and found to be safe and operational. Handover certificate has been issued, triggering a two-year defects notification period. The increased storage capacity of the reservoir results in an increased yield for both irrigation and water supply. The latter increases from a baseline of 69Mm3/yr to 81 Mm3/yr. Once the Bank-financed GWSEP is completed in April 2021, water from the dam can be abstracted, treated, and distributed to 560,000 residents in the GMMA. Page 40 The World Bank Water Resources Development (P107350) A.2 Intermediate Results Indicators Component: Component A: Water Resources Management Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion (AF/PPCR) Hydro-met Number 2.00 16.00 16.00 stations in Limpopo River basin upgraded to monitor 27-Sep-2013 27-Sep-2013 30-Jun-2020 extreme events Comments (achievements against targets): Indicator and target not relevant. The activities related to improve the hydro-meteorological system was transferred to a separate project and therefore, while the target was achieved, the results cannot be attributed to this project. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of agreements for Number 0.00 1.00 2.00 international waterways reviewed and updated 15-Sep-2011 15-Sep-2011 30-Jun-2020 Comments (achievements against targets): The target was exceeded. The target was 2 agreements to be reviewed, of which 1 agreement would also be updated. The actual achieved was 2 agreements reviewed and updated. Page 41 The World Bank Water Resources Development (P107350) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of other water Number 1.00 3.00 5.00 service providers that the project is supporting 15-Sep-2011 15-Sep-2011 30-Jun-2011 Comments (achievements against targets): The target was exceeded. The original plan was to provide assistance to ARA-Sul, ARA-Zambeze, and ARA-Centro Norte. During implementation, assistance was extended to ARA-Centro and ARA-Norte. Assistance included training, provision of equipment to control water quality, vehicles, cell phones, gauge readers, and computer equipment. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion (AF) Flood management Text Lack of studies Studies inform Studies inform studies integrating developed and applied decision making decision making infrastructure planning applied in government 27-Sep-2013 27-Sep-2013 30-Jun-2020 decision making Comments (achievements against targets): Target achieved. The study was carried out in a participatory manner and recommendations have been converted to a concrete action plan for local planning authorities. Indicator Name Unit of Measure Baseline Original Target Formally Revised Actual Achieved at Page 42 The World Bank Water Resources Development (P107350) Target Completion (AF) Integrated Flood Text Lack of integrated Operating effectively Operating effectively Management Decision flood management in wet season in wet season Support System operating for decision support Limpopo River basin system 27-Sep-2013 27-Sep-2013 30-Jun-2020 Comments (achievements against targets): Target achieved. The Integrated Flood Management Decision Support System has been completed, is operating effectively, and is informing decision making. Component: Component B: Completion of the Corumana Dam: Civil and Hydro-Mechanical Works Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Storage Capacity Corumana Cubic 720.00 1240.00 960.00 960.00 Dam Meter(m3) 15-Sep-2011 15-Sep-2011 12-Nov-2019 30-Jun-2020 Comments (achievements against targets): Target achieved. With the spillway gates installed and other major works completed, the Corumana Dam can now store 960 million cubic meters of water. Indicator Name Unit of Measure Baseline Original Target Formally Revised Actual Achieved at Page 43 The World Bank Water Resources Development (P107350) Target Completion Improved safety monitoring Yes/No No Y Yes system for Corumana Dam functioning 15-Sep-2011 15-Sep-2011 30-Jun-2020 Comments (achievements against targets): Target achieved. Dam monitoring system has been installed and is functioning. The pressure relief wells have been built, the drains are protected, and the v-notch outlet structures are being completed to measure the seepage from the toe drain. Devices for monitoring and measuring crack progression that may exist in the concrete structures have also been installed. The system includes a telemetric connection to five hydrometric stations to be processed by the SCADA system housed in the Incomati River Basin Management Unit. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Water license for Maputo Yes/No No Y Yes Water Supply granted 15-Sep-2011 15-Sep-2011 30-Jun-2020 Comments (achievements against targets): Target achieved. The Maputo water license was granted through a Memorandum of Understanding between ARA-Sul and FIPAG (the water utility) in 2015. Component: Component C: Completion of the Corumana Dam: Environmental and Social Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Page 44 The World Bank Water Resources Development (P107350) OLD Version of Indicator: Number 0.00 435.00 1,095.00 People in project areas with access to Improved Water 15-Sep-2011 15-Sep-2011 30-Jun-2020 Sources Comments (achievements against targets): Target exceeded. The integrated water supply system provided safe water to 1,095 people, including 660 people resettled into new houses, plus members of the host communities in Ndindiza, Baptine, Fungotine and Mutxipe, as well as the surrounding villages of Chavane, Lingongolo, Mahungo, Jone, Mabane. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion People provided with Number 0.00 435.00 660.00 improved sanitation 15-Sep-2011 15-Sep-2011 30-Jun-2020 Comments (achievements against targets): Target exceeded. The project benefited 660 resettled people with internal sanitary facilities and improved latrines in their new homes. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Environmental management Yes/No No Y Yes plan audited annually 15-Sep-2011 15-Sep-2011 30-Jun-2020 Page 45 The World Bank Water Resources Development (P107350) Comments (achievements against targets): Target achieved. The Environmental Management Plan was audited annually. A final audit report will be undertaken at the end of the defects liability period. Component: Component D: Water Resources Development Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Direct project beneficiaries Number 0.00 1452000.00 287,100.00 300,271.00 from interventions flood management interventions 27-Sep-2013 27-Sep-2013 15-Dec-2017 30-Jun-2020 Female beneficiaries Percentage 0.00 813267.00 143,050.00 180,162.00 09-Aug-2018 Comments (achievements against targets): Revised target achieved. This indicator was introduced during the Sept. 2013 Additional Financing, with the original target of 1.45 million people, representing the population of the entire Gaza Province (taking into account expected population growth rate of 2%). The target was reduced to 287,100 in Dec. 2017 to reflect the actual areas of the Province benefiting from flood management interventions. Actual achievement reflects current population numbers in those areas. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Page 46 The World Bank Water Resources Development (P107350) Investment studies for Number 0.00 5.00 4.00 3.00 additional dam projects completed 15-Sep-2011 15-Sep-2011 15-Dec-2017 30-Jun-2020 Comments (achievements against targets): The target was partially achieved. The original target of 5 was reduced to 4 in 2017 with the assumption that an AfDB-financed study (which was done in coordination with, but outside the framework of the project) would count toward the target. The project only financed 3 feasibility studies for the Gorongosa, Metuchira, and Ressano Garcia dams. The AfDB financed the Moamba Major dam. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion (AF) Dykes and levees Percentage 0.00 75.00 40.00 rehabilitated to withstand a 1:50 year flooding in the 27-Sep-2013 27-Sep-2013 30-Jun-2020 Limpopo River basin Comments (achievements against targets): Target partially achieved. The project financed the rehabilitation of dykes and levees back to their original capacity (1:20 year flood), not at greater capacity (1:50 year flood). Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion (AF) Macarretane barrage on Yes/No No Y Yes Page 47 The World Bank Water Resources Development (P107350) the Limpopo River has 27-Sep-2013 27-Sep-2013 30-Jun-2020 functioning hydro- mechanical works, gate control system, and reinforced foundations Comments (achievements against targets): Target achieved. The project financed the hydro-mechanical works, gate control system and the barrage is functional. A separate Bank project financed the reinforcement of the barrage's foundations. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion National water storage Cubic 77,474.00 78594.00 77,714.00 77,714.00 capacity Meter(m3) 15-Sep-2011 15-Sep-2011 12-Nov-2019 30-Jun-2020 Comments (achievements against targets): Target achieved. With the spillway gates installed at the Corumana Dam, the reservoir capacity has increased to 77,714 million cubic meters, which is the revised target for this indicator. Page 48 The World Bank Water Resources Development (P107350) B. KEY OUTPUTS BY COMPONENT Objective/Outcome 1. Strengthen the management of national water resources Outcome Indicators 1. Percent of hydro-climatic stations operational and reporting 1. Number of agreements for international waterways reviewed and updated 2. Number of other water service providers that the project is supporting 3. Integrated Flood Management Decision Support System operating Intermediate Results Indicators for Limpopo River basin 4. Flood management studies integrating infrastructure planning applied in government decision making 5. Hydro-met stations in Limpopo River basin upgraded to monitor extreme events 1. Training on international water law, negotiations, mediation, water conflict management 2. National Strategic Action Plan to Monitor the Compliance of Key Outputs by Component International Shared Watercourse Agreements (linked to the achievement of the Objective/Outcome 1) 3. Two new irrigation canals in the Limpopo Valley 4. Integrated flood management plan for the Limpopo 5. Training, capacity building, institutional strengthening activities Objective/Outcome 2. Strengthen the development of national water resources Outcome Indicators 1. River basins covered by water resources management plans 1. Investment studies for additional dam projects completed Intermediate Results Indicators 2. Direct project beneficiaries from interventions flood management interventions Page 49 The World Bank Water Resources Development (P107350) 3. Dykes and levees rehabilitated to withstand a 1:50 year flooding in the Limpopo River basin 4. Macarretane barrage on the Limpopo River has functioning hydro- mechanical works, gate control system, and reinforced foundations 5. National water storage capacity 1. Technical designs for three small dams and associated water supply networks and distribution systems Key Outputs by Component 2. Environmental and social impact assessments (linked to the achievement of the Objective/Outcome 2) 3. River basin plans for the Limpopo, Zambezi, and Lurio river basins 4. Rehabilitated dykes, levees 5. Rehabilitated Macarretane barrage Objective/Outcome 3. Increase the yield of the Corumana Dam to augment water supply for the GMMA 1. Incremental reservoir yield available from Corumana Dam for Outcome Indicators GMMA 1. Storage Capacity Corumana Dam 2. Improved safety monitoring system for Corumana Dam functioning 3. People in project areas with access to Improved Water Sources Intermediate Results Indicators 4. People provided with improved sanitation 5. Environmental management plan audited annually 6. Water license for Maputo Water Supply granted 7. Direct beneficiaries of the Environmental and Social Program 1. Completion of the Corumana Dam with new spillway gates, saddle dam, repairs to concrete pillars, installation of hydro-mechanical Key Outputs by Component and control equipment, control room (linked to the achievement of the Objective/Outcome 3) 2. New dam safety monitoring system, including pressure relief wells, v-notch outlet structures, toe drain 3. New O&M manual Page 50 The World Bank Water Resources Development (P107350) 4. New Emergency Preparedness Plan 5. Consultations with local authorities, basin management committees, disaster management personnel, first responders 6. Training, capacity building, institutional development activities Page 51 The World Bank Water Resources Development (P107350) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION TEAM A. TASK TEAM MEMBERS Name Role Preparation Marcus Wishart Task Team Leader Alberto Niñio Lead Counsel John Bryant Collier Operations Officer Jose Janeiro Senior Finance Officer Rafael Saute Communications Officer Satoru Ueda Lead Water Resources Specialist Supervision/ICR Pierre Francois-Xavier Boulenger Task Team Leader Antonio Laquene Chamuco, Amos Martinho Malate Procurement Specialist(s) Joao Tinga Financial Management Specialist Elvis Teodoro Bernado Langa Financial Management Specialist Sheila Lucrecia Francisco Pene Neves Financial Management Specialist Maria Isabel Nhassengo-Massingue Procurement Team Minerva S. Espinosa-Apurada Team Member Salma Omar Social Specialist Odete Duarte Muximpua Procurement Team Arlete Quiteria Comissario Nkamate Procurement Team Teofilo Domingos Munjovo Procurement Team Celia A Dos Santos Faias Procurement Team Paulo Jorge Temba Sithoe Environmental Specialist Eden Gabriel Vieira Dava Team Member Mark Doveton Wood Environmental Specialist Felipe Vicente Lazaro Team Member Page 52 The World Bank Water Resources Development (P107350) B. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY08 16.506 105,823.73 FY09 51.131 335,441.96 FY10 19.335 88,379.35 FY11 52.515 327,057.12 FY12 8.700 64,665.10 Total 148.19 921,367.26 Supervision/ICR FY12 33.850 214,541.00 FY13 33.646 186,422.81 FY14 9.120 41,631.60 FY15 8.658 59,757.78 FY16 19.516 126,666.25 FY17 20.414 153,407.72 FY18 24.951 205,365.84 FY19 30.499 192,860.24 FY20 17.083 139,341.19 Total 197.74 1,319,994.43 Page 53 The World Bank Water Resources Development (P107350) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Percentage of Approval Components (US$M) Closing (US$M) (US$M) COMPONENT A – WATER RESOURCES 15.0 16.15 107.6% MANAGEMENT COMPONENT B – CORUMANA DAM 42.0 39.21 93.3% COMPLETION COMPONENT C – CORUMANA 5.5 4.02 73.1% ENVIRONMENTAL AND SOCIAL COMPONENT D – WATER RESOURCES 35.0 27.53 78.6% DEVELOPMENT CONTINGENCY 4.5 Total 102.0 86.91 85.2% Page 54 The World Bank Water Resources Development (P107350) ANNEX 4. EFFICIENCY ANALYSIS 1. The objective of the Project’s intervention on Corumana Dam was to increase the yield of the Corumana Dam to augment water supply for the Greater Maputo Metropolitan Area (GMMA). An economic analysis was undertaken at the time of the ICR to evaluate the economic viability of the Project and compare the actual results against foreseen results at appraisal. 2. While the Project is designed to augment the bulk water supply for GMMA, realization of actual economic benefit (i.e. increased water supply for individual households in GMMA) cannot be achieved without completing associated infrastructures such as the transmission pipeline from the dam site to urban areas, WTP and distribution pipe network, all of which are implemented under Greater Maputo Water Supply Expansion Project (P125120). The economic analysis considers costs and benefits arising from associated components of both projects.76 3. At appraisal stage, there were a number of project design options, e.g. location of intake facility (intake at Corumana dam or intake at Incomáti/Sabie confluence) and water transmission system (gravity-fed or pumping system) that had not been determined yet, therefore the appraisal-stage economic analysis included the cases of each possible option. During project implementation, the PAMT adopted the option of an intake at Corumana Dam, bulk water transmission by gravity and incremental yield of 16.8 million m3/yr, therefore only the case of this particular option was compared in the current analysis. 4. Major changes since the appraisal include the delay in implementation. The original plan was to complete substantial civil works by 2016 and gain project benefits afterwards, but in the actual implementation, the major civil works contract started in 2017 and finished in 2020. Another major change was the cancellation of the fuse plug at the saddle dam and the addition of an auxiliary spillway.77 This design change requires the Corumana dam reservoir to be operated at full supply level (FSL) of 114 meters which is below the original FSL of 117m, until the auxiliary spillway becomes fully functional. As a result, the incremental water yield, which had originally been designed as 16.8 Mm3/yr at 117m FSL, was downscaled to 11.6 Mm3/yr at 114m FSL. According to the design study the construction of the auxiliary spillway would cost US$40 million and would take almost two years. To maximize the incremental water yield even during the construction period of the auxiliary spillway, the GoM is considering a plan to raise the FSL to 117m only during dry season when the probability of flood disaster is low. 5. Costs and benefits were converted to 2011 prices in order to compare them against the projection at appraisal. Taxes and duties included in financial costs of the project were deducted to convert to economic costs. Standard conversion factor, an inverse number of the shadow exchange rate factor, was also applied to convert the financial value 76 The design capacity of Greater Maputo Water Supply Expansion Project (60,000 m3/day) exceeds that of this project (16.8 Mm3/yr = 46,027 m3/day). In order to avoid overcounting the investment cost under Greater Maputo Water Supply Expansion Project, the ‘six-tenths-factor rule’ was applied to estimate the portion of cost corresponding to the achievement of the objective of this project. The six-tenths-factor rule is an empirical rule where if the cost of a given unit at one capacity is known, the cost of a similar unit with X times the capacity of the first is approximately (X)0.6 times the cost of the initial unit. 77 The original design of the project included the installation of six spillway gates and a saddle dam with an earth fuse plug to act as an emergency spillway. However, during project implementation, it became clear that the fuse-plug on the saddle dam could no longer play the expected role as the downstream area of this saddle dam, free of people during project preparation, was inhabited by hundreds of people who would be in harm if the fuse-plug were activated. Based on advice from the Panel of Experts (PoE), it was decided that instead of equipping the saddle dam with the fuse plug element, an auxiliary spillway will be constructed in the left embankment of the dam. Page 55 The World Bank Water Resources Development (P107350) to economic value. While the discount rate of 12 percent was used at appraisal, this analysis adopted 6 percent following the guidance issued by the Bank in 2016.78 Other assumptions, including a 45-year project life period, operation and maintenance (O&M) cost at 0.3 percent per year against capital expenditure, and additional maintenance fee that occur every 15 years, followed those of appraisal stage for comparison. 6. To transform costs and benefits expressed in nominal prices to 2011 prices, this analysis applied the annual inflation rates published by Mozambique National Institute of Statistics (INE). The costs were also converted to Mozambique Metical (MZN) to assess the impact of currency depreciation of MZN during the project period. Table 1. Comparison Between Investment Costs at Appraisal and Actual Costs Actual PAD Investment Nominal Prices 2011 Prices Mil. US$ Mil. MZN Mil. US$ Mil. MZN Mil. US$ Mil. MZN Corumana Dam heightening Civil work 40 1,162 37 2,315 23 1,475 Detailed design and 2 58 5 244 3 156 construction supervision Greater Maputo Water Supply Expansion Project79 Water treatment plant 58 1,698 34 2,154 22 1,372 Transmission 74 2,168 77 3,066 67 2,696 Distribution network 9 261 9 563 6 358 Subtotal without auxiliary 192 5,608 171 8,905 127 6,415 spillway Auxiliary spillway (planned) n/a n/a 40 2,543 23 1,418 TOTAL n/a n/a 211 11,448 150 7,833 7. As Table 1 shows, the cost for the main civil work, heightening of Corumana dam, in nominal dollar terms (US$37 million) were lower than the estimate at appraisal (US$40 million). When inflation is taken out, the price (US$23 million) is 42 percent lower than the estimate at appraisal. Similar cost decrease is seen with other investment packages except the supervision consultant work for Corumana dam, which underwent major contract amendment to design the aforementioned auxiliary spillway. 8. This cost decrease in dollar term is attributed to the devaluation of the local currency, Mozambique Metical (MZN). The exchange rate of MZN to US$ dropped significantly from 0.034 in 2011 to 0.016 in 2016, and it has remained at a similar standard until today. The sharp currency devaluation in 2016 was followed by high inflation rate (19.9 and 15.1 percent in 2016 and 2017 respectively), yet it did not fully absorb the impact of 50+ percent currency depreciation. Thus, 78The World Bank (2016) Discounting Costs and Benefits in Economic Analysis of World Bank Projects 79The costs under Greater Maputo Water Supply Expansion Project shown in this table are the full contract amount to achieve the project target of 60,000 m3/day supply. In the economic analysis, these values were deducted using the six-tenths-factor rule to estimate the cost portion that matches the attainment of the objective of this project (46,027 m3/day). Page 56 The World Bank Water Resources Development (P107350) the SDR-based fixed loan amount resulted in gain of value in terms of local currency even after taking out the inflation factor. 9. Benefits accrued through Corumana dam heightening were estimated using the same approach as at appraisal, that is: i) household benefits by increased water supply (valued in the form of willingness to pay [W2P]); ii) reduced flood damage; iii) increased hydropower generation and iv) increased catch of local fisheries. The measure of W2P is assumed to consider the positive health benefits accrued from improved water supply, hence no separate valuation of health benefits is considered either at appraisal or ICR. The benefit from increased water supply, represented by the W2P, remained the majority (80+ percent) of the total benefit composition, followed by reduced flood damage (11 percent). At appraisal, the W2P used was US$1.13/m3 based on the average water tariff of MZN32.77/m3 adopted by the water utility of Maputo (ARA-Sul). ARA-Sul revised the tariff to MZN43.25/m3 in 2018. When currency devaluation in 2016 is considered, the nominal tariff in foreign currency is US$0.69/m3, a significant drop from the value at appraisal. However, it is unlikely that people’s willingness to pay for basic services like water supply fluctuates because of macroeconomic turbulence factors such as exchange rates. It should also be noted that water tariff is just a reference proxy of W2P rather than a perfect representation of W2P. Therefore, this analysis assumed that the W2P as percentage of household income has remained at the same level throughout the project period and that the W2P fluctuates proportionally to the per capita GDP at the time, instead of mechanically interpreting the value of water tariff as W2P. Considering the 5-year moving average GDP per-capita at appraisal and at closure were $527 and $497 respectively, the ICR-stage W2P was set as US$1.06/m3. 10. Results of the economic analysis is summarized in Table 2. The ICR-stage EIRR was lower than appraisal-stage EIRR, but all the scenarios exceeded the revised breakpoint of 6 percent, denoting the economic viability of the project. The high NPVs in contrast to the US$2.3 million NPV at appraisal result from the revised discount rate from 12 percent to 6 percent. The major factors of lowered EIRR are: (i) prolonged preparation and construction period, and (ii) increased cost due to the cancellation of fuse-plug function of the saddle dam and the need to construct auxiliary spillway instead. The latter is unforeseen design change that also caused delay in benefit yield, since the Corumana Dam reservoir can only be operated up to 114 mASL, instead of the original target of 117 mASL, until the completion of auxiliary spillway in 2022 or later. Table 2. Results of the economic evaluation NPV EIRR (Mil. US$) PAD 2.3 12.3% Actual Scenario 1* 24.0 9.9% Scenario 2** 26.4 10.4% Scenario 3*** 15.2 9.0% * Corumana dam operated up to 114 meters filling level until the completion of auxiliary spillway. After the completion of auxiliary spillway, the dam is operated up to 117 meters level. ** Dam operated up to 114 m level during rainy season but 117 m during dry season until the completion of auxiliary spillway. *** Auxiliary spillway is not constructed. Dam is operated up to 114 m level permanently. Page 57 The World Bank Water Resources Development (P107350) 11. The analysis set three different scenarios with respect to the operation of Corumana dam reservoir in the coming few years until the auxiliary spillway is completed. The first scenario is status quo, where dam reservoir is filled up to 114 mASL during the absence of the auxiliary spillway. The second one, which is under discussion by the GoM, is a hybrid model to raise the filling level up to 117 mASL only during dry seasons, considering the low flood risk in that period, and lower the level to 114 mASL during the rainy season. The third one is to cancel the construction of the auxiliary spillway: this helps reduce the cost of US$40 million, but the maximum FSL of Corumana Dam remains at 114 mASL permanently. As Table 2 shows, the most economically viable scenario is the second option, whereas the least viable scenario is the cancellation of the auxiliary spillway. Even if the FSL is kept during the construction of spillway (Scenario 1), the economic performance exceeds the no spillway scenario (Scenario 3). Therefore, construction of the auxiliary spillway as currently planned is recommended from long-term economic viability point of view. In parallel, adoption of a hybrid reservoir operation plan during the construction period of the auxiliary spillway is also recommended to maximize the benefit from the existing facility. 12. Sensitivity analysis was also conducted to evaluate the impact on the project caused by possible future changes. The first case assesses the impact of delay in the completion and associated cost increase of the auxiliary spillway, the technical designs of which are under discussion. While scenario 2 remains the most viable under this case, the gap between scenario 1 and 3 narrows down significantly. This result indicates the importance of timely completion of the auxiliary spillway to maximize the benefit gained through the spillway. Further delay and cost overrun of the auxiliary spillway may put scenario 3 as a more viable option than scenario 1. Marginal impact is seen in the second case which assumes significant increase in future O&M cost. This result indicates the magnitude of project benefit is significantly larger than the running cost. The third case was analyzed in order to estimate the impact of future economic growth, which will boost the project benefit. The increase of benefit by 20 percent almost doubles the NPV in each scenario, which denotes the economic significance of provision of sufficient water to urban areas in growing economies like Mozambique. Table 3. Results of sensitivity analysis Scenario 1* Scenario 2* Scenario 3* Future scenarios NPV NPV NPV EIRR EIRR EIRR (US$ M) (US$ M) (US$ M) Base scenario 24.0 9.9% 26.4 10.4% 15.2 9.0% Case1: Spillway completion delayed by 20.5 9.4% 25.1 10.3% 15.2 9.0% 2 years and cost increase by 20% Case 2: Maintenance cost increased by 23.4 9.8% 25.8 10.3% 14.7 8.9% 50% Case 3: Benefit increased by 20% 40.1 12.1% 43.0 12.7% 27.8 11.2% * See Table 2 for the description of each scenario. 13. With respect to the efficiency of project implementation, the notable element was the delay in the civil works both under this project and the Greater Maputo Water Supply Expansion Project. However, in terms of economic viability, the impact associated with these delays is marginal, since the sequential relations between capital investment and benefit yield does not change with the delay in launch of works unless the implementation period of the work is prolonged. However, given the fact prolonged project period increases the chance of macroeconomic and political risk during that period, on-time implementation is always advisable. Page 58 The World Bank Water Resources Development (P107350) ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS Page 59 The World Bank Water Resources Development (P107350) Page 60 The World Bank Water Resources Development (P107350) ANNEX 6. EVOLUTION OF INTERMEDIATE RESULTS INDICATORS Intermediate Indicator Changes Component A. Enhanced capacity for water resource management at the national and regional level • Number of agreements for international waterways reviewed and updated • Improved cost recovery of ARAs Dropped in 2017 as it was found not to be achievable within the project scope. Although the project was to support improvements on revenue collection and improvements in cost recovery, this indicator required substantially more support including legal and institutional reform, investments and capacity building for water licensing and monitoring systems. • Integrated Flood Management Decision Support System operating for New indicator introduced in 2013; revised in southern Mozambique (Limpopo and Incomáti River basins). 2017 to focus only on Limpopo River basin. • Flood management studies integrating infrastructure planning applied in New indicator introduced in 2013 government decision making • Hydro-met stations in Limpopo and Incomáti River basins upgraded to New indicator introduced in 2013; revised in monitor extreme events 2017 to focus only on the Limpopo River basin. The MTR of Dec. 2014 had proposed dropping this from the NWRD project, but it was never carried out. Component B. Corumana Dam storage capacity restored to design capacity • Storage Capacity Corumana Dam Target was reduced from 1240Mm3 to 960 Mm3 to take into account design changes • Water license for Maputo Water Supply granted • Improved safety monitoring system for Corumana Dam functioning Component C. Social and environmental safeguards instruments complied with and programs implemented • Direct project beneficiaries of the Environmental and Social Program • People provided with access to improved water sources under the project • People provided with improved sanitation • Environmental management plan audited annually Component D. Strengthened capacity of ARAs for water resources development • Investment studies for additional dam projects completed • National water storage capacity The MTR of Dec. 2014 had proposed dropping this from the NWRD project, but it was never carried out. It’s was described as “redundant given the other measures of storage associated with Corumana Dam.” The AM proposed to have it “removed to safeguard sufficient funds for prioritized activities.” • Other water service providers that the project is supporting • Direct project beneficiaries from flood management interventions New indicator introduced in 2013; revised in 2017. The indicator was calculated based on the Gaza province population, while the Page 61 The World Bank Water Resources Development (P107350) project activities only cover specific locations along the Limpopo River. The indicator was therefore revised. In addition, the indicator was revised to specify linkage to the AF, since the initial credit also had an indicator on direct beneficiaries related to the resettlement of PAPs from Corumana dam. • Dykes and levees rehabilitated to withstand 1:50 year flooding in the Limpopo New indicator introduced in 2013; revised in and Incomáti River Basins 2017 due to capacity limitations, the flood management activities are focused on the Limpopo River basin which is the most critical river in terms of flood management. • Macarretane Barrage on the Limpopo River has functioning hydro-mechanical New indicator introduced in 2013 works, gate control system, and reinforced foundations Page 62 The World Bank Water Resources Development (P107350) ANNEX 7. STATUS OF IMPLEMENTATION OF ENVIRONMENTAL AND SOCIAL SAFEGUARDS 1. The project was classified as an Environmental Assessment (EA) Category A, requiring full assessment. The EA reflected the expected impacts of large-scale civil and hydro-mechanical works to increase the capacity of the Corumana Dam. Consequently, the following ESS policies were triggered: OP/BP 4.01 Environmental Assessment, OP/BP 4.04 Natural Habitats, OP/BP 4.11 Physical Cultural Resources, OP/BP 4.37 Safety of Dams, OP/BP 4.12 Involuntary Resettlement, and OP/BP 7.50 International Waterways. To comply with provisions under these policies the GoM prepared and applied an ESIA and RAP. The former included a suite of Environmental and Social Management Plans, including (i) a Construction and Workers Camp Management Plan; (ii) an Inundation Preparation Plan; (iii) a Reservoir Management Program; (iv) an Environmental Water Releases Program; (v) an Environmental Monitoring Plan. The RAP also included a community LRP. 2. Given the nature, magnitude and sensitivity of the proposed investments the project included a standalone Component C to ensure preparation of and compliance with Bank safeguards requirements as well as applicable national regulations. An Environmental and Social Service Provider (ESSP) was hired to review and update the original ESIA and RAP (completed in 2011) to take into account new, more demanding GoM regulations on resettlement and economic impact assessments. As discussed in Sections III.B.I and III.B.2, a number of challenges led to severe delays in finalizing and implementing the safeguards instruments. Project documents systematically reported safeguards performance, and during the four years it took to prepare the updated ESIA and RAP, performance was recurrently rated as MU (Moderately Unsatisfactory). 3. The 30-month cumulative extension of the closing date allowed good progress in implementing the EMP and RAP, despite the challenges discussed in Section III. The construction works has caused limited and largely temporary adverse environmental impacts, with deviations from the ESMP mainly concerning housekeeping and waste management, and initially poor OHS performance, which was corrected in later months through penalties, training, stringent monitoring and ongoing corrective action. One fatality occurred during construction, caused by a preventable vehicle accident, the root cause of which was determined to be mainly a lack of senior management commitment in the Contractor’s organization and to some extent to the inadequate borrower capacity to manage safeguards measures. Consequently, the Bank increased support to the project by implementing additional training on safeguards and recommended that the borrower commission an independent Environmental Audit to assess legacy issues, identify and report on any outstanding environmental non-compliances and propose corrective measures to be implemented simultaneously with the remaining project activities. The outcome of the Audit report concluded that, in general, the Contractor’s environmental management was partially compliant with the overall project requirements, whilst providing some remedial actions necessary to address housekeeping and improvements in waste management. At the time of this ICR, major construction has effectively ended and all rehabilitation around the works is satisfactory completed. A final audit will be undertaken towards the end of the construction liability period. 4. Pertaining to OP 4.04 Natural Habitats, embedding environmental flow requirements and water quality monitoring into the operating management system of the dam has been a major achievement and training of the dam operators is scheduled for this year. Flow management includes requirements for base flow and freshet releases, and for maintaining the dam at levels below 117 masl except in cases of severe floods, in order to minimize backwater impacts in Sabie Game Reserve and Kruger National Park in South Africa. Until the emergency spillway is built, water levels will be maintained well below this level at 114 masl. The client also maintains communications with Sabie Game Park about Page 63 The World Bank Water Resources Development (P107350) poaching and has an agreement in place which allows Sabie Game Park rangers to track and apprehend poachers who use the dam to hunt illegally in the South African nature reserves. 5. Pertaining to OP 4.37 Safety of Dams, An Emergency Preparedness Plan has been prepared and widely communicated and reviewed. There remains some concern about the Borrower’s capacity to manage long term safeguard issues. Provision is being made through the extended contract of the supervising engineers to build the client’s capacity to manage safeguards during the operational phase of the dam. 6. On OP/BP 7.50 International Waterways, GoM formally notified the Republic of South Africa (RSA) and Kingdom of Swaziland (KoS) during project preparation of its intentions to proceed with the Corumana Dam in accordance with the Revised SADC Protocol and the provisions of the Interim IncoMaputo Agreement (IIMA). The GoM has complied with the request from RSA for "constant feedback and regular consultation with relevant bodies" through the Tripartite Permanent Technical Committee. 7. On OP/BP 4.12 Involuntary Resettlement, the ESIA included a separate volume for the social impact assessment that provides an overview of the socioeconomic issues of project affected communities and the RAP, which looks at aspects related to resettlement and community livelihoods restoration. The project financed the construction of high- quality resettlement housing – including access roads, integrated water system, and electricity – as well as actual resettlement of the PAPs. But as detailed in Section II, the project closed with pending safeguards measures, including some compensation and allowances, several LRP activities, and social and environmental audits. These activities are on- going, to be paid by GoM and supported by the RAP service provider with support from the safeguards specialist in the PoE as well as from the Bank. The project closed with overall environmental and social safeguards performance consistently rated MS during much of the safeguards implementation phase and at project closure. 8. The Grievance Mechanism is active in the Corumana Dam and adjacent areas and in the resettlement villages of Ndindiza, Fungotine, Baptine and Mutxipe as evidenced by existing documentation on activities carried out by the service provider responsible for implementing the resettlement, as well as in annual social audit reports. Communities in general are aware and use the mechanism. DNGRH and ARA-Sul made an effort to ensure that the cases of complaints and grievances presented are resolved fairly and in accordance with the legislation in force in Mozambique and the Bank's safeguards policies. Page 64 The World Bank Water Resources Development (P107350) ANNEX 8. GOM LETTER ON SAFEGUARDS BUDGET, HANDOVER CERTIFICATE Page 65 The World Bank Water Resources Development (P107350) Page 66 The World Bank Water Resources Development (P107350) Page 67 The World Bank Water Resources Development (P107350) Page 68 The World Bank Water Resources Development (P107350) ANNEX 9. SUPPORTING DOCUMENTS Project Appraisal Documents (PADs), Project Papers (PPs), Implementation Completion Reports (ICRs), Legal Agreements PAD for the National Water Resources Development Project (P107350, Report No. 63794-MZ), World Bank, August 29, 2011 PAD for the Greater Maputo Water Supply Expansion Project (P125120, Report No. 75919), World Bank, June 27, 2013 PAD for the Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049, Report No. 75939), World Bank, April 1, 2013 PAD for the Mozambique Emergency Resilient Recovery Project (P156559, Report No. PAD1553), World Bank, September 29, 2015 PAD for Agriculture and Natural Resources Landscapes Management Project (P149620, Report No. 3917), June 12, 2016 PP for the National Water Resources Development Additional Financing (P146098, Report No. PAD702), World Bank, September 12, 2013 PP for the National Water Resources Development Project (P107350, Report No. RES29775), World Bank, December 8, 2017 PP for the National Water Resources Development Project (P107350, Report No. RES38756), World Bank, November 12, 2019 ICR for the Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049, Report ICR5203), World Bank, June 15, 2020 ICR Review for the Climate Resilience: Transforming Hydro-Meteorological Service Project (P131049, ICRR0022203), IEG, World Bank, February 2, 2021 Financing Agreement for Credit 5011-MZ, World Bank, December 5, 2011. Financing Agreement Amending and Restating for Credits 5011-MZ and 5307-MZ, World Bank, November 11, 2013 Country Partnership Strategies (CPS), Other Strategy Documents CPS for the Republic of Mozambique (Report No. 39395-MZ), World Bank, April 24, 2007 CPS for The Republic of Mozambique, FY17-21 (Report No. 104733-MZ), World Bank, March 30, 2017 Performance and Learning Review of the CPS for Mozambique, FY17-21 (Report No. 144024-MZ), April 3, 2020 Africa's Future and the World Bank's Support to It: Africa Regional Strategy (59761), World Bank, February 2011 Making Water Work for Sustainable Growth and Poverty Reduction, Country Water Resources Assistance Strategy for Mozambique, 2008-11 (Report No. 56855), World Bank, August 2007 República de Moçambique, Conselho de Ministros, Plano de Acção para Redução da Pobreza (PARP) 2011-2014, May 3, 2011 República de Moçambique, Estratégia Nacional de Desenvolvimento 2015-2035, July 2014 República de Moçambique, Conselho de Ministros, Proposta do Programa Quinquenal do Governo: 2020-2024, Maputo, 2020 República de Moçambique, Ministério para a Coordenação da Acção Ambiental, Estratégia Nacional de Adaptação e Mitigação de Mudança Climáticas, 2013-2025 Republic of Mozambique, National Directorate of Water Resources Management, Development of Master Plan for Water Resources Management in Mozambique, February 2018 Page 69 The World Bank Water Resources Development (P107350) Other Documents Aide Memoires of the Implementation Support Missions in May 2012, April 2013, November 2013, December 2014, May 2014, April 2015, October 2015, November 2016, January 2017, June 2018, March 2018, July 2019, February 2020 10th Report by the Panel of Experts for the Mission of November 25-December 7, 2020 Independent Project Completion and Evaluation for the National Water Resources Development Project, Ministry of Public Works, Housing and Water Resources, National Directorate for Water Resources Management, Government of Mozambique, June 2020 Bank Guidance, Implementation Completion and Results Report (ICR) for Investment Project Financing (IPF) Operations, World Bank, March 2, 2020 Draft Guidelines for Reviewing World Bank Implementation Completion and Results Reports, A Manual for IEG ICR Reviewers, IEG, World Bank, July 25, 2017 Page 70