Uganda Small Towns Water Supply and Rural Growth Centers Project. Project Completion Report TF UG TF UG TF UG

Transcription

1 Uganda Small Towns Water Supply and Rural Growth Centers Project Project Completion Report TF UG TF UG TF UG June 2015

2 2015 The Global Partnership on Output-Based Aid The World Bank 1818 H Street NW Washington DC Website: gpoba@worldbank.org All rights reserved.this report was produced the Global Partnership on Output-Based Aid (GPOBA). The findings, interpretations, and conclusions expressed herein do not necessarily reflect the views of GPOBA or the Board of Executive Directors of the World Bank or the governments they represent. Neither GPOBA nor the World Bank guarantees the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of GPOBA or the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Note: All currency amounts are in U.S. dollars. 2

3 Table of Contents A. Basic Information... 6 B. Key Dates... 6 C. Results Framework Analysis... 7 D. Disbursement Profile PROJECT CONTEXT, DEVELOPMENT OBJECTIVE AND DESIGN Context at Appraisal Project Development Objectives and Key Indicators Revised PDO (as approved by original approving authority) and Key Indicators, and Justification Project Design Main Beneficiaries Project Components Revised Components Other Significant Changes KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOME Project Preparation Implementation Project Outcome Monitoring and Evaluation Safeguard and Fiduciary Compliance Post-Completion Operation/Next Phase ASSESSMENT OF OUTCOMES Relevance of Objectives, Design, and Implementation Achievement of Project Development Objectives Efficiency Overall Outcome...30 The project was satisfactory based on the level of achievement of the PDO and the relevance of the project ASSESSMENT OF RISK TO DEVELOPMENT ASSESSMENT OF BANK AND BORROWER PERFORMANCE Bank Performance

4 5.2 Borrower Performance LESSONS LEARNED...32 APPENDIX A. PROJECT COSTS AND FINANCING...35 APPENDIX B. OUTPUTS BY COMPONENT...37 APPENDIX C: ECONOMIC AND FINANCIAL ANALYSIS...39 APPENDIX D: BRIEF ON IFC WORK...42 APPENDIX E: GENERAL RECOMMENDATIONS/WAY FORWARD

5 ACRONYMS AND ABBREVIATIONS CAS Country Assistance Strategy COWI/AS Consulting Engineers and Planners AS DWD Directorate of Water Development GPOBA Global Partnership for Output-Based Aid ICT Information, Communication and Technology IFC International Finance Corporation MoFPED The Ministry of Finance, Planning and Economic Development FM financial management MWE Ministry of Water and Environment NEMA National Environment Management Authority NPV net present value OBA NWSC output-based aid National Water and Sewerage Corporation OVC output verification certificate OVR output verification report PAD Project Appraisal Document PDO Project Development Objective PPP public-private partnership PwC PricewaterhouseCoopers PO Private Operators PWP public water point RGC rural growth center TA Technical Assistance WA Water Authority 5

6 A. Basic Information Country: Uganda Project Name Uganda Small Towns Water Supply and Rural Growth Centers Project Project ID: P L/C/TF Number(s): TF UG TF UG TF UG Completion Report Date: 30 August 2013 Type: GPOBA 1 Subsidy + Technical Assistance Lending Instrument: Grantee: Government of Uganda Original Total US$3,207,000 Disbursed Amount: US$2,973, Commitment Revised Amount US$3,207,000 Environmental Category: B Implementing Agency: Ministry of Water and Environment (MWE)/ Directorate of Water Development (DWD) Government Conditional grant to water authorities (WA) US$0.08 million), Operators investment (US$0.54 million) Water users connection fees (US$0.64 million) B. Key Dates Process Date Process Original Date Revised/Actual date Concept Review June 15, 2006 Effectiveness April Appraisal August 28, 2006 Restructuring Feb. 28, 2010 Dec. 31, 2011 Approval February 7, 2007 Mid-term review Not done Closing Feb. 28, 2010 June 30, 2012 Note: The project was set to end February 28, 2010 but was extended first to December 31, 2011 and then to June 30, The extension totaled 28 months, representing a 77 percent increase in project duration. Initial delays in the bidding process, and time required for environmental screenings and written approvals caused the first extension. The second extension resulted from complications with the land acquisition for Sipi rural growth center 1 Global Partnership on Output-Based Aid 6

7 (RGC) and the completion of works in Busembatia, a small town added to the project after appraisal. C. Results Framework Analysis Project Development Objectives (PDO) 2 The project aimed to increase piped-water access for Ugandans living in selected small towns and rural growth centers (RGCs) in a sustainable manner through increased participation of local private operators. Revised Project Development Objectives (as approved by original approving authority) The PDO was not revised. 2 Project Development Objectives as written in the Project Appraisal Document 7

8 PDO Indicators Indicator Baseline Value Original Target value (from approval documents) Indicator 1: Value quantitative or qualitative Formally Revised target value Actual Value achieved at Completion or Target Years Number of increased working connections (yard taps and public water points (PWP) or stand posts) 0 1,993 2,416 connections connections No formal revision made to number of connections Date achieved August June 2012 Comments (incl. percent achievement) The project exceeded its target by 21 percent, instituting 1,671 connections in small towns and 745 in RGCs; connection types comprised 2377 yard taps and 39 PWP Indicator 2: Value quantitative or qualitative Number of people provided with access to improved water supply under the project 0 45,000 persons No formal 54,486 persons served revision made to served number of persons served Date achieved August 2006 June 2012 Comments The target was exceeded by 21 percent, affecting an additional 9,422 persons 8

9 D. Disbursement Profile At the beginning of the project, disbursements were adversely affected by delays in tendering. Since this was the first time that a competitive output-based aid (OBA) tendering had been undertaken in the water sector in Sub-Saharan Africa, actors had no precedents on which to base their decisions. 1. PROJECT CONTEXT, DEVELOPMENT OBJECTIVE AND DESIGN 1.1 Context at Appraisal In 2002, just 56 percent of Uganda s population had access to a safe and reliable water source. 3 Access was especially limited in small towns and rural growth centers (RGCs). By the time of this project s appraisal (August 2006), Uganda s government already had been implementing reforms in the sector to improve access to safe water. The outputbased aid (OBA) pilot involving private sector participation was welcomed as part of the reform agenda, with the caveat that participants adhere to the tariff caps set by the government in June The government aimed to provide access to improved drinking water to 70 percent of all urban residents by 2005, reaching 100 percent by 2015; and to provide access to 77 percent of rural residents by An estimated US$410 million investment was required to meet this goal. Government-imposed budget ceilings pertaining to macro- 3 World Development Indicators: Little Data Book 2005 World Bank. 9

10 economic targets limited annual allocations to the water sector, and constrained the conditions under which the goal was to be achieved. By 2008, improved water coverage extended to about 46 percent of the population in 160 small towns, with systems functioning an average of 89 percent of the time. Private water operators began supplying water in small towns in 2001, enhancing the management of water supply schemes. A significant number of private operators had efficient operations with unaccounted for water levels averaging 22 percent and collection efficiencies of 86 percent. A study of 10 years of public-private partnership (PPP) in Uganda s small towns shows that the number of towns under the private sector participation (PSP_ model rose steadily from only 15 in to over 90 by , covering a population of over 1.5 million. The PSP approach achieved the following results: Water Connections expanded from 4,883 in 2002 to over 13,000 in Tariff collection rose from almost zero to approximately US$2 million (U Sh 5 billion), collected across the 88 towns reporting in Collection efficiency rose by more than 15 percent on average within three years of the introduction of private operators. Private Operators (POs) successfully collected nearly 90 percent of issued bills in , as the metering ratio the key element of a rational payment based system rose to over 90 percent. Despite these gains, Ugandans still spent much time collecting water from unprotected springs and streams, which resulted in disease and even death. There was a desire to develop a more sophisticated PPP arrangement. An output-based aid (OBA) pilot project was conceived to leverage private sector finance and expertise in system design, construction, and operation within the existing institutional framework. In November 2005, the Ministry of Water and Environment (MWE) approached the Bank s Global Partnership on Output-Based Aid (GPOBA) for help in designing and funding a pilot OBA project in small towns and rural growth centers to provide affordable safe water to lower income customers. The project was expected to deliver 2,000 connections (yard taps and public kiosks) to almost 45,000 people. The cost was covered by a US$3.2 million grant from GPOBA, co-financing of about 20 percent from user contributions, and, in some towns, a small number of conditional grants made by the central government to local water authorities. The small towns included Wakiso, Wobulenzi, Luwero, Kalisizo, Rukungiri, Kachumbala and Busembatia. RGCs included Sipi, Magale, Masafu, and Namutumba. 10

11 1.2 Project Development Objectives and Key Indicators The key project development objectives were to increase piped-water access for Ugandans living in selected small towns and RGCs in a sustainable manner through increased participation of local private operators. Indicators of these objectives included: Local populations would spend less time collecting water at distant water points, and therefore potentially spending more resources on other productive activities, including attending school. Traditionally, most water carriers are women and young girls; There was expected to be a reduction of disease and death related to waterborne illness. Localized and cost-effective solutions for system design, including greater coordination between system design, maintenance and operation, would be implemented. Existing water supply systems would demonstrate improved financial viability, including availability of alternative sources of finance for the sector. The use of funds in the water sector would be transparent. Key Performance Indicators The overall development objective was to be measured by the number of increased working connections (the number of yard taps and public water points (PWP)) of a specified quality in the respective small towns and RGCs. Success of the competitive tendering process was to be measured by the resulting lowest subsidy (for the small towns) or tariffs (for the RGCs) required to deliver outputs and manage systems. 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and Justification The PDO and key performance indicators were not revised during project implementation. 11

12 1.4 Project Design The OBA pilot for small towns and RGCs water supply set the following goals: PPPs would design, build and operate (DBO) contracts for RGCs based largely on the existing institutional framework. Operators contracts would include tariffs with simple escalation clauses intended to cover at least 10 percent of expected investment costs in RGCs and up to 30 percent in some small towns. In small towns, where private operators were extending existing systems, operators were compensated for investments after targeted connections had been made and independently verified. Ten percent of the payment was withheld until bills paid demonstrated a six-month period of water delivery. In RGCs, operators received compensation in phases for intermediate outputs, with 45 percent withheld until after verification of connections and demonstration of six months of water delivery. The private operators were contracted through a competitive process and had to pre-finance delivery of outputs. Private operators bore the performance risks. Private operators were expected to direct, maintain and manage the water supply systems and ensure a safe and reliable supply to all connections. Funding that exceeded the off government budget was channeled through a private fiduciary agent, PricewaterhouseCoopers (PwC), which had been selected through a competitive bidding process. PwC was to disburse funds to the operators based on verified and certified outputs, and report to GPOBA. The Ministry of Finance, Planning and Economic Development approved this arrangement in a letter dated August 10, The MWE and the Directorate of Water Development (DWD) were jointly responsible for the performance of water authorities and procurement and monitoring during the project s implementation. PwC signed an agreement delegating implementation to MWE/DWD. An independent verification and monitoring agent/local consultant checked the quality and quantity of outputs and reports to the MWE and PwC. Subsidies were to be one-off investment subsidies, and tariffs covered operation and maintenance (O&M) costs plus a margin. 4 Contracts were set at five years for small towns, and seven to 10 years for RGCs. 4 The resulting tariff will also demonstrate required cost recovery levels in small towns and RGCs to the new Minister of Water, who has expressed openness to tariff reform. 12

13 Private operators were expected to extend networks and provide a specific number of connections (yard taps and public water points) in small towns. Operators were to expand the production and/or storage capacity where required. All connections were to be metered. In RGCs, private operators were expected to construct an entire water supply system, including drilling/searching for appropriate water sources, production, storage, conveyance, and final connections. All connections were to be metered. 1.5 Main Beneficiaries The major beneficiary of the grant was the Government of Uganda. Other beneficiaries included private operators: Kol Kagulu Joint Venture, Trandint Limited, WSS Services Uganda Limited and Jobatov Joint Venture. Direct beneficiaries comprised the 2,377 lower income households in small towns and RGCs. A total of 54,486 persons were served. Original components 1.6 Project Components The project components at appraisal were as follows: Component 1: Installation of water service connections for selected households Component 2: Bank/GPOBA Supervision Component 3: (i) Operating costs for fiduciary agent; (ii) Consulting services for technical evaluation and contract implementation support; (iii) Consulting services for monitoring and verification of outputs; and (iv) Auditing services The grant agreements split the project in two parts, explained below: Part A: Pilot Subproject in Small Towns and Rural Growth Centers Part A consisted of improving water connectivity in six small towns and four rural growth centers, resulting in the following outputs: In small towns, water connectivity was to be achieved via: (i) connection of functional new yard taps; (ii) establishment of new public water points; and (iii) functionality of connections and stand posts after a period of 12 months, later revised to six months. In rural growth centers, water connectivity would be achieved through: (i) establishment of raw intake; (ii) installation of pump house and water supply facilities; (iii) completion of distribution networks; (iv) installation of storage capacity; (v) connection of 13

14 functional new yard taps; (vi) establishment of new public water points; and (vii) functionality of connections and stand pots after a period of six months. Part B: Project Advisory, Monitoring and Audit Services Part B involved strengthening the capacity for project management, including supporting the DWD in procurement, monitoring and evaluation function, through the provision of technical advisory services, and providing auditing services to the recipient. 1.7 Revised Components The components were not changed during the implementation of the project 1.8 Other Significant Changes Originally, 10 subprojects (six small towns and four RGCs) were targeted for implementation. One small town, Busembatia, was added to the pilot as a result of the savings made from the efficiency gains from the competitive tendering process. 2. KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOME 2.1 Project Preparation Project preparation and Design: The project gestation period took 15 months from submission of the request to the signing of the grant agreement. Milestones are listed below: The request was submitted by MWE/DWD to GPOBA in November The Project Information Document (PID) was prepared on June 15, The project appraisal was conducted in August The Ministry of Finance, Planning and Economic Development wrote to the World Bank (referred to as the Bank hereafter) endorsing the private sector as the fiduciary agent on August 12, MWE announced PricewaterhouseCoopers as the fiduciary agent on January 16, The grant agreement was signed between the Bank (GPOBA) and Price Water House Coopers Limited on February 12, Project preparation involved selecting towns and RGCs for implementation and determining the subsidy level for each. A total of 19 small towns and RGCs were identified by the MWE and submitted to GPOBA for consideration. GPOBA engaged a consultant, COWI A/S to assess the viability of the 19 towns and RCGs. COWI A/S identified eight towns and four RGCs as potential OBA candidates, however two of the 14

15 towns had running contracts and could not be included in the initial pilot. Selected towns, RGCs and original subsidy levels are listed in table 2.1, below. Table 2.1: Pilot Small Towns and RGCs and Estimated Subsidy Amounts Inv. costs 1) Tariff (ex VAT) Funding Sources for Investments User GPOBA fee, WA, grant CG, PO loan 2) 15 GPOBA Grant Efficiency Subs/ Subs/ connxn cap 3) US$, thousands U Sh/m3 percent perc ent US$, thousands US$ US$ Small towns Kachumbala 83 1, Wakiso 114 1, Luwero 94 1, Wobulenzi 68 1, Rukungiri 168 1, Kalisizo 93 1, RGCs 2,275 (7-yr) 90 2,056 Sipi 271 1, , Namutumba 820 2, , Magale 370 1, , Masafu 824 1, , Source: PAD 1) Inclusive of interest on pre-finance and investment loan during the GPOBA disbursement period 2) WA = local water authority; CG = Ministry of Water conditional grant; PO = private operator 3) Variance in per capita pop dependent on number of PWPs versus yard taps. Official Ugandan estimates 18 people per yard tap (3 families), and 300 people per PWP. Since the OBA approach was relatively new, the government and the GPOBA consulted stakeholders at all levels to secure buy-in. GPOBA also engaged the services of transaction advisor to assess the viability of towns and RGCs, scope the works and determine the subsidy levels, as Uganda s government lacked the internal capacity to design OBA projects. Lessons Learned at Project Preparation Data: Initial bidding had to be repeated due to a dearth of good quality data to make realistic projections of revenues and subsidy requirements. For example, unreliable data regarding water resources resulted in unnecessary subsidy allocations to the small towns of Luwero, Wobulenzi and Kalisizo, and a shortfall in the targeted number of connections in Rukungiri. To guide future planning, the sector needs to invest in systematic and continuous capturing of quality performance data. A water asset management policy should be developed to guide data collection and sharing. Communication: An effective communication strategy is crucial at all stages of the project cycle, and particularly important at project inception. OBA concepts need to be

16 effectively communicated to stakeholders to allay concerns and improve understanding of the OBA mechanisms. Since OBA involves off-government budget financing, it was not featured in major sector discussions or reviews. It was therefore unsurprising that even after successful implementation of the OBA pilot, skepticism and misconceptions about OBA remained. Future OBA projects should have a well-resourced communication strategy. Tariff: The tariff regime is crucial in the planning OBA schemes since it forms the basis for determining subsidy levels. Since Uganda s government had implemented a policy capping the tariff, the projected tariff had to be embedded in the DBO contracts with escalation clauses to assure the operators that they would recover their investments. Risks: The project appraisal document highlighted a number of risks to the project, which are listed in table 2.2. Most risks did not occur and therefore had no impact on the project outcome. Table 2.2: Risk Assessment Risk Occurrence Impact There are no interested bidders There are few bidders, and mainly incumbents win back their contracts The bids result in only modest investment efficiency gains Government reneges on ring-fenced tariffs DWD does not fulfill its obligations under the Memorandum of Understanding (MoU) Outputs are not achieved by Small Town operators Outputs are not achieved by RGC operators The risk did not occur; there were many bidders in the first and second rounds of bidding The bidders were many bidders; however, the incumbent operators won back most of their towns This risk did not occur; the efficiency gain from the competitive bidding process was 20 percent This risk did not occur; however, there was resistance from water authorities of Sipi and Magale on implementing the tariff escalation clause of the contract This risk did not occur; DWD fulfilled its obligations under the MoU with GPOBA and the agreement with PwC This risk did not occur; all small towns but Rukungiri achieved the planned outputs and some were higher This risk did not occur; all RGCs achieved the planned outputs No impact Low No impact Low No impact No impact No impact 16

17 2.2 Implementation The project s implementation was delegated by the fiduciary agent (PwC) to the MWE/DWD. Table 2.3 shows the original implementation plan provided in the project appraisal document. The table illustrates the delays in project implementation. Table 2.3: Project Original Implementation Plan and Actual Progress Milestone Expected Completion Actual Completion GPOBA provides subsidy commitment End-August 2006 Project design finalized October 2006 Fiduciary Agent/Grant Recipient November 2006 January 16, 2007 delegated by GoU based on bids quoted GPOBA Grant Agreement signed November 2006 February 12, 2007 Tender process for Small Towns and December March, RGCs launched 2006 Contracts signed between GoU and Operators March-April 2007 August to October 2008 Extensions/construction begins May 2007 Small Town subsidy disbursement begins June 2007 RGC subsidy disbursement begins June 2007 Small Town subsidy disbursement ends December 2008 June 30, 2012 RGC subsidy disbursement ends September 2009 June 30, 2012 Challenges affecting project implementation included the following: (i) Unanticipated and significant delays: The original timelines were ambitious and did not consider the difficulty in undertaking a PPP in the water sector in Sub-Saharan Africa involving small, local providers. OBA involved different bidding parameters than normal, and the tendering process had to be repeated due to late or no bid submissions for some lots, and some bids exceeding the maximum allowed ceiling. More time and resources than anticipated for small local providers. Shifting from input- to output-based approaches required further capacity development for public contracting authorities and the private operators than anticipated. (ii) Limited access to finance. Private operators needed affordable access to credit to pre-finance investments before output-based disbursements were made. Financing water projects was new to most financial institutions, which were reluctant lend money to operators. The banks wanted MWE to guarantee the loans to operators, and that was not possible. The International Financial Corporation (IFC) and Acumen Fund tried to address financing needs by introducing measures, such as including phasing outputs to reduce the pre-finance capital required and providing capacity building for private 17

18 operators and local banks. But most operators preferred to avoid high interest rates and collateral security, and relied on their own savings and suppliers credit instead of bank loans. DFCU Bank came on board later and provided a loan to Trandint for Busembatia water supply at a 17 percent interest rate. (iii) Higher advisory and financial intermediary costs (12 percent of the total subsidy amount) than traditional projects (four percent to eight percent). Clustering of towns might have lowered some transaction costs and lead to additional economies of scale. (iv) Availability of adequate water resource. The unreliability of water resources was a particular hindrance to project implementation in Rukungiri town. In Kacumbala, the quality of water deteriorated during rainy seasons. (v) Unclear authority over land acquisition. Under the project agreement, local water authorities were responsible for provision of land. However, some water authorities, such as in Sipi, did not have resources to pay for land. Consequently, the operator paid for the land, and this payment was not part of the operator s bid. Namutumba also had land issues that delayed project completion. In future OBA contracts, land acquisition should be a responsibility of the contractors. This is however, contrary to the Water Policy (1996) which vested the responsibility in the community to secure the land and meet any compensation to the affected landowners. (vi) Political interference in the location of connections, land acquisition and setting the tariff. It took more than a year for Wobulenzi town council to agree on the location of the new connections. Land disputes in Sipi and Namutumba were partly fueled by politicians. The politicians in Sipi and Magale refused revision of the tariff to cover the costs of operation. (vii) Local fiscal issues. Some towns, such as Kacumbala, had unsettled debt for electricity bills. This information was not shared with the operators at the time of bidding. The operators had to pay the debts because the water authorities could not do so. (viii) Approval delays. The National Environmental Management Authority s (NEMA) approval of the Environmental Impact Assessment was delayed. The Wildlife Authority s grant for a water abstraction permit in Sipi was delayed. Lessons Learned from Implementation The highly competitive bidding process resulted in efficiency gains but left operators with little room to fund unexpected or emergency operational costs. For example, pumps in Rukungiri and Wobulenzi broke down and operators had to replace them. The replacement cost was not included in their bids. The Ministry of Water and Environment needs to explore ways of mitigating such risks in future. 18

19 Given increasing constraints to available water resources, the risks related to availability should be reassessed and shared among providers, authorities and users. Any expansion of the OBA must consider stakeholders capacity to monitor and regulate private operators throughout the life of their contracts. 2.3 Project Outcome The immediate project outcome was increased access to clean and safe water supply. Over 54,000 people were estimated to have access to clean and safe water, compared to the original target of 45,000. The number of persons served by town/rgc is shown in table 2.4, below. Table 2.4: Population Served under OBA Towns/ RGCs New Of which New population connections PWPs covered Small Towns 1, ,282 Kachumbala ,044 Wakiso 300 5,400 Luwero 310 5,580 Wobulenzi 200 3,600 Rukungiri* 131 2,358 Kalisizo 150 2,700 Busembatia ,600 RGCs ,204 Sipi ,102 Namutumba ,950 Magale ,578 Masafu ,574 Factors that hindered the project s implementation included: Availability of alternative water sources, including springs (protected or unprotected), boreholes and wells or ponds. In the rainy seasons, operators reported a reduction in water sales since residents had access to alternative sources. It was recommended that improved water sources (boreholes and protected springs) within the towns/rgcs be operated and managed by the operator of the scheme. Disconnection for non-payment of water bills. Several reasons explain non-payment of water bills: 19

20 - Poverty was cited as a major factor in RGCs where the majority of people depended on agriculture and petty trades. Incomes were seasonal during harvest periods. - Men reportedly gave a higher priority to expenditure on alcohol than water in some places. - Some government institutions, such as the local police, took a long time to pay because of the bureaucracy involved in processing payments at the headquarters in Kampala. The table below shows the functional/active connections by town/ RGC. The percentage of functional/active connections ranged from 66 percent in Sipi to 95 percent in Luwero. Table 2.5: Functional Connections by Town/RGC Town/RGC Connections Total Active/ functional PWP percent Active 1 Wobulenzi Luwero Kacumbala Masafu Besembatia (4) 91 6 Namutumba (6) 85 7 Magale (8) 80 8 Sipi (2) 66 9 Kalisizo Rukungiri 1, Note: For PWP, numbers in brackets show active connections Unreliable water supplies in Rukungiri and Kalisizo. In Rukungiri, households received water once in a week. In Kalisizo, the water source was too small to serve one section of town. In Busembatia, one community was not regularly receiving water due to low pressure. Mismanagement of PWP. Some PWP attendants embezzled revenues and others remitted less money than expected. As a result, many PWP were disconnected, leaving some households without access to safe water. The mismanagement of PWP was not unique to OBA project; DWD and the National Water and Sewerage Corporation (NWSC) have experienced similar problems with PWPs in the past. NWSC is planning to replace PWP in Kampala with pre-paid meters. 2.4 Monitoring and Evaluation 20

21 The DWD was responsible for the overall monitoring and verification of output deliveries and for compliance with contractual covenants in the operate-and-manage (O&M) contracts between the local water authority and the operator. An independent monitoring and verification engineer oversaw the monitoring and verification of project outputs. This consultant was mainly responsible for approval of design and issue of design certificate to the private operators; verification of completion and functionality of the water distribution system; issuance of the output verification report (OVR) and reporting on the project's monitoring and verification indicators. The OVR was certified by the DWD before issuing output verification certificates (OVC) to the contractors. Operators prepared the disbursement requests after works had been certified. PricewaterhouseCoopers Limited also conducted due diligence, including physical inspection of the works done by the operators. The World Bank carried out quarterly supervision missions to monitor the performance of the project. The missions involved meeting with project staff, government officials and field visits to the project sites. Semiannual reports were prepared by PwC and submitted to the World Bank. The reports provided the physical and financial progress of the project including challenges encountered. Indicators used to monitor the performance are shown in table

22 Table 2.6: Matrix of Project Performance Indicators Indicator Unit Baseline Actual as of June 2012 Comments/ Output: Number of new connections established Number of household yard taps (incl. number of people served per yard tap) Number of public water points (incl. number of people served per water point) Number Small Towns A total of 1,649 yard connections and 22 Public Water Points verified and functional after 6 months Rural Growth A total of 728 yard taps and 17 public water points verified and functional after 6 months Remarks Total number of households benefiting from increased access Average monthly household expenditure on alternative sources of water prior to project Service availability (daily average) Unscheduled water supply failures lasting more than 30 minutes (number per year) User assessment of project Number US$ Hours per day no/poor/fair /good/very good U Sh per jerry can of 20L (U Sh 25 was the official rate per jerry can but is hardly if ever applied, and definitely not in the dry season.) 2,356 connections were verified and total beneficiaries are 54,486 (About 9,000 households) The average monthly expenditure for a household with a yard tap was U Sh 5000 (or 3 M 3 ). The average cost of a 20 liter jerry can was U Sh 100. Most towns had water 24 hours except for Rukungiri, which had once in a week because of water resource constraint. The users interviewed rated the project as very good. Average number of people per household 6 people per household and each yard tap serves 3 households or approx. 18 people; and each PWP was estimated to serve 300 persons. Rukungiri did not have a reliable water source. There was no accurate data captured to measure this indicator The reason given was the project saved them from drinking contaminated water. Average monthly household consumption of water Cubic meters liters/capita/day liters per capita per day The household water consumption ranged from 3 5 units per month. Average household has 6 persons 22

23 Indicator Unit Baseline Actual as of June 2012 Comments/ Remarks Water sales per year Cubic meters Wobulenzi = 46,258 Luwero = 114,528 Kacumbala = 2,395 Busembatia =32,476 Rukungiri = 45,925 Kalisizo = 56,172 Namutumba =29,207 Magale = 14,971 Masafu = 16,018 Sipi =4,710 Average tariff for beneficiary households Collection Rate (U Sh collected / U Sh billed) U Sh / cubic meter percent Wobulenzi = 2,036 Luwero = 1,511 Kacumbala = 3,273 Busembatia =1,842 Rukungiri = 1,918 Kalisizo = 1,469 Namutumba =2,103 Magale = 1,150 Masafu = 2,053 Sipi =1,600 Wobulenzi = 77 percent Luwero = 90 percent Kacumbala = 81 percent Busembatia =74 percent Rukungiri = 94 percent Kalisizo = 96 percent Namutumba =99 percent Magale = 98 percent Masafu = 80 percent Sipi =68 percent Average income of beneficiary household Investment cost per person given access to safe water Average subsidy per person given access to safe water Affordability ratio (average water bill per household per month divided by average monthly household expenditure) Specify currency US$ US$ percent Less than US$1 per day per person Between US$60-70 per person (small towns around US$21/cap, RGCs closer to US$100/cap) 6-12 percent of household income but noting the very low consumption levels (so this ratio needs to be calibrated to the actual consumption when comparing across projects) US$0.99 Yard tap (RGC) = US$2,941 Yard tap (Town) = US$500 US$69 6 to 10 percent of household income was spent on water bills The range of household water consumption was 3 to 5 units per month. On average a unit cost U Sh

24 Indicator Unit Baseline Actual as of June 2012 Comments/ Remarks Range from U Average was U Sh 100 per Sh50 U Sh liter jerry can per 20 liter jerry can Retail price (per Jerry Can) paid by nonconnected households from (PWP, Neighbors Yard Tap, Vendors) Are there plans for scaling up the project further, and/or is the project design being replicated elsewhere? Comparator information: average subsidy required for similar output under similar circumstances. Other development impacts recorded: e.g., results from consumer surveys, cost or time savings compared to pre-project situation, health benefits recorded, etc.] Gender impact: Please explain any specific impact this project has/will have on women/girl-children, both positive and negative US$ Plans for scaling up by GOU. Expected time savings, fewer expenditures on treating diarrhea, or time off from work, etc. Expected to have clear benefits to women and girlchildren who spend substantial time collecting water. Greater safety with closer water points. Data sources: Monitoring and Verification Reports and semiannual reports. Government plans to scale up OBA but lacks resources Time saved for collecting water from far leading to increased school attendance by children and attending to other domestic chores. Expenditure on treating waterborne diseases saved and used for other domestic purposes. Increase income from the sales of water by households with yard taps and from businesses that depend on water e.g. restaurants, hotels, bars etc. Girls have more time for attending school saved from collecting water from long distances in the morning and evening. Reduced burden of collecting water from far distances leading to better quality of life for women and children. Biggest challenge is reliability/ensuring timely payments to the private sector, under the GOU funding. Data was not readily available 24

25 2.5 Safeguard and Fiduciary Compliance (i) Environmental and Social Safeguards The project s effect on the environment was classified under category B. DWD was responsible for ensuring that operators comply with environmental and social safeguards. NEMA executed and approved environmental impact assessments (EIAs) for greenfield subprojects. Public disclosures and consultations were carried out during EIAs. In the case of Sipi, where the intake works were located in a forest reserve, an abstraction permit (permit to extract and use ground water) was obtained from the Uganda Wildlife Authority before works could commence. Resettlement Action Plans were implemented and affected persons were compensated. 2.6 Post-Completion Operation/Next Phase The pilot project was intended to inform future planning for water supply in small towns and RGCs. The project was still ongoing because most of the contracts were still running. DWD will continue to monitor the performance of these contracts. The Water Regulation Unit in DWD reviews operators business plans and receives periodic reports that are used as the basis for monitoring. Plans are underway to handover some of the water supply contracts of the OBA towns to NWSC. At the time of this writing, water supplies in towns Wakiso, Kalisizo and Rukungiri are due to be handed over to NWSC at the end of the OBA contracts. The Ugandan government has no immediate plan to expand the OBA project, partly due to the challenge of ensuring prompt payments given the government s mainstream financial cash flows. For example, the government could take over six months to pay contractors and consultants in the event of a budget cut. The government has asked the World Bank to consider funding the second phase of the project. Other development partners have yet to decide whether to adopt an OBA approach. The government s next step should be to share the achievements and lessons learned from the pilot project with all relevant stakeholders, including The Ministry of Finance, Planning and Economic Development (MoFPED ) and development partners. 25

26 Replication of the Project As authorities consider the next phase of the project, local actors are successfully implementing the OBA approach. For example, the Mpigi Water Authority and the private operator (Trandint) adopted an OBA approach, with MWE/DWD providing materials for the expansion of the water supply, and the private operator agreeing to pre-finance the cost of labor and recover from the tariff. Such initiatives should be considered in other towns as well as in urban areas. OBA should be one of the approaches for implementing urban water supply-based performance-based subsidies and leveraging Uganda s local private sector expertise and financial resources. Part of the conditional grant earmarked for small towns investments could be implemented in an OBA manner through the design, build, and operate approach. The lessons and experiences of the OBA in the health, water, and information, communications and technology (ICT) sectors should be shared with all relevant stakeholders to inform future OBA projects. 26

27 3. ASSESSMENT OF OUTCOMES 3.1 Relevance of Objectives, Design, and Implementation The PDO was to increase piped-water access for Ugandans living in selected small towns and RGCs in a sustainable manner through increased participation of local private operators. The PDO met Ugandan government objectives in the following ways: (i) The Government s goal was to provide 70 percent of all urban residents access to safe drinking water by 2005, with 100 percent access by 2015; it aimed to provide access to 77 percent of rural areas by The PDO contributed to Uganda s attainment of both these goals and the United Nations Millennium Development Goals (MDGs). (ii) The Government introduced local private operators in 2001 in the management of small town water supply. Increasing participation of local private operators was consistent with the Government Water Policy (1999). (iii) The PDO was also consistent with the Bank s Joint Assistance Strategy (JAS) for Uganda ( ). Pillar 5 of JAS had the objective of improved access to safe water supply and sanitation. It is also in line with the Uganda Country Assistance Strategy (CAS) FY11-15 Strategic Objective 2 to enhance public infrastructure and to increase access to and quality of water and sanitation services. The implementation challenges related to land acquisition and pre-financing were overcome through the combined effort of the operators, the MWE and the Bank and did not affect the achievement of project outcomes other than to delay them. Intermediary Output Indicators (IOI): the number of increased working connections and number of people provided with access to improved water supply targets were surpassed by 21 percent. 3.2 Achievement of Project Development Objectives The PDO to increase piped-water access for Ugandans living in selected small towns and RGCs in a sustainable manner through increased participation of local private operators were achieved: the plan aimed to provide 45,000 persons with access to safe water, and it surpassed this number by 9,000. Table 3.1 shows the achievement by town/ RGC against planned targets at appraisal. Since Busembatia was added to the pilot after initial appraisal, it did not have original targets. 27

28 Table 3.1 Achievements against Planned Targets Town At completion June 2012 At Appraisal August 2006 YT PWP Pop. Served YT PWP Pop. Served Small Towns , ,444 Kachumbala , ,304 Wakiso , ,640 Luwero , ,500 Wobulenzi , ,800 Rukungiri , ,500 Kalisizo , ,700 Busembatia , RGCs , ,620 Masafu , ,300 Magale , ,080 Sipi , ,940 Namutumba , ,300 Total ,486 1, ,064 Specific PDO achievements included: Time savings: The installation of yard connections and PWPs reduced time spent collecting water at distant water points. Traditionally, children are woken to collect water for the household before going to school. Many children would arrive late to school and others missed entire days. Yard taps solved this problem. Women in the towns and RGCs reported that yard taps also reduced the burden of carrying water from distant places. In Sipi, women were especially grateful that the project saved them from climbing mountains to search for water. They reported using the time saved to improve the hygiene of their homes and for agriculture production. Reduction in waterborne diseases: Households reported a reduction in the incidence of diarrheal diseases compared to the situation before the water supply. Households also did not have to spend as much money and resources on treating diarrheal related ailments. Financial viability: Other than in Kacumbala, the project proved financially viable in all small towns. Three of the seven towns did not need subsidies. Kacumbala reportedly lacked sufficient connections, reliable water quality (with water quality deteriorating in rainy seasons), and had access to alternative water sources. Among the RGCs, the project was not viable in Sipi due to few connections, low tariffs and high overhead costs. Schemes in the remaining RGCs were breaking even at the time of reporting. Transparency: Systemic corruption was minimized through separation of verification and certification works from the payment agency. The overseeing engineer did not have 28

29 discretionally powers to pay for the water connections, as was the case with conventional contracts. 3.3 Efficiency Efficiency gains: The competitive bidding process (based on lowest subsidy required) resulted in a 20 percent efficiency gain. Three towns did not need a subsidy and the investment cost was recovered from the tariff revenue. Investment cost: The subsidy per person gaining access was US$0 to US$36 (for extension projects) and US$83 to US$108 (for greenfield projects), compared with US$30 to US$138 and US$37 to US$250 in traditional input-based projects respectively (COWI 2009). Table 3.2 below shows the investment cost efficiency at appraisal and completion. Table 3.2: Investment Cost Efficiency at Appraisal and Completion Town/ RGC Investment cost efficiency Appraisal (US$) Investment Investment cost per cost per new connection access to piped water Investment cost efficiency Completion (US$) Investment Investment cost cost per per capita connection Small Towns Kachumbala Wakiso Luwero Wobulenzi Rukungiri* Kalisizo Busembatia RGCs , Sipi 2, Namutumba 2, , Magale 5, , Masafu 2, , Processing contracts and payment requests from operators improved. The time to processing payment requests averaged two weeks, compared to more than 30 days under the conventional system. 29

30 Revenue collection became more efficient as operators gave staff targets to meet in order to recover the investment and meet monthly loan obligations. Average collection efficiency was 90 percent. Non-revenue water reduced under most schemes from about 22 percent to 15 percent. Savings of US$792,998, representing 24.7 percent of the grant total, were obtained by surpassing output targets by 21 percent. 3.4 Overall Outcome The project was satisfactory based on the level of achievement of the PDO and the relevance of the project. 4. ASSESSMENT OF RISK TO DEVELOPMENT The risk of not achieving the development outcome was diminished as targets were surpassed. Operators have continued to connect additional people albeit at higher cost. For example, Namutumba added 89 new connections after the project s completion. Similarly, Masafu added 62, Magale, 43, and Sipi, 9 new connections, respectively. 30

31 5. ASSESSMENT OF BANK AND BORROWER PERFORMANCE 5.1 Bank Performance (a) Bank performance in ensuring Quality at Entry It took less than eight months from preparing the Project Information Document (PID) in June 2006 to signing the grant agreement in February However, project completion required 28 months beyond initial estimates. (b) Quality of supervision The Bank s team carried out five missions and two reviews during the life of the project. The team provided timely no-objection requests on procurements and payments. The government team and fiduciary agent were satisfied with the Bank s oversight. (c) Overall performance The Bank team provided the necessary technical guidance and ensured active participation of all stakeholders. The project team worked closely with Uganda s government in resolving issues that arose during the implementation of the project. The Bank team was amenable to proposals from stakeholders, as shown by the revision of the Operational Manual and restructuring papers. 5.2 Borrower Performance (a) Government performance The government (MWE/DWD) performed all the roles outlined in the MoU with GPOBA and the grant agreement, including procurement of operators for 11 towns and RGCs; engaging a verification and monitoring engineer; certifying works; signing performance agreements with all water authorities; ensuring compliance with environment and social safeguards during project implementation; and monitoring performance indicators and reporting. The government collaborated with water authorities and operators to resolve land disputes and other implementation issues that arose. Stakeholders held a workshop on the completion report, and results from the workshop are described in Annex 5. (b) Other implementing agencies The Fiduciary Agent fulfilled its obligations as outlined in the Grant Agreement including; delegation of implementation function to DWD, submission of withdrawal applications to 31

32 the Bank; disbursements of funds operators based on certified outputs and other service providers, submission of reports to the World Bank; financial management, and preparation financial statements. Water authorities signed agreements with operators, participated in the procurement and agreed on the tariff, provided land, and monitored the operators. (c) Overall performance (i) (ii) (iii) (iv) The borrower fulfilled all obligations in the MoU and Grant Agreement; The borrower resolved most of the issues which during implementation and achieved the targets; The borrower achieved the PDO and surpassed the target by 21 percent, and achieved 25 percent saving on the grant. There was no evidence from mission supervision reports indicating the borrower s performance was unsatisfactory. 6. LESSONS LEARNED Application of OBA Concepts Risk: OBA aims at transferring the performance risk to the service provider. However, in this case the performance risks transferred to the operators were inadequately defined. As a result, private operators took over unanticipated performance risks as demonstrated by inadequate ground water resources in Rukungiri, difficulty with land acquisition in Sipi, and poor estimation of demand in Magale and Masafu. Future OBA projects should analyze all the performance risks and allocate them to the right party or share some of them. Pre-financing: Private operators were supposed to pre-finance the delivery of outputs and be reimbursed by the project. It was not easy for the private operators to secure loans because most financial institutions had little experience financing water projects. The IFC and Acumen Guarantee Fund tried to market the venture, but banks remained reluctant to lend money to private operators. Meanwhile, operators themselves were reluctant to borrow at interest rates ranging from 17 percent to 27 percent, and at collateral values ranging from 105 percent to 120 percent of the value of the loan. MWE and PwC issued letters of comfort to the banks but this did not sway them. As a result, operators had to resort to their savings and obtain materials on credit. It was suggested that financial institutions used to channel OBA funds should be compelled to provide loans to operators at favorable terms. More advocacy and education about the water industry is also necessary. Independent verification: The role of an independent consultant to verify and monitor the project was extolled by private operators as it reduced systemic corruption tendencies experienced in conventional approaches. The verification/certification of 32

33 works was delinked from the payment system, as the verification consultant lacked discretionary powers to certify works and pay, as was the case in conventional contracts. However, independent verification does not guarantee the quality of the works as the consultant checks the final out (yard tap or PWP). It was difficult to verify the quality of buried materials and parts. Water supply schemes are designed for 20 years and the effects of poor quality materials may not manifest in the first five to 10 years, when the operator is still running the scheme. A proper technical audit is needed to guarantee quality. Efficiency: OBA provided incentives for the private operators to be efficient in order to recoup the returns on their investment. As a result they set performance targets for their employees, which improved billing and revenue collection efficiency. Non-revenue water totaled less than 20 percent for most towns, and collection efficiency surpassed 90 percent. Sustainability: The subsidy was one-off, and the systems were supposed to sustain themselves based on the tariff revenue. A flexible tariff policy (based on the market forces) is necessary for the long-term sustainability of these schemes. In addition, the revenue sharing ratios embedded in the management contracts (95 percent for the operator and five percent for the Water Board expenses) need to be reviewed because they do not allow for any savings for major repairs. Future contracts need to consider longer term financing of major repairs from the revenues generated. Most RGCs will not break-even in the short term partly because of the small number of connections and high operational costs. Tariff revenues cannot be relied upon to sustain these schemes. The subsidy for RGCs had catered for shortfall in tariff revenue. Government has to plan for O&M grants for these schemes once the OBA contracts expire. For the OBA approach to be mainstreamed in the sector, government payment systems must improve: contractors/operators must trust that they will be timely compensated for their investments. Conclusion The OBA pilot project demonstrates advantages in financing infrastructure services. Contrary to the view that the private sector has little appetite for taking risk in the water sector, the pilot has shown that relatively small private companies are willing to take on pre-finance and performance risk, and have strong incentives to roll out improved services quickly. By shifting performance risk to the service providers, the project increased accountability for results and efficiency. Given the successes and lessons learned, the water sector is in a strong position to lobby government and development partners to adopt OBA as a method of choice for financing the expansion and management of small towns water supply. 33

34 The World Bank and GPOBA are working with the government and its development partners to explore how to scale up the OBA in Uganda s small towns and rural growth centers. 34

35 APPENDIX A. PROJECT COSTS AND FINANCING Table A.1: Budget, Actual Cost and Balance on Release (US, dollars) Component Budget Actual Balance on Release Component 1: Installation of water service connections for selected households. Component 2: Bank/GPOBA Supervision Component 3: 2,800,000 1,908, ,127 50,000 $ n/a 360, ,096 (145,096) (i) Operating Costs for Fiduciary Agent; (ii) Consulting Services for technical evaluation and contract implementation support; and (iii) Consulting services for monitoring and verification of outputs (iv) Auditing services 210,000 95,000 45,000 10, ,970 90,205 54,921 0 By the close of the project in June 2012, US$792,973 had not been utilized and was returned to the World Bank. Detailed of the receipts and payments are shown on the following page. 35

36 Table A.2: Receipts and Payments 36

37 APPENDIX B. OUTPUTS BY COMPONENT Table B.1: Project Outputs by Small Town and RGC Component/ Project Outputs Small Towns Water Supply Kachumbala Wakiso Luwero Wobulenzi 158 yard taps and 4 PWP completed 6.3 km of pipe work laid well above the planned 5.5 km 2 bulk meters installed 5.5 km of water mains laid 300 yard taps 30 KW standby generator was procured and construction of a generator house 250 yard connections 6 km of distribution network laid 310 connections 200 yard tap connections 2.5 km of pipe work laid Rukungiri 131 yard connections out of the planned km pipe work laid for the distribution system Kalisizo 150 yard tap connections. 3 km of distribution system laid 0.7 km for the transmission system laid Busembatia 1 Borehole for intakes 1 submersible pumps 1 pump house supply and installation of transmission system bulk water meters and distribution system 1 storage reservoir (100 m 3 ) 418 connections Rural Growth Centers (RGCs) Sipi Namutumba raw water intake (1 surface intake) settling tank (9,400m) of raw water & clear water transmission system 2 bulk meters and 2 pressure filters chlorination equipment and house 6 km of distribution network 2 break pressure/storage tanks (96 m 3 each) 89 meter yard connections and 5 metered public stand posts raw water intake (1 borehole and 1 spring intake) pump house 3.6 km of transmission mains 2 bulk meters 2 pressure filters 37

38 Component/ Project Outputs chlorination equipment and house storage tank(200 m 3 ) 7.15 km distribution network 275 functional yard taps Magale Masafu raw water intake (1 surface intake) settling tank (6,855 m3) of raw water & clear water transmission system 2 bulk meters and 2 pressure filters 7.2 km distribution network (7,247 m) storage reservoir (75 m 3 ) 2 break pressure tanks 221 yard taps and 12 PWP raw water intake (2 boreholes) 2 pump houses transmission mains 2 bulk meters 7.2 km distribution network Storage tank (200m 3 ) 143 functional yard tap connections Table B.2 below provides a summary of the percentage of additional distribution pipeline and connections. Table B.2: Additional Connections and Length of Pipeline after OBA Town/ RGC Pipeline (km) Connections Total OBA percent increase Total OBA percent increase 1 Wobulenzi Luwero Kacumbala Busembatia Rukungiri Kalisizo Namutumba Magale Masafu Sipi Wakiso

39 APPENDIX C: ECONOMIC AND FINANCIAL ANALYSIS No detailed cost benefit analysis was taken at project appraisal to determine the net present value (NPV) and economic internal rate of return (EIRR) of the project. The focus of this section therefore is to analyze the economic and financial viability of the towns/rgcs after the expiry of project support. The analysis covered the production and sale of water, tariff levels, billing and revenue collection and cost of operations for small towns and RGCs as shown below. Water production and sales: In small towns, annual production ranged from a low of 2,395 m 3 in Kacumbala to a high of 114,528 m 3 in Luwero. Production in the remaining towns ranged from 41,725m 3 to 69,117m 3. In RGCs, the annual rate of water production ranged from 12,581 m 3 in Sipi to 39,775 m 3 in Namutumba. The billing efficiency for small towns ranged from 78 percent in Busembatia to 100 percent in Rukungiri. For RGCs, the billing efficiency ranged from 40 percent in Sipi to 95 percent in Magale. Details are shown in table C.1 below. It is apparent from table C.1 that Kacumbala small town and Sipi RGCs did not perform well in terms of the production and sale of safe water. Table C.1: Performance on Key Economic and Financial Indicators Indicator Small Town RGC Wobule nzi Luwero Kacum bala Busemb atia Ruku ngiri Kalisi zo Namutu mba Magal e Masa fu Sipi 1 Water produced 39, ,58 (units) 53, ,52 2,395 41,725 45,12 69, Water sold (units) 46, ,09 1,996 32,476 45,92 56,172 29, , Billing efficiency (percent) 7 Average rate per unit 2,036 1,511 3,273 1,842 1,918 1,469 2,103 1,124 2,053 1,600 (U Sh/M3)-Tariff 4 Amount billed (U Sh 94, ,25 6,533 59,826 88,52 82,535 61, ,031 thousands) Amount collected (U , ,665 Sh thousands) 0 6 Unit running cost (U , ,106 Sh/M3) 9 Collection efficiency (percent) 10 Non-revenue water (percent) 3 Total operating costs (U Sh thousands) 45,096 74,186 7,680 40,407 65, ,806 39, ,551 Source: Operators annual reports (2012/13) Revenue: The tariff level in small towns ranged from U Sh 1,469 in Kalisizo to U Sh 3,273 Kacumbala. The average was U Sh After discounting for inflation of 15 percent, the tariff level in small towns was higher than the average level at appraisal of 39

40 U Sh 1,000. In RGCs, the tariff ranged from U Sh 1,124 in Magale to U Sh 2,103 in Namutumba. Sipi and Magale were using gravity flow and therefore had no pumping costs compared to Namutumba and Masafu. After discounting for inflation, the average tariff level for RGCs was slightly lower than at appraisal, from U Sh 1,500 to U Sh 2,000. The annual revenue collected in small towns ranged from U Sh million in Kacumbala to U Sh million in Luwero. The average was U Sh 69.7 million. The annual cost of operation ranged from U Sh 7.680m in Kacumbala to U Sh m in Luwero. The average was U Sh 48.7 million. Kacumbala was experiencing losses. Improving collection efficiency (81 percent) and non-revenue water (17 percent) may reduce loss. The annual revenue for RGCs ranged from U Sh 7.6 million in Sipi to U Sh 60.8m in Namutumba, with an average of U Sh 28.4 million. The annual operational cost for RGCs ranged from U Sh 5.5 million in Sipi to U Sh 39.0 million in Namutumba. This means that the RGCs can meet their operational costs. Sipi could do better if it reduced non-revenue water from the current level of 60 percent. Recommendation: Future OBA projects should consider combining a number of towns and/or RGCs under a single management contract to minimize overhead costs and facilitate cross subsidies. 40

41 Length of the Operators contracts Table C.2: Operators Contract Duration and Sources of Funds Operator Trandint Ltd Town/RGC Duration of contract Source of funding Luwero 5 years Barclays Bank Wobulenzi 5 years (loan) 27 percent Kacumbala 5 years interest rate Savings Busembatia 5 years DFCU Bank(loan) 17 percent interest rate Jobatov JV Wakiso 5 years Stanbic (overdraft) 19.5 percent interest rate Savings Kol Kagulu JV WSS-Services Sipi 7 years Savings Magale 7 years Masafu 7 years Namutumba 10 years Rukungiri 5 years Savings Kalisizo 5 years Credit facility from suppliers Credit facility from suppliers 41

42 APPENDIX D: BRIEF ON IFC WORK Background Improved access and use of safe water and sanitation facilities are among the key priority areas for the government of Uganda. To achieve this goal, the government decided to decentralize rural water supply delivery. Implementing this plan relied on local governments, which depend heavily on grants from the central government which are financed in turn by donors. Since 2001, the government has implemented PPPs in suburban and rural areas, and established management contracts with private operators in over 70 small towns. These contractual arrangements were generally weak and plagued with capacity challenges, both at the national and regional levels. IFC s Work Uganda s government leveraged IFC s experience in public-private partnership transactions and business development support. The IFC provided support for three program components: transaction advice, public sector capacity, and access to finance. IFC also tapped into the resources of its financial markets and small- and mediumenterprise departments to address financing constraints. IFC also developed a training program based on the proposed generic management contract to address weaknesses in the area of contract administration among public sector stakeholders. Additional elements were built into the program to facilitate continuity upon completion, including a workshop curriculum that could be easily replicated. Transaction Structure As part of its role as lead advisor, the IFC conducted due diligence in 10 small towns, uncovering contracts of short duration and varying performance indicators. To address these shortcomings and ensure ease of management and administration, the IFC proposed a generic contract with a minimum term of five years to appeal to both private operators and lenders. The IFC also analyzed the geography of areas in need and proposed clustering towns within close proximity to one another. Clustering allows for contracts that capture economies of scale, specifically on human resources and capital investments. However, since funding was secured for only one of the 10 towns, this approach could not be implemented and a bidding process was ultimately implemented solely for the town of Busembatia. Prior to the IFC s involvement, private operators in Uganda raised financing by using overdraft facilities provided by the banks or secured loans using other existing business. Improving access to financing was a priority. Post-Tender Results 42

43 In June 2010, IFC awarded the first five-year management contract to a private operator for the town of Busembatia. The lowest bidder, Trandint Limited, won the contract with a bid of US$270,000. Advisory work was supported by the Austrian Development Agency. The Global Partnership on Output Based Aid provided funding for capital investment costs to support the private operator. Key project results are listed below. Residents of Busembatia are enjoying expanded access to water at the same tariffs for the next five years. Four hundred new water connections were installed. DFCU Bank, a Ugandan commercial bank, loaned approximately US$100,000 to the winning bidder for the Busembatia contract. Seventy representatives from local authorities participated in two IFC-designed training programs for public sector stakeholders. USAID is using IFC s relationship with local banks as a model for developing a risksharing product for banks to lend to private operators. Unfortunately for this project, access to finance for POs did not work out due to difficulty ensuring timely payments under the Government of Uganda s financial cash flow system. 43

44 APPENDIX E: GENERAL RECOMMENDATIONS/WAY FORWARD Table E.1 Completion and Results Report for the OBA in Water Supply in Uganda s Small Towns and RGCs No. Questions Response to Question 1 Could some of the OBA successes be translated to sanitation? General Recommendations Criteria of measuring outputs ought to be more precise, for instance, targeted number of people versus served people. A percentage of monthly revenue should be saved to cater for corrective maintenance. The critical considerations of the OBA approaches need to be summarized in the Report. Concrete recommendations on what needs to be done next must be developed. The OBA project did not cover sanitation. However, it is possible to apply the experience and lessons learned from the OBA for water supply to sanitation, such as solid waste management. The challenge is that most of the available local operators do not have experience and/or skills in waste management. 2 To what extent was the OBA project evaluated in terms of technical efficiency, i.e. inputs versus outputs? 3 Why did three of the pilot towns not require subsidies and why was Kachumbala town not financially viable? The technical efficiency of OBA small towns and RGCs was shown in appendix C (table C.2) of the Completion Report. The parameters evaluated included water produced, water sold, non-revenue water, etc. The three pilot towns which did not require subsidy included Wobulenzi, Luwero and Kalisizo. These towns did not require subsidy because the bid prices for the successful bidders could be covered by the revenue generated from the tariff within the contract period. Kacumbala was not viable because (a) it had very few customers, (b) the existence of many alternative water sources particularly during rainy season, and (c) water quality during rainy season deteriorated and the cost of treatment was high. 44

45 4 What was meant by a flexible tariff? A flexible tariff refers to the tariff based on market forces. In 2006, the then Minister for Water capped the tariff at U Sh 2,500 per cubic meter. 5 What does tariff indexation mean in the OBA environment? Tariff indexation means providing for tariff escalation to compensate for increase in costs (e.g. power, inflation etc.). OBA contracts had a tariff 6 How did competitive bidding affect performance? 7 Were there constraints during land acquisition and were beneficiaries sensitized. If so, which institution was responsible? escalation clause to cover increases in costs of operations. The competition was stiff and the prices of bids returned were low. As a result, the winning bidders made small profits not adequate for them to grow. There were challenges pertaining to land acquisition in Sipi (for intake works), and Namutumba (for power transmission line). Sipi did not have money for land compensation and the contractor had to pay his own costs. In Namutumba, the problem was caused by local politicians who wanted to frustrate the project. The communities were sensitized by MWE/DWD during the inception phase and their contribution in form of provision of land was highlighted. 8 What should the next steps be to move forward? The next steps are included in the completion report subsection 2.6 on page 28. MWE/DWD should share widely with sector stakeholders including MoFPED and Development Partners the achievements and lessons learned from the OBA pilot project, agree on what needs to be improved and how to replicate the project. 9 How good was the reporting and were there any aspects that affected the OBA approach? The reporting on the OBA activities was good because there was the Independent Verification/ Monitoring Consultant with the responsibility of reporting on subprojects, and PwC responsible for project reporting on a semiannual basis. The operators also submitted quarterly reports to MWE/DWD on the performance of the water supply system. The factors that affected the OBA approach were discussed under Section 2 of the completion report regarding project preparation, implementation and outcome levels. 10 Will the OBA report be adopted? The Government and the World Bank will consider the final report and decide whether to adopt it or not. 11 Will local authorities solicit for new The MWE/DWD will formerly guide the Water Authorities on what to do 45

46 bids after the expiration of the OBA contracts? 12 How do you decide which schemes qualify for the OBA subsidy? 13 What went well? What did not go well? 14 Can benefits from this business be documented so that banks can rely on that information to give loans? 15 How can aid be delivered more efficiently, i.e. performance for results? when the OBA contracts expire. However, the government, through the MWE/DWD, plans to transfer 89 small towns to National Water and Sewerage Corporation for better management in the next five years. Three OBA small towns (Wakiso, Kalisizo and Rukungiri) are among candidate towns to be transferred to NWSC at the end of the OBA contracts. The decision for which scheme qualified for OBA was based on the feasibility study done by COWI and bid prices returned. The schemes that had projected revenues sufficient to recover the investment during the contract period did not require subsidies. What went well included the following: outputs exceeded targets by 20 percent and there was a saving on the project budget by the same percentage. What did not go well was that the operators/ contractors had trouble securing financing from financial institutions. The benefits will be summarized in a pager and shared with banks and other stakeholders. This will be accompanied by informing banks and other financial institutions of investment opportunities in water supply. OBA provides an efficient way of delivering aid because beneficiaries pay for outputs, not inputs and processes. There is a high degree of transparency and low risk of corruption. 46

47 Map F.1 Small Towns and RGCs covered by the Project Legend Rural Growth Centers (RGCs) Small Towns 47