Document of The World Bank Report No: ICR00003433 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-41380, IDA-44290, IDA-49060) ON A CREDIT IN THE AMOUNT OF SDR 69.1 MILLION ($100 MILLION EQUIVALENT) CREDIT IN THE AMOUNT OF SDR 59.6 MILLION ($98.1 MILLION EQUIVALENT) CREDIT IN THE AMOUNT OF SDR 63.6 MILLION ($100 MILLION EQUIVALENT) TO THE DEMOCRATIC SOCIALIST REPUBLIC OF SRI LANKA FOR A ROAD SECTOR ASSISTANCE PROJECT June 22, 2016 Transport and ICT Global Practice Sustainable Development South Asia Region CURRENCY EQUIVALENTS (Exchange Rate Effective as of February 21, 2016) Currency Unit = Sri Lanka Rupee LKR 1.00 = $0.00712824 $1.00 = LKR 140.2871 FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS $ All dollars are in United States dollars unless otherwise indicated ADB Asian Development Bank AF Additional Financing AM Aide-memoire CAS Country Assistance Strategy CPS Country Partnership Strategy DO Development Objective EIRR Economic internal rate of return EMP Environmental Management Plan ERR Economic rate of return FM Financial management GoSL Government of Sri Lanka HDM Highway Development and Management Model ICR Implementation Completion Report IRI International roughness index ISR Implementation Status Report JICA Japan International Cooperation Agency MoF Ministry of Finance MoH Ministry of Highways MoT Ministry of Transport M&E Monitoring and evaluation MTR Midterm Review NPV Net present value NRMP National Road Master Plan NWSDB National Water and Sewage Development Board PAD Project Appraisal Document PDO Project Development Objective PMU Project Management Unit RAMS Road Assets Management System RDA Road Development Authority RMTF Road Maintenance Trust Fund RSAP Road Sector Assistance Project TAC Technical Advisory Committee VoC Vehicle operating cost SRI LANKA ROAD SECTOR ASSISTANCE PROJECT CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design ............................................... 1 2. Key Factors Affecting Implementation and Outcomes .............................................. 8 3. Assessment of Outcomes .......................................................................................... 17 4. Assessment of Risk to Development Outcome......................................................... 26 5. Assessment of Bank and Borrower Performance ..................................................... 27 6. Lessons Learned ....................................................................................................... 30 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners .......... 31 Annex 1. Project Costs and Financing .......................................................................... 33 Annex 2. Outputs by Component ................................................................................. 35 Annex 3. Economic and Financial Analysis ................................................................. 37 Annex 4. Bank Lending and Implementation Support/Supervision Processes ............ 42 Annex 5. Beneficiary Survey Results ........................................................................... 44 Annex 6. Stakeholder Workshop Report and Results................................................... 45 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ..................... 49 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ....................... 62 Annex 9. List of Supporting Documents ...................................................................... 63 Annex 10. QAG Report ……………………………………………………………….64 Annex 11. MAP ……………………………………………………………………….74 Regional Vice President: Annette Dixon Country Director: Francoise Clottes Senior Global Practice Director: Pierre Guislain Practice Manager: Karla Gonzalez Carvajal Task Team Leader: Amali Rajapaksa ICR Team Leader: Radia Benamghar DATA SHEET SRI LANKA ROAD SECTOR ASSISTANCE PROJECT A. Basic Information SL - Road Sector Country: Sri Lanka Project Name: Assistance Project IDA-41380 Project ID: P086411 Credit Number(s): IDA-44290 IDA-49060 ICR Date: 06/22/2016 ICR Type: Core ICR Democratic Socialist Lending Instrument: Specific Investment Loan Borrower: Republic of Sri Lanka Original Total XDR 69.10 million Disbursed Amount: XDR 180.16 million Commitment: Revised Amount: XDR 180.16 million Cancelled amount XDR 12.14 million Environmental Category: B Implementing Agencies: Ministry of Ports and Highways through Road Development Authority Ministry of Provincial Council and Local Government Ministry of Finance and Planning Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 12/18/2003 Effectiveness 03/16/2006 03/16/2006 06/23/2008 06/23/2008 03/15/2011 03/15/2011 Appraisal: 09/26/2005 Restructuring(s): 04/08/2013 04/08/2013 09/23/2014 09/23/2014 06/30/2015 06/30/2015 Approval: 12/15/2005 Midterm Review: 01/12/2009 03/02/2009 Closing: 09/30/2011 06/30/2015 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: High Bank Performance: Moderately Satisfactory -i- Borrower Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Satisfactory Government: Moderately Satisfactory Implementing Quality of Supervision: Moderately Satisfactory Moderately Satisfactory Agency/Agencies: Overall Bank Overall Borrower Moderately Satisfactory Moderately Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Project at Quality at Entry No 2 (Likely to achieve DOs) any time (Yes/No): (QEA): Problem Project at any time Quality of Supervision No None (Yes/No): (QSA): DO rating before Satisfactory – – Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Central government administration 6 6 Rural and Inter-Urban Roads and Highways 90 90 Subnational government administration 4 4 Theme Code (as % of total Bank financing) Administrative and civil service reform 17 17 Environmental policies and institutions 16 16 Infrastructure services for private sector development 33 33 Other social development 17 17 Rural services and infrastructure 17 17 E. Bank Staff Positions At ICR At Approval Vice President: Annette Dixon Praful C. Patel Country Director: Francoise Clottes Peter C. Harrold Practice Manager/Manager: Karla Gonzalez Carvajal Guang Zhe Chen Project Team Leader: Amali Rajapaksa Isabel Chatterton - ii - ICR Team Leader: Radia Benamghar ICR Primary Author: Radia Benamghar Mirtha Pokorny Tabjeel Ashraf F. Results Framework Analysis Project Development Objectives (from Legal Agreement) The objective of the Project is to lower transportation costs through sustainable delivery of an efficient national road system. Revised Project Development Objectives (as approved by original approving authority) The PDO remained unchanged throughout the life of the Project. Note on Indicators: For the purposes of a split evaluation, indicators below refer to either RSAP-1 (original project and first Additional Financing), or RSAP-2 (second Additional Financing). Questions regarding validity of some of the baseline data for RSAP-1 arose during implementation. This issue is discussed in sections 1.3, 2.3 and 3.2. (a) PDO Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years Lower transportation costs through sustainable delivery of an efficient national road system RSAP–1 Reduction in average network vehicle operating costs for average vehicle of Indicator 1 (a) 3.8% by 2010 Value (quantitative or LKR 23.9 per km LKR 22.9 per km LKR 14.97 per km qualitative) Date achieved 11/16/2005 09/30/2011 09/30/2011 Target achieved by more than 100%. Comments RSAP-1: The final value well surpassed the 3.8% target (by several hundred percent); (including % however, an exact determination of the extent of the overachievement for this indicator achievement) cannot be precisely quantified, due to uncertainties related to baselines, mentioned above. (See section 1.3, 2.3 and 3.2 for further discussion.) Indicator 1 (b) RSAP-2 Reduction in average network vehicle operating costs for average vehicle Value (quantitative or LKR 14.6/km LKR 10.5/km LKR 13.88/km qualitative) Date achieved 03/15/2011 03/15/2011 06/30/2015 Comments Original target 128% achieved; revised target not achieved. While the VoC was (including % reduced by 4.9% (exceeding the original 3.8% RSAP-1 target), target was revised under achievement) RSAP-2, increasing the percentage reduction to 28% (LKR 13.88/km). The lower - iii - achievement may have been partially affected by post-conflict country priorities, discussed in section 3.2. RSAP-1 Reduction in average network roughness from 9.5 in the International Indicator 2 (a) Roughness Index (IRI) in 2005 to 8.39 IRI in 2010 Value (quantitative or 9.5 8.39 6.45 qualitative) Date achieved 11/16/2005 09/30/2011 09/30/2011 Target achieved by more than 100%. As with Indicator 1 (a) the target was exceeded Comments (272%); however, there are uncertainties regarding baseline. (See sections 1.3, 2.3 and (including % 3.2.) achievement) RSAP-1I Reduction in average network roughness from 9.5 on the International Indicator 2 (b) Roughness Index (IRI) Value (quantitative or 6.2 5.9 5.9 qualitative) Date achieved 03/15/2011 09/30/2014 06/30/2015 Comments 100% Achieved. (including % achievement) RSAP-1 Reduction in percentage of the network in poor and bad condition from 52% Indicator 3 (a) in 2005 to 35% in 2010 Value (quantitative or 52% 35% 40% qualitative) (%) Date achieved 11/16/2005 09/30/2011 09/30/2011 Comments Partially achieved (70%) (including % achievement) Indicator 3 (b) RSAP-1I Reduction in percentage of the network in poor and bad condition to 35% Value (quantitative or 38 35 – 35 qualitative) (%) Date achieved 11/16/2005 09/30/2011 – 06/30/2015 Comments 100% Achieved. Although there had been considerable progress, the target had not (including % been fully achieved under RSAP-1 so target was not changed. achievement) RSAP-1 Increase in annual maintenance expenditure from $13 million to $46.3 million Indicator 4 (a) in 2010 Value (quantitative or 13 million 46.3 million – 44 million qualitative) ($) Date achieved 11/16/2005 09/30/2011 – 09/30/2011 95% Achieved. The Borrower reported maintenance expenditures had been Comments substantially higher than the target ($56 million vs $46.3 million). The task team noted (including % the $44 million figure excludes funding for periodic maintenance in 2010 that has been achievement) channeled outside of the RMTF, and when these funds are added, the target is - iv - substantially exceeded. However as the higher figures do not appear in relevant ISRs or the Project Paper for the 2nd AF, the lower figure is reflected here. See related comment for 4 (b) below. Indicator 4 (b) RSAP-2 Progressive increase in annual road maintenance expenditure Value (quantitative or LKR 4.2 billion/year LKR 6.6 billion/year – LKR 5 billion qualitative) Date achieved 03/15/2011 12/31/2014 – 06/30/2015 Not achieved. This indicator is a continuation of indicator 4 (a) under RSAP-1. The term “progressive” was added (in the Project Paper) to capture annual improvements Comments over the life of the Project. Unit of measurement changed from $ to LKR. (including % As above, Borrower reported as achieved and task team noted maintenance expenditures achievement) had been substantially higher, and that the intent was to increase annual maintenance, while this indicator measured the amount channeled through the RMTF. RSAP-2 Improved level of satisfaction of road users (for 3 roads: A002, A003, and Indicator 5 A006) (Percentage) A002 40.38% A002 62.16% Value A003 36.22% A003 74.14% (quantitative or – Combined: 38.3% Combined: 68.1% qualitative) A006: 81.2% Date achieved 03/15/2011 – 06/30/2015 Achieved. While user satisfaction surveys for both A002 and A003 were carried out Comments using standard methodology, including baselines, A006 was assessed using a recall post (including % survey which did not include a baseline. Satisfaction improved substantially, however, achievement) no target was established. (b) Intermediate Outcome Indicator(s) Component 1: Maintenance and Rehabilitation of National Roads Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years Indicator 1(a) RSAP-1 About 620 km of national roads resurfaced Value (quantitative 0 620 – 617 or qualitative) Date achieved 11/16/2005 09/30/2011 – 09/30/2011 Comments (incl. % 99.5% Achieved. All activities carried out as planned. achievement) Indicator 1(b) RSAP-2 134 km of national roads resurfaced Value 43 (Orig. section) (quantitative 0 134 A2 17 km 78 or qualitative) A3 17.7 km Date achieved 03/15/2011 09/30/2014 09/30/2014 06/30/2015 -v- 100% Achieved (of revised target) New indicator for 134 km of national roads under RSAP-1I. It later became evident that widening the full 134 km from 2 to 4 lanes would Comments entail substantial land acquisition issues which could not be resolved during the life of (including % the Project. Therefore, 134 km section was divided into 2 phases: phase 1 retained 43 achievement) kilometers of the original section; phase 2 replaced the remainder with road sections A2, 17 km and A3, 17.7 km. 43 + 35 = 78 km. Indicator 2a RSAP-1 Reduction in IRI for project roads from 7.8 to 2.9 (for 620km) Value (quantitative 7.8 2.9 – 2.8 or qualitative) Date achieved 11/16/2005 09/30/2011 – 09/30/2011 Comments (including % 103% Achieved. achievement) Indicator 2b RSAP-1I Reduction in IRI for project roads Value 6 2.8 2.33 (quantitative 4.3 (A002) 2.4 1.9 or qualitative) 3.1 (A003) 2.4 2 Date achieved 03/15/2011 09/30/2014 09/30/2014 06/30/2015 More than 100% achieved. Comments This indicator refers to the road section resurfaced that was split into two phases (See (including % indicator 1 (b) above.) The original target for phase 1 remained. New targets of 2.4 IRI achievement) were set for the two new sections. Road sections A6, A002 and A003 were achieved by 115%, 127%, and 151%, respectively. RSAP-1 50% of relevant staff in the Planning and Maintenance Division of the RDA Indicator 3 trained in road asset management Value (quantitative or 0 50 – 67 qualitative) (%) Date achieved 11/16/2005 09/30/2011 – 09/30/2011 Comments (including % 134% Achieved. achievement) Indicator 4 RSAP-2 Reduced travel time on project national roads in minutes Value 66 42 40 (quantitative 13.8 for (A002); 12.5 11.31 or qualitative) 26.5 for (A003) 23.9 23.58 Date achieved 03/15/2011 09/30/2014 09/30/2014 06/30/2015 Comments More than 100% achieved. (including % This indicator was added for the reasons mentioned above in indicator 2b. Road achievement) sections A6, A002 and A003 were achieved by 108%, 191%, and 112%, respectively. Indicator 5 RSAP-2 About 60 km of national roads on which periodic maintenance was carried out. Value (quantitative or 0 60 – 64.50 qualitative) Date achieved 03/15/2011 09/30/2014 – 06/30/2015 - vi - Comments (including % 107% Achieved. achievement) Component 2: Maintenance and Rehabilitation of Rural Roads Pilot Indicator 6 Rural roads improved (635 km) - RSAP-1 Value (quantitative or 0 635 km 157 km 167 km qualitative) Date achieved 11/16/2006 09/30/2011 09/30/2014 09/30/2015 Achieved 106% of revised target RSAP-1 and RSAP-1I. Comments The initial target length of 635 km for improvement of rural roads was based on (including % assumption that only minor maintenance would be required. However, upon achievement) formulation of a new Rural Road Strategy Study, and after extensive stakeholder consultation, it was decided to carry out more extensive improvements, thus necessitating a reduction to 157 km, which was achieved after restructuring in 2013. Indicator 7 RSAP-1 Reduced travel time on project rural roads by 10% Value 10% reduction (quantitative or 0% – 60% (overall average) qualitative) (%) Date achieved 11/16/2005 09/30/2011 – 09/30/2011 Target exceeded substantially. While detailed travel-time baselines could not be Comments prepared for the PAD, large scale monitoring and assessments were carried out. In (including % many cases, even 60% improvement does not capture the full level of improvement as achievement) many of the roads had been non-motorable. Component 3: Institutional Strengthening and Policy Support Indicator 8 Annual road maintenance program is approved by the RMTF Value Approved (quantitative or Not approved Approved annually qualitative) Date achieved 03/15/2011 06/30/2015 06/30/2015 Comments 100% achieved (including % This is an RSAP-1I indicator. achievement) G. Ratings of Project Performance in ISRs Date ISR Actual Disbursements No. DO IP Archived ($, millions) 1 06/21/2006 Satisfactory Satisfactory 4.39 2 12/12/2006 Satisfactory Satisfactory 13.61 3 06/22/2007 Satisfactory Satisfactory 24.59 4 12/20/2007 Satisfactory Satisfactory 32.67 5 06/30/2008 Satisfactory Satisfactory 64.85 6 10/01/2008 Satisfactory Satisfactory 74.21 - vii - 7 04/30/2009 Satisfactory Satisfactory 80.79 8 08/06/2009 Satisfactory Satisfactory 87.21 9 02/07/2010 Satisfactory Satisfactory 156.78 10 12/16/2010 Satisfactory Satisfactory 159.52 11 02/16/2011 Satisfactory Satisfactory 159.52 12 08/20/2011 Satisfactory Satisfactory 164.08 13 04/14/2012 Satisfactory Moderately Satisfactory 181.01 14 10/16/2012 Satisfactory Moderately Satisfactory 186.79 15 05/15/2013 Satisfactory Moderately Satisfactory 221.64 16 12/01/2013 Satisfactory Satisfactory 234.07 17 02/07/2014 Satisfactory Satisfactory 240.26 18 03/24/2014 Satisfactory Satisfactory 240.26 19 05/05/2014 Satisfactory Satisfactory 244.63 20 12/23/2014 Satisfactory Moderately Satisfactory 269.46 21 06/22/2015 Satisfactory Satisfactory 279.95 H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Restructuring Disbursed at Reason for Restructuring and Approved PDO Date(s) Restructuring Key Changes Made Change DO IP in $ millions Additional Financing of $98.1 million to help finance the costs associated with the cost overrun of civil works 06/23/2008 -- S S 64.85 under the national roads component of the Project due to unprecedented rise in world fuel prices in 2007/08. A second Additional Financing of $100 million to scale-up the activities through additional road sections in two phases for the rehabilitation of 134km in 2 phases: Phase 1. Rehabilitation of 43km of roads from Kanthale to Trincomalee on the A6. Phase 2. Rehabilitation of 91 km of 04/12/2011 -- S S 162.73 roads Ambepussa to Dambulla on the A6. The operation also included (a) financing of periodic maintenance through the Bank assisted RMTF, (b)TA of RMTF, (c) capacity building of RDA; and (d) the piloting of a safety corridor. In addition, the original project was extended by three years from September 30, 2011 to September 30, 2014. A Restructuring to address the 04/08/2013 -- S MS 209.06 following changes: Phase 2 related to rehabilitation of 91 km of roads - viii - Ambepussa to Dambulla on the A6 was replaced by priority urban sections in the western province. It included rehabilitation and upgrading of approximately 34.7 km of ‘A’ class urban roads. It also included design review and supervision of these sections. In addition reallocation of categories. The proposed restructuring was for extension of closing date by 6 months from September 30, 2014 to March 31, 2015. In addition, reallocation of financing proceeds from category 1 09/23/2014 -- S MS 262 (supporting parts A, B3, and B4 of the Project) to category 2 (supporting Part B1 of the Project) an amount of $1.4 million to finance the periodic maintenance contracts through RMTF. The proposed restructuring was for partial cancellation of the Project proceeds in the amount of SDR 8.5 06/30/2015 -- S S 279.95 million (approximately US$12 million). In addition, the Project was extended by three months from March 31, 2015 to June 30, 2015. - ix - I. Disbursement Profile -x- 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal Country Context 1. At the time of appraisal of the Road Sector Assistance Project (RSAP) (Credit 4138-CE, ‘the parent Project’) the Government of Sri Lanka (GoSL) expected the economy to grow at annual rates of 6 to 8 percent during the next five years. Infrastructure was considered paramount to achieving these expected growth targets. The GoSL planned to increase the level of public expenditure to reach 8 percent of gross domestic product, a substantial increase from the 5 percent in 2004. To that end, a broad agenda was laid out for the road sector, including connecting poor regions and production centers to domestic and international markets, building a national highway system and an integrated road network, enhancing road safety, and promoting private sector participation in the sector. Sector Issues 2. Road infrastructure in 2004 was inadequate to support economic growth. Uncontrolled roadside development, years of neglect, and poorly maintained roads had resulted in low traffic speeds and poor levels of services, discouraging long-distance traffic and hindering the spread of economic activities and development away from the Colombo metropolitan area. Given Sri Lanka’s road density of 1.5 km per km2, there was an urgent need to address the issues of asset preservation. 3. The focus on the road sector was taking place at a time when there was substantial progress, albeit slow, on the peace front. There were considerable improvements in the lives of the people in the north and east, and the entire country was benefitting from the cease-fire of 2002. Roads were being reopened, suspended rail services were resumed, and considerable reconstruction had occurred. In this scenario, the GoSL and the Asian Development Bank (ADB) agreed on a medium-term sector reform framework as part of ADB’s Road Sector Development Project that was under implementation. The World Bank, Japan Bank for International Cooperation (JICA), and ADB agreed in November 2004 to coordinate their activities in the sector to support a medium-term reform program that was based on three pillars: (a) strengthening the Road Development Authority (RDA), a statutory institution assigned to the Ministry of Highways (MoH) and responsible for implementing the development strategy for the sector, as well as strengthening the provincial road agencies; 1 (b) establishing a mechanism for providing 1 The Ministry of Transport (MoT) has overall responsibility for policies concerning land transport services, including bus transport and railways. The National Transport Commission, a regulatory body under the MoT is responsible for policy formulation and subsidy disbursement. The MoH has overall responsibility for policies and programs concerning the national road network. The RDA under the MoH is responsible for planning and managing the national roads. Provincial roads are the responsibility of Provincial Councils and local roads under the responsibility of local authorities. -1- sustainable funding for financing road maintenance; and (c) developing the domestic private sector in the road industry. Rationale for Bank Involvement 4. The Bank had remained disengaged from the transport sector in Sri Lanka for almost a decade. However, there was an increased demand for the Bank to support infrastructure development and the IDA Credit, the subject of this report, was seen as the appropriate vehicle for reentering the transport sector at a time when GoSL was showing signs of abandoning its previous reluctance to commit to sector reform. The emerging sector reforms included the closure of a state-owned construction company that was causing serious distortions in the market, and the decision to implement a Road Maintenance Trust Fund (RMTF) as an interim mechanism for financing road maintenance. GoSL expected that, eventually, dedicated fuel taxes would fund the RMTF. Furthermore, under the right circumstances, the Project would help to put in place the necessary legal and regulatory framework for a ‘second generation’ road fund, that is, with a governance structure that included major stakeholders. 5. Within this framework, the Bank agreed to take the leadership to address the issue of sustainable road maintenance financing; JICA agreed to take leadership of the issue of private sector development in the road construction industry; and ADB agreed to continue to lead the dialogue on institutional strengthening and capacity building. 1.2 Original Project Development Objectives (PDO) and Key Indicators 6. The Project Development Objective (PDO) in the Project Appraisal Document (PAD) differs slightly from the one in the Credit Agreement. In both documents, the PDO is stated as: The objective of the Project is to lower transportation costs through sustainable delivery of an efficient national road system. In the PAD, this statement is followed by the phrase that serves the needs of road users and the Sri Lankan public at large. The latter phrase is not included in the PDO in the Credit Agreement. The difference is not material within the context of the project and results framework; therefore, the Credit Agreement PDO is treated as operative for the purpose of this Implementation Completion and Results Report (ICRR). 7. The achievement of the PDO was to be monitored and evaluated using the following indicators (as described in the PAD): PDO Indicators  Reduction in the average network VoC for standard commercial vehicles of 3.8 percent by 2010;  Reduction in the average network roughness from 9.50 in the international roughness index (IRI) in 2005 to 8.39 IRI in 2010;  Reduction in percentage of the network in poor and bad condition from 52 percent in 2005 to 35 percent in 2010;  Increase in annual maintenance expenditure from projected US$13 million in 2005 to US$46.3 million in 2010. -2- Intermediate Indicators  About 620 km of national roads resurfaced;  Reduced IRI for project roads from 7.8 to 2.9;  50 percent of relevant staff in Planning and Maintenance Divisions of the RDA to be trained in road asset management;  About 635 km of rural roads improved;  10 percent reduction in travel time;  An increase in annual maintenance expenditure from projected $13 million in 2005 to $30 million in 2010; allocation based on Road Assets Management System (RAMS); regular consultation with road users; publication of the RMTF Annual Reports in public domain. 1.3 Revised PDO (as Approved by Original Approving Authority) and Key Indicators, and Reasons/Justification 8. The PDO remained unchanged throughout the life of the Project but the outcome and intermediate indicators were revised to reflect new activities supported by a $100 million Second Additional Financing in 2011. Changes affecting the indicators included the rehabilitation of additional road sections in two phases for 134 km in the East-West corridor, and the financing of road maintenance through the Road Maintenance Trust Fund (RMTF). (See Section 1.6., Revised Components.) 9. A subsequent restructuring was approved in 2013 to drop one of the National road sections from the Second AF, replacing it with two urban road sections. The replacement was in response to GoSL’ s request to widen the National road section from two to four lanes due to increased traffic; however, the widening would have entailed land acquisition adding a significant amount of time to the contract period, exceeding the timeline envisaged under the Project. 10. The new activities undertaken for RSAP-1I also provided an opportunity to utilize and incorporate improved road data that was being collected through enhanced monitoring of the road network. Inadequate data had, in some cases, resulted in indicators with deficient baselines. It was not until several years into the project that more reliable data became available. 11. Table 1 below summarizes the changes in the indicators and targets that occurred over the life of the project, including the original credit, the additional financings, and the restructuring in 2013. -3- Table 1. Indicator and Target Changes during the Project Original targets to be Type of Target revisions made at the achieved by 2011 as set out in Restructuring in 2013 change second AF the PAD made Reduction in the average network Reduction in the average network Target value VoCs for standard commercial VoCs for standard commercial vehicles from 23.9 (baseline) to vehicles to LKR 10.5 per km – 22.9 Reduction in the average network Reduction in the average network Target value roughness from 9.50 in the IRI to roughness from 6.2 to 5.9 in the – 8.39 IRI IRI Reduction in percentage of the Target value Reduction in percentage of the network in poor and bad condition network in poor and bad condition from 38 to 35 percent (not – from 52 percent in 2005 to 35 technically a revision, but PAD percent target had not been fully achieved so it was continued for RSAP2) Target value Increase in annual maintenance Progressive increase in annual expenditure from projected $13 maintenance expenditure from – million in 2005 to $46.3 million in LKR 4.2 billion ($45.1 m equiv.) in 2010 2011 to LKR 6.6 billion in 2014 New Improved level of satisfaction of – – Indicator for road-users (new indicator) RSAP-1I Type of Intermediate Indicators- PAD Second AF - 2011 Restructuring Paper - 2013 change made About 620 km of national project roads resurfaced About 134 km of national roads to New be resurfaced in two phases. Indicator for – (Revised in 2013 restructuring – RSAP-1I see cell one down to the right) 134 km revised to 78 km of national Target value – – roads to be resurfaced. Phase 2 split into 2 roads (A002 & A003) Reduced IRI for 620 km national project roads from 7.8 to 2.9 (same – – km as cell above) Reduced IRI for 134 km project New roads from 6.0 to 2.8 (same km as – Indicator for cell above) RSAP-1I Split into separate roads (reduced km) Target value necessitated having individual IRI – – targets: from 4.3 to 2.4 for A002 road and from 3.1 to 2.4 for A003 Reduced travel time for 134 km New – project roads from 2.3 to 1.8 (same – Indicator for km as cell above) RSAP-1I Split (reduced km) required new travel Target value – – times: from 13.8 to 12.5 for A002 -4- Original targets to be Type of Target revisions made at the achieved by 2011 as set out in Restructuring in 2013 change second AF the PAD made road and from 26.5 to 23.9 for A003 About 60 km of national roads on New – which periodic maintenance has – Indicator for been carried out. RSAP-1I New Annual road maintenance program – – Indicator for is approved by Board of Trustees. RSAP-1I 50% of relevant staff trained in – – road asset management 157 km of rural roads improved (This Target value. About 635 km of rural roads – was changed based on the new Rural improved Road Strategy Study) Reduced Travel Time (rural roads) – – (Same km for indicator above) Routine and periodic maintenance – – funding increased to $46.3 million 1.4 Main Beneficiaries 12. The primary beneficiaries of the Project are the road users and road-related transport services. More specifically, road users and road transport services directly benefit from improvement in key linkages, with decreased transport costs and travel times, which improve connectivity and road safety for its users, and promote economic growth in the region. In the socioeconomic context, the improvement of local roads directly influences the daily social life of the population who live in rural areas. As a result, rural populations have better access to public amenities such as schools, health care, business centers, and markets. In addition, the execution of rural roads generated new jobs and stimulated the local economy through the indirect purchasing of local goods and services. The rural development improved safety for working women. Furthermore, the highways safety component of the project helped to decrease road accidents and fatalities and contribute to lower accident-related road costs to the economy. 1.5 Original Components 13. The parent Project (RSAP-1) consisted of the following three components (see Table 2 for costs).  Component 1: Maintenance and Rehabilitation of National Roads. Consisted of the following: (a) civil works along national roads, comprising resurfacing and improvement of about 620 km of Class A and B roads; (b) technical assistance for the selection, design and construction supervision of these works; (c) training and institutional strengthening; and (d) incremental operating costs.  Component 2: Maintenance and Rehabilitation of a Rural Roads Pilot. Consisted of the following: (a) rehabilitation and maintenance of about 635 km of rural roads, and strengthening of the rural road management and maintenance systems in three project Provincial Councils and nine project Pradeshiya Sabhas; (b) feasibility, design, and construction supervision consultant services; (c) training and institutional strengthening; -5- and (d) incremental operating costs. This was a pilot project to help the GoSL for future rural road improvement programs.  Component 3: Institutional Strengthening and Policy Support. This component comprised technical assistance to improve sector policies for road maintenance financing. 1.6 Revised Components 14. The original components were revised and updated at the time of the second AF for $100 million and the 2013 Restructuring. The components below include the new activities funded under the AF. Component 1: Rehabilitation of Priority National Roads in the East-West Corridor Providing Connectivity to the Eastern and Northern Provinces. This component consisted of resurfacing and improving about 134 km of class A roads which was part of the government’s strategy to provide connectivity to the Northern and Eastern Provinces. It was designed to be implemented in two phases. While phase 1 was completed as planned, Phase 2 was modified under a 2013 restructuring, due to the government’s request that the road sections be widened from two to four lanes, because of increased traffic. The proposed widening, however, required significant land acquisition which could not be completed within the timeline envisaged under the Project. Therefore, priority urban sections in the western province were selected for upgrading in Phase 2. Component 2: Institutional Strengthening and Improvements in Asset Management Practices. This component provided funding for periodic maintenance, technical assistance and strengthening of the systems within the RMTF and RDA relating to road maintenance, institutional strengthening of the RDA, and addressing issues identified through the capacity assessment relating to road safety. 15. The component allocations under each key phase of the project – the parent project, the first and second additional financings, and the restructuring in 2013 – are shown in Table 2. -6- Table 2. Revised Project Costs by Components ($ millions) Components 2005 2008 Original + First Additional Finance (Original Project) (1st AF) Component 1 - Maintenance and Rehabilitation 82.64 66.64 of National Roads Component 2 - Maintenance and Rehabilitation 11.04 of Rural Roads Pilot Component 3 - Institutional Strengthening and 0.66 Policy Support Unallocated (physical and price 31.46 5.67 contingencies/First AF) Includes VAT Sub-Total 100.00 98.10 **Total Project Costs (original +First AF) 198.10 Components with Second Additional 2011 2013 Finance and Restructuring 2013 (2nd AF) Restructuring Component 1 - Maintenance and Rehabilitation 85.50 81.00 of national Roads Component 2 - - Institutional Strengthening and 14.50 19.00 Asset management practices Sub-Total 100.00 Reallocation **Total Project Costs 100.00 TOTAL (Original + First AF + Second AF + 298.00 RP) 1.7 Other Significant Changes 16. First Additional Financing 2008. The First AF under RSAP-1 (for $98 million equivalent) was approved on September 16, 2008 to cover financing gaps due to: (a) the unprecedented oil price increase in the world market that had led to the escalation of prices in oil-based materials; and (b) the use of higher standards for national highways to improve safety environment as stipulated in the National Road Master Plan (NRMP) adopted in December 2007. (See Factors outside the control of the Government, pages 9-10, for additional detail.) 17. Second Additional Financing 2011. As mentioned above, the purpose of the AF was to scale up the parent project and enhance its impact through the rehabilitation of national roads, providing funding for periodic road maintenance, technical assistance to the RMTF and RDA, and address safety issues identified in a capacity assessment. The Government was keen to achieve quick results in order to: (a) capitalize on the benefits of the cessation of the conflict and its impact on the development of the previously conflict affected areas, and (b) maximize the usage of the already established implementation mechanisms and institutional arrangements in order to scale-up activities within the agreed extended timeframe. 18. Restructuring 2013. In addition to the changes already mentioned, the length of rural roads targeted for improvement was reduced from 635 km to 157 km in Component 2 of the parent project. Subsequent to the formulation of the Rural Roads Strategy Study financed under the Project, and after extensive stakeholder consultations carried out during project preparation, it was decided that a full upgrade of the rural roads should be carried out, rather than proceeding with a relatively low cost but temporary maintenance regimen, i.e., pothole filling, patching, etc. The -7- decision was made to provide a more comprehensive upgrade, which required a reduction in the number of kilometers of road. 19. Restructuring 2014 was a six-month project extension to allow sufficient time to complete the rehabilitation on sections of A002 and A003. 20. Restructuring 2015 included a partial cancellation of $12 million equivalent at the request of the Borrower due to underuse of price and physical contingencies and savings resulting from sliding fuel prices, and lastly, the depreciation of the Sri Lanka Rupee. The restructuring also included a final project extension of three months to accommodate the completion of unfinished works. 21. Establishing and functioning of the RMTF. The Ministry of Finance (MoF) was responsible for the implementation of all activities related to the set-up and adequate financing for a trust fund related to sustainable road maintenance finance. However, the RMTF under the MoF was not fully operationalized because of administrative difficulties and lack of capacity. This resulted in re-establishment of the RMTF under the Ministry of Ports and Highways to take the agenda forward through the Second AF. 22. RMTF and road maintenance funding. The funding for road maintenance was channeled to the RDA through the RMTF under an annual budgetary allocation. Although it was originally envisaged that taxes from the sale of petrol and diesel would be made available to the RMTF, the funding for road maintenance was provided as a budgetary allocation. The annual funds collected from the sale of fuel were only $8 million to $10 million—far less than the maintenance needs. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 23. Soundness of the background analysis and design. The project responded to the government’s transport priorities where civil conflict had been prolonged for many years and had contributed to the poor condition of the road network. Immediate rehabilitation was required to augment the condition of the road network. The project design was highly relevant to Sri Lanka’s needs and in line with the priorities of the Country Assistance Strategy (CAS). The FY2003–06 CAS aimed to improve social and economic inclusion of the poor and facilitate broad-based economic growth. Both the government and the Bank strategies placed high priority on developing road infrastructure and recognized its contribution to the development of the private sector and the desired increase in revenues to achieve high levels of public investment. The lessons learned from earlier operations included: (a) providing adequate post-project maintenance; (b) the importance of environmental and social management; (c) improving rural roads management; (d) providing targeted training and capacity building for local governments; (e) the need for the Project’s design to include the first steps toward developing an appropriate sector policy framework, (f) empowering and strengthening the local authorities to discharge responsibilities, and (g) providing targeted and practical training programs built into the rural roads pilot component. -8- 24. Paticipatory process. The project was prepared in a participatory manner with the involvement of key stakeholers, including government ministries, designated government implementation agencies, local and provincial government units, development partners and local communities, and road users. 25. Risk and mitigation measures. A range of risks were identified during project preparation and appropriate mitigation measures were considered. The main risks were related to: (a) the consolidation of the peace process and political stability to oversee meaningful sector and economic reforms; (b) exogenous shocks to the economy impairing aggressive revenue raising strategy; (c) vested interests supporting the status quo; (d) weak national construction capacity to undertake the works; and (e) institutional capacity for the day-to-day implementation of the Project. These risks were mitigated by the already proven strong commitment of the government and the resources committed by the ADB and the Bank to facilitate the process of implementing reforms. Finally, the risk of a weak national construction capacity to undertake works under the Project was addressed by suitable packaging of contracts to accommodate both National Competitive Bidding and International Competitive Bidding. One risk that was not forseen at appraisal, yet materialized during implementation was the effect of the rapid surge of the construction activity in the country, which contributed to the shortage of human and material resources and the resulting increase in construction costs. The issue of counterpart funding was well foreseen during preparation and eventually, remained under monitoring. 26. A Quality at Entry (QAG) assessment was carried out by the Quality Assurance Group on November 7, 2008 and the project was rated Satisfactory (likely to achieve its development objectives). In particular, the report comments, “The DOs expressed exactly what they neeeded, and they are still highly relevant.” The report also remarks, “The project design adequately evidences that measures were taken to promote positive social impacts…With regard to poverty, project covers the poorest segments of the country including provinces close to the conflict areas and proposes labor intensive interventions…Stakeholder consultations were organized in a manner that ensured gender equality.”2 2.2 Implementation 27. The project implementation period could be broadly divided into two phases: (a) RSAP-1 (parent project and First AF)—for a period of six years (December 2005– September 2011) (b) RSAP-1I (additional components through the Second AF and two Restructurings) for an additional period of four years (March 2011–June 2015), for an overall project period of about 10 years. 28. Project implementation, including compliance with both the original Project and its two AFs, was generally rated satisfactory in all of the Implementation Status Reports (ISRs) and 2 The report from the Quality Assurance Group is in annex 9 . -9- slipped into moderately satisfactory for about 18 months (April 2012, May 2013, and December 2014). Project implementing agencies had amassed considerable experience through implementation over the 10-year period and developed an adequate technical capacity in procurement, contract management, social and environment related aspects. There were some events that occurred during the Project’s life that delayed implementation. These are outlined below. Factors outside the control of the Government 29. The increase in construction costs in Sri Lanka necessitated an AF of $98 million, i.e. 59.2 percent of the Project cost. This cost overrun was due to two reasons:  A worldwide escalation of oil and commodity prices resulted in an increase in construction costs in Sri Lanka following the Project start-up. The analysis undertaken during project implementation showed that all key cost drivers under Components 1 and 2 suffered major increases between 2005 and 2009: Bitumen (433 percent), fuel (241 percent) among other factors. Works on the national roads under the parent project experienced cost overruns averaging 86.5 percent, and ranging from 51.4 percent to 132.4 percent.  The use of higher standards for national highways to improve the safety environment as stipulated in the National Road Master Plan (NRMP) adopted in December 2007. Under the Project, this new policy of the Ministry of Highways and Road Development necessitated widening of roads to 6.2 meter national standards with a 0.65 meter hard shoulder on the inside of the numerous bends. As a result of the change in width of carriageway, all culverts had to be extended and other related matters addressed. Factors generally subject to Government control  Unrealistic estimates. Bid prices were 6.8 percent above estimated prices at a time in which a construction boom brought about by reconstruction works after the end of the civil war, and rebuilding after the tsunami of December 2004 resulted in shortages of resources that affected the cost and pace of implementation.  Limited availability of bitumen during the first two years of project implementation due to the high demand created by construction works in the country. The state monopoly supplier, Ceylon Petroleum Corporation, was not able to increase supply to match the new demand. Eventually, after a concerted effort by international donors, the GoSL approved, starting in January 2008, importing bitumen for interested contractors through Maga Naguma, a subsidiary of RDA.  Unrealistic estimation of contract periods for works and designs. The original periods had been understated and the contracts neither fully appreciated nor reflected the work involved in difficult terrains. They also did not reflect the additional work due to variations in time needed to obtain environmental clearances, which later required additional time for expansion of the scope of work. - 10 -  Delay in obtaining Environmental Protection Licenses and/or Explosive Permits, processes that in many cases took more than 10 months.  Lack of coordination with utilities agencies. Civil works were slowed because of difficulties in getting sufficient crews mobilized to work simultaneously on several contracts, particularly for utilities relocation. Despite monthly coordination meetings being chaired by RDA’s management, the utility companies performed slower than expected. This was particularly true of the National Water and Supply and Drainage Board (NWSDB). For works on the Second AF, coordination improved substantially through a different allocation of responsibilities. In particular, RDA procured the water pipes and their laying was part of the civil work contracts. This allowed for the water and road works to be completed on these corridors simultaneously avoiding wastage due to uncoordinated activities that the country has been suffering for a long time.  Frequent turnover of the PMU staff. The PMU staff turnover which resulted in losing institutional capacity and memory was a major challenge to project implementation. Four different project directors headed the PMU during the course of the Project.  No appropriate approval delegation, ownership, and accountability from the MoF with regard to operationalizing the RMTF, which necessitated the reestablishment of the RMTF within the Ministry of Port and Highways.  Cost overruns on the national roads parent project could range from 51.4 percent to 132.4 percent, and time overruns from 7.6 percent to 180.7 percent. Cost overruns in the civil works under the Second Additional Financing were substantially lower. The cost improvements were achieved through better distribution of responsibilities between RDA and consultants for the different stages of implementation. For the original Project, design and supervision of works were entrusted to consultants. The lack of involvement of RDA’s provincial staff had, in some cases, led to variation orders being received several years after commencement of contracts.  With respect to supervision, the parent Project was affected by multiple problems. These ranged from changes in staffing of the consulting firm, to the difficulties in dealing with multiple projects spread over a large area. These problems were also compounded by (some) poorly qualified consultants both with respect to design as well as supervision.  During a workshop for the preparation of this ICR that took place in Colombo in June 2015, there was a consensus among the main stakeholders that the pressure to quickly start projects contributed to less than optimal designs. Time overruns were also substantially contained although contractors claimed that the time estimated to complete projects was too optimistic. 30. The Midterm Review (MTR) was carried out in March 2009 with 41 percent of funds disbursed (of RSAP-1). Both the PDO and IP were rated satisfactory. The MTR highlighted overall project progress, including lessons learned from the challenges discussed above. The MTR discussed some revisions to the Results Monitoring Framework and some changes were suggested, notably that the number of km of rural roads targeted for spot improvements would likely need to - 11 - be reduced from 635 to 157, and changing the road condition target. Although these changes were noted in the aide memoire, there was no formal restructuring of the results framework until the AF two years later. Part of the reason for this seems to have been government uncertainties during a period of rapid change, i.e., identifying and agreeing on the number and location of roads to be rehabilitated and a change in the government’s view post-MTR regarding changing the road condition targets because of budget considerations. That said, adjustments could have been made sooner in a restructuring. The relevant observations and findings from the MTR were used to corroborate the ICR’s team’s observations and analysis with respect to implementation challenges and progress. 31. The Rural Roads Pilot. The objective of the Rural Roads pilot was to develop a sustainable system that could be scaled up in future programs. The emphasis was on the preparation of a Rural Roads Strategy/Manual, development of a rural roads improvement program, providing maintenance on the targeted roads, and the creation of a rural roads database. While the number of kilometers was reduced in order to provide greater rehabilitation to a smaller network of roads, the works, as well as the strategy, database and other activities, were successfully completed and the pilot has been expanded to other provinces, an important step toward a sustainable system. 32. The Roads Maintenance Trust Fund (RMTF) was established in order to improve transparency in the allocation of resources and disbursement of funds, as well as to carry out efficient monitoring of the maintenance programs. The intention was to ensure a sustainable well- maintained road network in the long term by maximizing the use of resources applied to the rehabilitation of the network. 33. The Roads Maintenance component focused on supporting GoSL to set up an interim mechanism for road financing in the form of a trust fund under the Trust Ordinance for earmarked road maintenance funds. At the time, it was to support the writing of the draft bill, and its presentation to Parliament, and necessary legislation to establish a dedicated Road Maintenance Fund. The Fund would be financed by fuel taxes and managed as a statutory body by an independent board and funding. The RMTF was set up, as agreed, at the beginning of the Project under the MoF, and the GoSL appointed a Board of Trustees and a Technical Advisory Committee (TAC). A separate budget line allocated funds to the RMTF and these funds were transferred to RDA after the approval of its annual maintenance plan. 34. The Fund, however, only attained a meaningful role as an overseer of maintenance practices after the approval of the Second AF, late in 2011. Lack of strong commitment of the main stakeholders in fully launching the Fund combined by frequent changes in the Board of Trustees and TAC contributed to the delays. 35. During the preparation of the Second AF it was decided to return momentum to the Fund through the inclusion of a road maintenance component to be cofinanced by the Bank and RMTF. The Credit Agreement included a disbursement condition for adequate staffing of the Fund that, after problems finding suitable candidates, was accomplished by transferring some RDA staff. As a result, the role of the Fund in approving the annual road maintenance plans was carried out with an increased technical capability. In addition the Fund started to play an active role in overseeing works and ensuring the adequate use of the approved allocations. - 12 - 36. The Road Asset Management System (RAMS) has been successfully implemented. The implementation phase witnessed not only the enhanced capacity of the RDA and PMU for implementation but also improvement in the construction industry overall. The institutionalization of RAMS has resulted in sustained maintenance grants through the RMTF with more emphasis given to collecting scientific data. Though much improvement is needed for the analyses of such data, the presence of RAMS will lead to the sustainable functioning of a maintenance system for the highways in Sri Lanka. Awareness of environmental and social safeguards during construction activities has increased among the practitioners in the road sector. The public participatory process in these activities through stakeholder consultations, and the approval process of yearly maintenance plans, may open opportunities for enhanced quality, transparency, and accountability. 37. Road traffic safety pilot. Concerned by the fact that the substantial road improvements taking place in Sri Lanka were allowing increased speeds and potentially detrimental effects on traffic accidents, the Second AF included a pilot safety corridor to address these concerns. Twenty-seven km of the A002 was selected as the road safety demonstration corridor due to its existing high-level of road accidents. A committee was formed from multiple organizations that had not cooperated previously, and a considerable amount of time and effort was used to mobilize them and the communities. Organizations included, among others, RDA, the Traffic Police, Ministry of Health, four local authorities on the corridor, the University of Moratuwa, and the National Council for Road Safety (NCRS). The committee was tasked to scope and implement the demonstration corridor. Traffic Police coordinated public awareness measures and enforcement, working closely with the public to mitigate initial reluctance. In addition to awareness programs, interventions included placing of pedestrian crossings, especially in view of schools and bus stands, street lighting, provision of safety barriers, raised foot walks, availability of ambulance and fire facilities and training of staff for emergency care, provision of an emergency number to access these facilities, safety awareness for schools, three wheelers and private buses, enforcement of speeds, etc. Table 3. Changes in Accident Patterns Along A002 Pilot Safety Corridor 2015-2016 Category 2015 2016 Fatal 10 5 Grievous 40 5 Minor Injured 101 22 Total 151 30 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Use 38. M&E design. The project M&E systems were planned using the Results Framework designed at appraisal. The PDO outcome and intermediate results indicators were adequate from a technical standpoint to measure improvements in road conditions (IRI), reduction in VoCs, sustainability, and efficiency. The indicators were, in principle, appropriate to measure the PDO to reduce transportation costs through sustainable delivery of an efficient national road system. - 13 - 39. However, three out of the five PDO indicators encompassed the whole network, in particular, VoC, IRI and the percentage of roads in poor and bad condition. While the PDO and associated outcome indicators were to reduce the transport costs through sustainable delivery of an efficient national road system for the whole network (12,000 km), the Project works were responsible directly for only 755 km, which is about 6 percent. To address the greater network, the project incorporated a number of institutional strengthening activities; however, the results framework falters in showing direct causal links between the project’s institutional activities and the national targets. These weaknesses raise attribution issues. 40. The PDO could have been more focused by separating the sub-objectives more distinctly, i.e., lowering the transportation costs for the targeted project roads, and supporting sustainable delivery of an efficient national road system. In fact, the project indicators that measure the reduction of transport costs for the targeted roads in the project are precise and easily measurable, and closely related to the road improvements and related activities. It is the references to the national roads that weaken the framework. A relatively simple change in the language of the PDO and the three PDO indicators could have gone a long way towards strengthening the evaluative framework, i.e., substituting “targeted roads, or corridors” for “national network”. Many projects with similar components do just this: frame the objective to “reduce the transports costs along the targeted corridor,” instead of “reduce transport costs for the whole national network”. 41. The establishment of baselines and target values for RSAP-1 was also marred by flawed data. According to one aide-memoire,3 the baseline value of the network IRI for RSAP-1 was fixed at 9.5 based on 226 km out of 9,913 km of road network in the year 2004. Yet even the methodology used in assessing the 226 kilometers was not clear. This resulted in indeterminate baselines the VoC and IRI of the parent project. The Bank recognized the flaws and at various points took steps to improve the framework, but reliable data – even four or five years into the project – was not always available. It was only by the time of the Second AF in 2011, when annual national road surveys had been instituted, that the Bank team was able to agree with the government on significant changes to the results framework and establish clear and reliable baselines and targets. (The PDO target values for the VoC and IRI were revised because the Project had already achieved a much lower value than the target value in 2009.) 42. M&E Implementation and Use. RDA and the Ministry of Provincial Councils and Local Governments (MPCLG) were responsible for collecting, analyzing, and reporting project performance indicators. The Bank had access to the data necessary to analyze performance and other monitoring issues. The Bank used this information, to the extent possible, to carry out the project’s restructuring and prepare the AFs. The Results Framework matrix was adjusted accordingly, particularly at the Second AF. 43. An independent M&E report in 2009 found that monitoring alone was not sufficient and more focus needed to be placed on analysis and evaluation to improve strategic decision making. Improvements were made, albeit gradually. 3 Annexure 2 of aide-memoire of June 2008. - 14 - 44. Progress towards achievement of the final targets was monitored through regular project progress reports by RDA and the Bank’s implementation Status and Results Report (ISR). Progress under the institutional development component and compliance with related covenants, were reported in semi-annual progress reports prepared by RDA, while regular supervision missions and comprehensive aide-memoires monitored progress and results achieved. An MIS system was used by RDA to provide monthly progress reports to sector stakeholders, including the Bank, on physical and financial, as well as technical progress for all road sections and contracts. HDM-4 was used to monitor selected project indicators. The project was able to train an adequate number of persons on RAMS and RDA is well equipped with the latest survey equipment, including laser road profilometers and falling-weight deflectometers. Survey teams now collect data annually. 2.4 Safeguard and Fiduciary Compliance Environment Safeguards 45. The project triggered Bank safeguards policies Environmental Assessment (OP/BP 4.01) and Natural Habitats (OP/BP 4.04), and was classified as an environmental category B project. An Environmental Management Plan (EMP) was prepared by the RDA. Overall, the EMP was satisfactorily implemented by contractors and monitored under the supervision of the RDA. During implementation, this was confirmed with no major impacts or safeguard compliance issues arising. The successful adoption and implementation of the Bank’s environmental guidelines on the initially reluctant local and international construction firms is one of the important achievements of the Project. Sri Lankan law requires environmental assessments and actions for road constructions, but not for rehabilitations. However, the Project helped introduce and include environmental clauses as part of the contract documents of these rehabilitation works. By the end of the Project, many of the local contractors were adopting the same preventive and remedial environmental protection measures on other projects with no such requirements. Social Safeguards 46. Social safeguards management of the Project showed gradual progress throughout the Project implementation period. In keeping with the Bank’s safeguard policies, the RDA prepared an Environmental and Social Management Framework for the overall project. 47. The project triggered OP 4.12 (Involuntary Resettlement) to mitigate impacts to the properties and structures along the roads. However, no physical or economic displacement of persons or resettlement impacts were experienced by the Project and most of the lands used for expansion of the roads were within the domain of the RDA or obtained through voluntary land donations from project beneficiaries. 48. Most of the adverse social impacts were due to construction-induced issues related to expansion of the width of roads, affecting property owners residing along the roads, and drainage lines that obstructed planned improvements. Social Impact Assessments and site-specific Resettlement Actions Plans were prepared to identify the potential, adverse social impacts and mitigation measures were implemented to reduce them. A key highlight of the Project was the systematic implementation of a grievance redress mechanism according to the Policy Framework. Also, for the first time in a road project in Sri Lanka, an independent third-party monitoring system - 15 - was introduced to engage project-affected persons to provide feedback and systematic monitoring of project implementation. In addition to the above measures, the Project undertook a series of awareness creation programs on HIV/AIDS in compliance with the Bank’s policy. Overall, implementation of social safeguards was Satisfactory throughout the Project period. Procurement 49. Procurement of goods, works, technical assistance, and non-consulting services was carried out in accordance with the Bank’s procurement guidelines. All contracts, with minor exceptions, were signed and completed before project closing. At appraisal, the procurement capacity of the implementation agency was assessed and a Procurement Plan was prepared. The procurement activities were managed by the related project implementation agencies. Procurement under the Project was rated Satisfactory throughout most of the Project implementation period, with very few exceptions, where it received a Moderately Satisfactory rating due to procurement and contract delays. However, it was noted in two aide-memoires (December 2010 and April 2015) and interviews that procurement and contract management were at times affected by a lack of proper monitoring and supervision by the implementing agency which, in turn, affected fiduciary compliance. The project experienced substantial time and cost overruns largely due to unrealistic estimates, and poor performance in works that showed premature pavement failures on three packages under the national roads component of the parent Project. At the end of completion of the original Project in December 2010, about $10 million was financed by the government’s own funds due to additional works that were carried out without the Bank’s agreement, and which were found to be ineligible due to unacceptable variations in price indices for the calculation of price adjustments and the use of unacceptable market rates. Financial Management 50. The performance of the implementing agencies in financial management (FM) has been satisfactory in a majority of the implementation support review missions. On two occasions, during November 2011 and December 2013, the overall FM rating was downgraded to Moderately Satisfactory. The reasons for the downgrade was primarily due to one or more of the following: inadequate FM staffing, issues in submission of interim unaudited financial reports, inadequate internal audits, delayed external audit, and inadequate counterpart funds flow on a timely basis. However, other than on these two occasions, basic arrangements handled at the PMU, Project Consultancy Unit, and RMTF level, including budgeting, financial reporting, external auditing arrangements, accounting, FM staffing, funds flow, and internal controls, did not have major issues and hence, the overall implementation was managed satisfactorily. It should be noted, however, that during the last few years of implementation, the Project’s internal audit failed to perform at its optimum level due to internal audit staffing constraints that were prevailing at the ministerial level. This was identified as a systemic issue that could affect any project going forward. It is envisaged that this area could be given special focus and perhaps training might be provided to enhance the capacity to perform internal audits. 2.5 Post-completion Operation/Next Phase 51. The project roads from the parent project and Phase 1 of Component 1 of the Second AF have now been in operation for some years. Despite cracks detected in pavements of three road - 16 - sections (parent project) and surface irregularities in pavements of a Phase 1 section, maintenance is under way on these roads. The projects of Phase 2 of Component 1 of the Second AF have been taken over and post-project activities such as issuance of Taking over Certificates and Defect Liability periods have been completed. 52. An annual maintenance plan is being developed by the RDA and approved by the RMTF for periodic maintenance and routine maintenance of roads. The funding is channeled through the RMTF using budgetary grants. During 2005 to 2013, the budgetary allocation to road maintenance through the RMTF increased from $13 million to $63 million. However, more sustainable avenues for funding need to be explored and an output and performance-based contracts system may be implemented for rehabilitation and maintenance of projects for longer duration. The project has been transformational in its attempts to address the backlog of maintenance due to years of neglect. In addition, this project helped both the Bank and the RDA undertake several institutional changes and reforms, such as financing the piloting of several initiatives to address adverse issues within the sector, including development of the contract management capacity within the RDA. 53. Based on the experience of the RSAP, the Bank is considering another project for long- term development of the capacity of the RDA as well as the maturity of the construction industry with renewed focus on good governance. Given the country’s history of inadequate maintenance of existing assets and the Bank’s strong support for maintenance reflected through its entire support to the road sector in Sri Lanka, an asset management approach is being proposed through design-build-maintain-operate-transfer methodology. 4 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Rating of Relevance of Objectives: High 54. The project objectives at appraisal were highly relevant to Sri Lanka’s needs and in line with the priorities of the CAS for FY2003–06 that aimed to improve social and economic inclusion of the poor and facilitate broad-based economic growth. Both the government and Bank strategies highly prioritized the development of road infrastructure and recognized its contribution to the development of the private sector and the desired increase in revenues to achieve higher levels of public investment. Improvement of the road network to maintainable conditions was considered a necessary condition to sustain the expected surge in economic growth. 55. It cannot be overemphasized that civil conflicts had severely impacted the country for a number of years and had contributed to the poor condition of the road network. Its rehabilitation was essential to the country’s recovery process. 4 A new project with many of these features was approved on May 9, 2016. - 17 - 56. The project objectives were and continue to be highly relevant and fully aligned with the current FY2013–16 CPS, which identified three areas for the World Bank Group’s engagement to assist Sri Lanka in addressing its long-term strategic and structural development challenges and middle-income country agenda. The CPS’ areas of focus are linked to the three central goals of the government's Mahinda Chintana vision, namely: (a) facilitating sustained private and public investment; (b) supporting structural shifts in the economy; and (c) improving living standards and social inclusion. The Bank supports Sri Lanka’s transport sector with the objective of improving the quality and sustainability of roads. 57. The project was linked to the third central goal of the government’s Mahinda Chintana vision and the third pillar of the CPS. The third goal was to ensure access to basic, related services and provide improved quality of services, and the Project was fully aligned with this objective. The project supported the government’s efforts to improve connectivity and enhance competitiveness through lowering of transportation costs and providing better and all-weather accessibility to related services for 80 percent of Sri Lanka’s population that live in rural areas. The improved level of services resulted in enhanced highway safety, reduced travel time, and lower transportation costs, which resulted in more social inclusion and improved living standards. Rating of Relevance of Design: Modest 58. The project design had a balanced mix of civil works and institutional strengthening components which were closely aligned with the government’s reform priorities. The lending instrument was Specific Investment Lending (SIL) and was found to be the appropriate instrument for the Bank to re-engage in the Transport sector after being absent for decades. The Project’s design recognized the need to lower transportation costs and the sustainable delivery of a road system, which is an appropriate description of the objectives. 59. However, as mentioned previously, the M&E framework did not have an adequate mechanism to attribute the impact of these activities directly to the project which raises attributional questions. Given the scope of the project and its components, the formulation of the PDO was overly ambitious. In addition, the establishment of some baselines and target values were shown to be problematic because of weaknesses in the data which was not fully resolved until the second additional financing. 60. While the above weaknesses should not be understated, and are the direct cause for the Modest rating for Relevance of Design, it should not be lost that the project design had some highly relevant and innovative features that have had, and continue to have, a transformative influence on road maintenance, contracting, safety, and financing in Sri Lanka. 3.2 Achievement of Project Development Objectives Rating: Substantial (RSAP-1) and Substantial (RSAP-2) 61. The project was initially scheduled to close on September 30, 2011, although two additional financings and several restructurings extended the closing date to June 30, 2015. In addition to the closing date changes, there were significant changes to the results framework. With this in view, the ICR used a split evaluation as outlined in the ICR Guidelines. As the primary demarcation over the life of the project was the second AF, which was a scale-up, and which - 18 - introduced new activities and significant changes in the results framework, the project has been evaluated based on this milestone. The two phases comprise RSAP-1, which includes the parent project up through the first AF; and RSAP-2, which includes the project from the second AF through its closing. 62. The PDO was to reduce the transport costs through sustainable delivery of an efficient national road system comprising a network of 12,000 km. While most of the civil works were limited to about 6% of the network roads, the project counterbalanced the limited civil works with a broad range of institutional activities that included capacity building and financial support to the Road Maintenance Trust Fund. Support funded improved network maintenance, promoted greater transparency in the allocation of resources and disbursement of funds, and carried out efficient monitoring of the maintenance programs. Between 2005 and 2013, the allocation for road maintenance channeled through the RMTF increased from $13 million to $63 million. 63. Support was also provided for the establishment of a Road Asset Management System, which was a key component of the second additional financing aimed at building capacity and strengthening systems within the RMTF and RDA relating to road maintenance, and geared to better maintaining and prioritizing the country network of national roads. Funding was provided for periodic maintenance, training of staff, developing the right tools and developing a strategy for moving from a post-conflict system of historically poor road maintenance to a modernized and competitive working national system. 64. The Project PDO and some of the associated outcome indicators were not ideally formulated to align with the project scope. As mentioned in the M&E section, this weakens certain causal links. At the same time, discounting achievements because of poor formulation risks discounting substantial outcomes. First Objective - Lowering Transport Costs 65. Lowering transportation costs has been monitored and evaluated through outcome indicators via reducing average network VoCs, average network roughness, share of network roads in poor and bad condition, and maintenance expenditure for project roads. Taking into account that each indicator is split by the multiple implementation periods, the project achieved, and in many cases exceeded, the targets of the large majority of these indicators. Reduction in VoC and Roughness 66. Reduction in VoC was fully achieved for RSAP-1. The target was a 3.8 percent decrease; the actual decrease was closer to 35 percent. As mentioned above, however, the result should be interpreted with caution as the VoC was overestimated at the appraisal stage of RSAP-1. For RSAP-2, reduction in VoC was not fully achieved. Although the decrease achieved about a 5 percent reduction (improving upon RSAP-1’s target), the target had been reset in the 2nd AF from 3.8 percent to 28 percent. In contrast, the roughness (IRI) of the Project roads was decreased substantially and the target was fully achieved. This duel result suggests that despite significant improvements in road quality there was only a modest reduction in VoC. This could be due to increased traffic on the roads due to the road improvements, or increased speeds of vehicles on the roads, resulting in higher fuel consumption (or both). The Borrower has suggested, although it - 19 - awaits confirmation, that subsequent to the cessation of the conflict in 2009, the amount of funding to the road sector increased substantially, but was disproportionately targeted to rebuild the network in areas that had suffered severely from the conflict. However, the road network in the conflict affected areas does not carry as much traffic as the road network in the urban areas, and thus, fewer users would result in a smaller impact on the network’s VoC. 67. The achievement of the first objective was also linked to project outputs measured by intermediate indicators, namely, to lowering transportation costs through resurfacing of national roads, improving rural roads, increasing of periodic maintenance, and increasing annual maintenance expenditure, which were almost all achieved, and in several cases, exceeded. Second Objective – Sustainable delivery of an efficient national road system Increase in Annual Maintenance Expenditure 68. The achievement of this second objective was linked to project outputs that were implemented successfully, including: capacity building, the establishment of the RMTF, the rural roads maintenance pilot that included the development of a rural roads maintenance strategy, country rural roads investment plan and database; safety capacity assessment; and a 5 percent reduction in traffic accidents in the selected demonstration corridor as a result of measures supported by the Project. 69. The expenditures of $0.66 million and $19 million for institutional strengthening in (RSAP-1 and RSAP-2, respectively) resulted in building capacity for data collection and analyses, design of enhanced decision support systems, better internal organizational management, and additional support for maintenance financing for periodic maintenance work. 70. As noted under the first objective, though with different emphasis, the project supported the establishment of RAMS and the RMTF, two of the highlights of the project which have had a transformative impact. RMTF was set up to ensure the sustainable and transparent allocation of resources and funds, but also carries out efficient monitoring of road maintenance, while RAMS aids RDA in building capacity to optimize the selection of roads for maintenance and align its processes toward minimizing costs. A related notable achievement has been the strengthening of the capacity of local governments to prioritize road expenditures and select appropriate technical interventions for rural road maintenance. The Project substantially achieved the intermediate indicators that assess this objective. 71. The outcome indicator related to sustainability was to monitor the funding levels for yearly maintenance for periodic and routine maintenance works. The original target for RSAP-1 was substantially achieved; $44 million (up from $10 million) against the target value of $46.3 million. The target for RSAP-1I was not achieved, LKR 5 billion against the target of LKR 6.6 billion. The indicator, however, does not take into account that GoSL has increasingly channeled funds for periodic maintenance through other projects not reflected under the budget item for RMTF. In addition, GoSL channels funding for improvements which include periodic maintenance expenditure which cannot be easily separated as a result of streamlining budgetary processes, and recently moving to a multi-year budgetary system. - 20 - 72. In sum, although the results framework had inherent weaknesses, the project accomplished most of its major objectives: resurfacing, rehabilitating or maintaining over 800 km of roads, developing budgetary systems that are generating institutional changes and new approaches to road maintenance and financing; innovating in programs for road safety; and piloting new methods and strategies for rural roads maintenance. Taken together, the cumulative impact of these activities, together with the achievement of most of the project targets, suggests the project had a major impact on the improvements to the national network. Quoting from the Borrower’s ICR: “The culture change that RSAP catalyzed that ranged from the improved practice of safeguards and an effective grievance redress system to better efficiency in all aspects of project design and implementation will be key to the sustainability of its outcomes. The setting up of the RMTF and maintenance systems, the building of capacity within the RDA, including the capacity to manage contracts, and the development of the private contracting industry, are all key factors that influence sustainability.” The efficacy of the project is rated Substantial. Table 4. Assessment of Efficacy Pre- and Post- Restructuring Table 1. Assessment Targets Actual Values Percent Achieved of Efficacy Pre- and Post- Restructuring PDO Indicator Aggregate Rating of Efficacy: Original Efficacy – Substantial ; Revised Efficacy - Substantial Specific Objective 1: Main objective - Lower Transport Costs through sustainable delivery of an efficient national road system Rating (against revised target): Substantial RSAP-1 Original: Actual value: 14.97 Original: Achieved (i) Reduction in average LKR 22.9 per The value achieved in 2006/08 was 15.37/km network VoC by 3.8%, km and the mission considered VoC to be very low and after examination of the estimation, it was found that the methodology was unknown (AM-June 2008 (ii) Reduction in IRI in Original: 8.39 Actual value: 6.45 Original: Achieved average network Similar as above, the survey data collected in 2004 was being used. It was found that only 226 km out of 9913 km of national highways were surveyed for IRI in the year 2004 or before. Therefore the source of the information of the baseline year 2004 of IRI 9.5 could not be known. (iii) Reduction in Original:35% Actual value: 44% Achieved: 70% percentage of the network in poor and bad condition (iv) Increase in annual Original: Actual value: 44 Achieved 95% maintenance expenditure $46.3 million - 21 - RSAP-1I (i) Reduction in average Original: Actual value: 13.88 Original target of 3.8% achieved; revised network VoC LKR 10.5 target of 28% not achieved (ii) Reduction in IRI in Original: 5.9 Actual value: 5.9 Achieved average network (iii) Reduction in Original: Actual value: 35% Achieved percentage of the network 35% in poor and bad condition (iv) Progressive increase Original: $66 Actual value: $50 Not achieved in annual road million million maintenance (v) Improved level of Original:38.3 Actual value: 68.15 Achieved satisfaction of road users 0 Intermediate outcomes RSAP-1 Original: 620 Actual value: Achieved (i) # of KM of National 617km roads resurfaced (ii) reduction in IRI for Original: 2.9 Actual value: 2.8 Achieved project roads (620km) (iii) Reduced travel time Original: 42 Revised (split) Achieved on project roads (620km) (iv) # of km of Rural Original: 635 Actual value: Original: Revised from spot improvement to Roads improved Revised: 157 167km rehabilitation due to rural strategy. Revised: Achieved (v) Reduced travel time Original: 4% Actual value: 60% Achieved on project rural roads by 10% (vi) Relevant staff to be Original: Actual value: 67% Achieved trained in road asset 50% management RSAP-1I Original: 134 Actual value: Revised: Achieved (i) # of KM of National Revised: 78 78 roads resurfaced (ii) reduction in IRI for Original:2.8 Actual value: 2.33 Original: Achieved project roads Revised: 2.4 Revised value: 2 Revised: Achieved (iii) Reduced travel time Original: 42 Actual value: 40 Achieved on project roads Revised: 12.5 Actual value: 11.31 Achieved Revised: 23.9 Actual value: 23.58 Achieved (iv) Periodic maintenance Original: Actual value: 64.50 Achieved 60km - 22 - 3.3 Efficiency Rating: Modest (RSAP-1) and Substantial (RSAP-2) 73. The national roads sections rehabilitated under the parent project were selected on the basis of development priorities and economic viability and assessed during appraisal using the Highway Development and Management Model (HDM-4) software. The net present value (NPV) of the investments in the 15 original road sections was estimated at $103 million and the economic internal rate of return (EIRR) at 31 percent. Sensitivity analysis showed that changes in costs and benefit scenarios would have limited impact on the results. The analysis of the ICR is conducted for 12 road sections. After project completion, the NPV is estimated to be $225 million and the economic internal rate of return (EIRR) is 29 percent. This difference in the NPV values (between the PAD and ICR) is driven by a difference in the assumptions of the PAD and ICR—the analysis in the PAD is conducted for an 11-year time horizon while the analysis of the ICR is conducted for a 20-year time horizon. If we limit the time horizon of the ICR analysis to 11 years then the resulting NPV is $23.6 million and EIRR is 17 percent. This reduction in EIRR and NPV is consistent with the increase in construction costs between the initial economic analysis and the revised analysis. With cost overruns ranging from 51.4 percent to 132.4 percent, the efficiency of this phase of project, that is, RSAP-1, is rated as Modest. 74. In the roads financed under the Second AF, the economic analysis was carried out for both sections using HDM-4 by the RDA. Here we compare the results for the ICR to those in the Restructuring Paper. The estimates of NPV and the EIRR for the K-T section, A002, and A003 increased after project completion. For the KT section, the NPV and EIRR at the time of restructuring was $24 million and 21 percent, respectively—the NPV is $8 million higher and the EIRR is 3 percentage points higher after project completion. For A002, the NPV and EIRR at the time of restructuring was $101 million and 29 percent, respectively—the NPV is $28 million higher and the EIRR is 10 percentage points higher after project completion. For A003, the NPV and EIRR at the time of restructuring was $132 million and 47 percent, respectively—the NPV is $14 million higher and the EIRR is 4 percentage points higher after project completion5. With cost overruns ranging from negative to 20.8 percent, the efficiency of this phase of project RSAP- 2 is rated as Substantial. 75. Table 5 summarizes the main results, comparing the EIRRs before and after the Project. All EIRRs after project completion remain above the 12 percent threshold. For both RSAP-1 and RSAP-1I the benefits from travel time savings are higher than the benefits from vehicle operating cost savings. 5 The analysis at both the appraisal and ICR stage is conducted using HDM4 which is calibrated for Sri Lanka, it is based on the same time horizon and accounts for the same benefits (VoC and VoT). - 23 - Table 5. Results of Economic Analysis-Before and After the Project EIRR at NPV at appraisal EIRR at ICR NPV at ICR Length Appraisal ($, millions) (%) ($, millions) (%) 620 km(using different time 31 103 29 225 horizon) 620 km (using same time 31 103 17 24 horizon) Before Restructuring 134 km 51 522 Dropped and replaced with A002, A003 (A006) After Restructuring K-T Section 21 24 24 32 A002 29 101 39 129 A003 47 132 51 146 76. Final results were heavily dependent on the soundness of input data and quality of analyses. This remained a matter of concern during project life. The Bank mission highlighted the issue of quality of data and more need of analyses to apprehend the scenarios better.6 Therefore, the ICR team has taken cost overruns into account as well for assessment of efficiency. The project's overall efficiency was rated as Modest (rounding down the efficiencies of two phases of the project because the parent project is given more weightage due to more cost). 3.4 Justification of Overall Outcome Rating Rating: Moderately Satisfactory 77. The ICR considers the overall outcome of the Project as Moderately Satisfactory. The rating is based on the relevance of objectives, relevance of design, achievement of PDOs, and efficiency, and using the weighted system, before and after the Second AF, which was a major change in scope of work. The related indicator values during project life were rounded off to the nearest whole number. The assessment is given in Table 6. 6 AM of mission in December 2013. The mission discussed the shortcomings in development of asset management system in the RDA, highlighting data management issues. - 24 - Table 6. Assessment of Overall Project Outcome Against Original Against Revised Outcome Outcome Indicators Indicators before after 2nd 2nd Restructuring Restructuring Overall Comments Relevance to 1 Substantial Substantial – – objectives and design Achievement of 2 Substantial Substantial – – project outcome 3 Efficiency Modest Substantial – – Moderately 4 Overall rating Satisfactory – – Satisfactory 5 Rating value 4 5 – – Weight (% disbursed 6 before/after 2nd 57 43 – – Restructuring) 7 Weighted value 2.28 2.15 4.43 Rounded to 4, which arrives at Moderately 8 Final rating – – Moderately Satisfactory Satisfactory 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 78. The project is expected to have had an indirect positive impact on poverty and social development—improving transport conditions fuels the economy, creating jobs, and access to markets. Improved transport conditions facilitate better connected communities, foster access to social facilities, and reinforce social nets. Poverty impact and social development were higher- level objectives of the Project success and have not been specifically assessed. (b) Institutional Change/Strengthening 79. The project collaboration has had important benefits in driving the institutional reform agenda for road asset management. The Project financed several courses to improve the capacity of the RDA and the construction industry in areas of construction capacity, maintenance, asset valuation, preservation, highway safety, environment and social safeguards, FM, and procurement management. 80. The project made significant contributions to institutional strengthening of Sri Lanka’s transport sector. In particular the Project:  Provided technical assistance for the strengthening of procurement, planning, and contract management.  Improved the capacity to resolve social and environmental safeguards issues by training RDA and other implementing agencies,  Provided capacity building in the functionality of the RMTF, - 25 -  Strengthened RDA to design and supervise consultants’ and contractors’ work including laboratory tests, etc.,  Strengthened RDA to work with other inter agencies such as utilities, water body, police etc, and  Supplied technical Assistance in Asset Management, HIMS, PMS, HDM-4, inventory, selection, prioritization, database for rural roads. (c) Other Unintended Outcomes and Impacts (positive or negative) 81. There were no noticeable or irreversible negative impacts of this project. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 82. The ICR team carried out a workshop with stakeholders during the mission of June 8-12, in order to obtain their impressions and lessons learned. The workshop was attended by a large audience from the various government agencies, PMU, contractors, consultants, and beneficiaries who provided valuable input for the ICR. This workshop sought to capture the lessons learned to improve the design of future transport projects in Sri Lanka. The findings of this workshop were rich and are presented in annex 6. 4. Assessment of Risk to Development Outcome Rating: High 83. The joint strategy by the GoSL to rehabilitate the national road network along with the rural road project was supported through a collaborative effort of the Bank, ADB, and JICA that agreed on funding the maintenance backlog which affected the road network. Sustainability of the road network largely depends on adequate maintenance; thus, the focus of the Project on supporting the RDA in developing capacity in assets management and reducing the backlog in maintenance was critically important. Proper maintenance of the road assets remains one of the priorities in the transport sector. 84. The risk to the development outcome is High. It is clear that through persistent efforts and consistent programs, GoSL has managed to improve the road network during the past 10 years. It has continued gradual progress in strengthening road management at the national and rural levels, brought increased transparency to the allocation of resources through the use of the RMTF, and is working to put in place the right institutions and policies to secure continued inroads in traffic safety. However, there is a significant risk to relaxing the level of funding for road maintenance as maintenance budgets for roads infrastructure may be crowded out by large investments in new construction and for works stated in this section. 85. Political will to invest in the maintenance of roads is critical. In this case, it was witnessed that with the change of regime, the interim government reduced the budgetary allocation for road maintenance from LKR 6.6 billion to LKR 5 billion. - 26 - 86. From a broader perspective, the financing of the sector continues to raise doubts in its capacity to meet the actual needs of the road network with regard to rehabilitation and maintenance. Although the increases in road maintenance budget allocations made it possible to partially catch up with the road maintenance backlog, the current budget is insufficient to mitigate road deterioration in the coming years.7 This means that any additional resources allocated to the road rehabilitation and maintenance will reduce the extent of the expected deterioration rather than contribute to absolute improvements on the road network. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 87. The Bank's performance during the preparatory period was Moderately Satisfactory. Overall, the Project was well prepared and reflected the government’s priorities at the time of the appraisal. The project design included: (a) the incorporation of most lessons learned; (b) the pragmatic approach for dealing with unknowns in the rural roads sector; (c) identification of most of risks early in the Project; and (d) strong attention to safeguards. 88. The Bank's performance, with regard to the strategic relevance and approach and paying adequate attention to fiduciary and safeguards arrangements was satisfactory. The Bank team was comprised of experienced professionals who worked with the GoSL and development partners on project preparation. The project was also designed based on the GoSL’s development strategy and in the framework of the CAS and donor transport strategy. In this respect, there was a strong Bank effort to coordinate its interventions in the transport sector with ongoing efforts from other international financial organizations, particularly ADB and JICA. 89. The project was designed taking into account the insurgency in Sri Lanka and its impacts, which exacerbated the lack of available vital data at entry, the nonexistent or limited capacity of the implementing agency in RAMS practices, the relative inexperience of the local private construction industry, and the need to address the country’s huge backlog of road maintenance through a systematic and programmatic approach. The urgency of the crisis made moving quickly a high priority, thus limiting the amount of time for preparation. The team handled most of these challenges well. However, as explained in previous sections, the PDO and associated indicators were overly ambitious and disproportionate to the Project scope, particularly for a post conflict affected state. The Results Framework was inadequately designed, and in addition there were 7 Final Book; Directions and Destinations by World Bank Group Sri Lanka. Page 6; “road maintenance has been chronically under-funded across the different types of roads. The total funding requirement for periodic and routine maintenance of roads from 2007 to 2013 was approximately LKR 81 billion, but only LKR 38 billion was made available during the period.” - 27 - some causal disconnects between inputs, outputs and outcomes which made it difficult to assess the contribution and achievement of the Project activities as stated in the PAD. 90. Keeping this in mind, the quality at entry is rated as Moderately Satisfactory, although QAG’s rating was satisfactory. (Annex 9). (b) Quality of Supervision Rating: Moderately Satisfactory 91. The Bank provided constant and appropriate implementation support to the borrower and took proactive action on issues that arose during implementation. Clear examples of this have been the prompt attention to process two AF operations, explained previously in this ICR. The operation took into account the priorities of the borrower and the need to ‘fine tune’ the design of the project, incorporating a focus on road traffic safety issues and tailoring the design and supervision of civil works to the specific needs of Sri Lanka. The timely processing of the 2nd AF was pivotal in nature to ensure continuity of the Bank's engagement in the country with already well performing implementation arrangements. The Bank also provided special training on the use of HDM-4, procurement, contract management, social and environment management to RDA. 92. The Bank team carried out at least two implementation support missions per year. The field-based Task Team Leader and other team members worked directly with the client and ensured timely responses. Continuity in the team was maintained from preparation through completion. Implementation support extended beyond operational missions. The task team possessed the appropriate skills mix with the technical, institutional, fiduciary, and safeguards knowledge. The Task Team had the experience to work effectively with the client, review documentation and provide advice on the quality of works. The team recommended new approaches in the implementation of institutional support components which was appreciated by the client. This resulted in an excellent integration of procurement supervision activities with the overall supervision effort, and for the consistency, clarity, and comprehensiveness of documentation of safeguards supervision. 93. In addition, the Bank helped to introduce and include environmental clauses as part of the contract documents of rehabilitation works. Because of this project, many of the local contractors adopted the same preventive and remedial environmental protection measures on other non-Bank financed projects that previously had had no such requirements. Therefore, the Project was instrumental in elevating the awareness of using high standards of safeguards practiced by several contractors. 94. The Bank Team, however, needed to have updated the M&E Results Framework more vigorously and on a more timely basis. As the weakness of the data became apparent in the early years of the project, efforts should have been made to prioritize data collection and analysis. If it became clear that data was not going to become available, then baselines and targets (and perhaps, objectives) should have been reassessed. While the results framework was revised during the second AF, this was five years into the project. In addition, at no time during project implementation – including the second AF, or the restructuring in 2013 – was the issue of attribution ever considered. If addressed earlier, indicators/objectives affecting attribution could - 28 - have been revised, creating stronger links to the PDO, and bringing a more robust coherence to the results framework. (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 95. The Bank team put substantial effort in project preparation and implementation support, which led to the Project (mostly) achieving its development goals. The implementation support was provided amidst the escalating conflict where in certain parts of the project period the particular roads could not be accessed due to terrorist activities. The Bank, albeit with some shortcomings in design and supervision, achieved implementation of a project with complex civil works, challenging safeguards features, and the need to manage a large multidisciplinary team, all that in a pre and post conflict state for decade. 96. Therefore, considering the ratings for the Bank's performance at entry and quality of supervision Bank performance is rated Moderately Satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Moderately Satisfactory 97. The GoSL remained fully supportive of the efforts for smooth implementation of the civil works component of the Project, the expansion to the whole country of the strategy for rural roads maintenance, and the successful implementation of a pilot traffic safety effort. The GoSL was instrumental in reshaping the road maintenance regime in the country; its commitment made it possible to achieve a majority of the performance indicators. 98. Shortages in counterpart funds were evidenced at different points of project implementation; however, the government was able to provide the funds at a later stage. Actions toward fully launching the RMTF dragged on during the first three years of the Project but with continuous engagement of the Bank's Task Team and approval of the Second AF, the fund became operational, and the GoSL was able to establish the RMTF. Maintenance allocations during the life of the Project were mostly in line with the agreed amounts, and the output indicator related to maintenance expenditure was mostly achieved. 99. The Government's performance is rated Moderately Satisfactory due to delays in the release of counterpart funds and delay of a fully functioning RMTF, which impeded the implementation of the maintenance component. - 29 - (b) Implementing Agency or Agencies Performance Rating: Moderately Satisfactory 100. The project was implemented by a dedicated PMU. Despite initial problems with staffing due to internal regulations affecting outside recruitment and a rather high rotation of staff because of the then ongoing construction boom and high demand for skilled professionals, the PMU was helpful in carrying out the implementation of the Project related activities. The PMU had adequate reporting systems, submitting bi-annual progress reports to the Bank that contained updated information on project financing, procurement, physical progress of works, safeguards, and key implementation issues. 101. However, the implementing agency failed to provide key technical performance data (roughness, condition, and traffic survey report) which was reflected through ISRs and AMs. The provision of this data on a yearly basis was a covenant. 102. The main weakness of the PMU and the RDA related to contract management and close control with respect to growing variations which were found to be ineligible. For instance, physical variations on the 12 civil work contracts ranged from 7% to 88%. The cost increase by the end of the project was 34% more than the original contracts cost and 11% over the variations at MTR which stood at LKRs. 3.9 billion. The implementing agency showed a lack of proactivity in the use of remedial actions to address the issue of underperforming supervision consultants. The project faced cost and time overruns throughout project life. Many quality-related issues were highlighted and were attributed to non-satisfactory performance of the supervisory consultant, poor designs, and engineering practices. Premature failures in pavements in three road sections were witnessed. 103. The PMU and RDA could have avoided many of these issues through better monitoring and supervision and proper contract management. Overall, the implementing agencies performance is Moderately Satisfactory. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 104. Based on the above considerations, the Borrower’s performance is rated Moderately Satisfactory. 6. Lessons Learned 105. The long-term and continued Bank’s engagement and coordination ef forts in the country’s road subsector provides a robust platform for well performing projects. The Bank’s continued presence in the transport sector helped to strengthen and enhance supervision effectiveness and deepen understanding issues in Sri Lanka’s road subsector. The Bank has been active in the subsector since 2004 with consistent objectives. This continued presence has allowed for the strengthening of the transport sector across the board through the introduction of innovative road stabilization programs and sound periodic and routine road maintenance policies. - 30 - 106. Extensive turnover of the key RDA project team weakens the institutional memory. Continued RDA key staff is pivotal for a project's success. Continuity of the same project team and task team leader of the Bank from the preparation to completion helped achieve most of its PDOs. The high turnover of many project team members of the RDA (four project directors were changed during the Project) resulted in many problems related to cost and schedule overruns and quality of civil works. 107. A dedicated project management consultant will help both the Bank and the Borrower to have more effective control on project implementation. The RSAP project suffered major cost overruns and issues with quality and risk management. These issues are related to project management implementation. In the case of the RDA, the authority was assisted by supervisory consultants and individual consultants but close investigation of their terms of reference indicated that these consultants were subject to area specialists (engineering, finance, etc.) and they were given additional segmental project management activities. In the era of project management role delineation, the engagement of a dedicated project management consultant on big projects will help the Borrower in their project management effort. 108. Adopting sustainable funding sources for the RMTF is key for its optimum functioning. Though setting-up institutional arrangement for the creation of a road fund seems simple, implementation of associated Policy and Legislative Frameworks throughout the sector becomes difficult. Nevertheless, autonomous arrangement of the RMTF and road funds provide an opportunity for progress on management, accountability, transparency, and increased awareness on the need to address long-neglected maintenance. Whereas the concept of highway maintenance and development financing through marginal social costing is very popular and remained the essence of this project, another revenue stream from road user fees might have added to the much- needed additional source of funding. 109. Formulating a proper M&E system (with improved and adequate Results Framework) is instrumental for the evaluation of project. The choice of the PDO indicators should take into account the type of data that can be collected during implementation. The Bank should ask the Borrower to provide reliable data with the frequency stated in covenants. The baseline and target values should be based on reliable data, should be realistic, and should be based on sound engineering. 110. Importance of RAMS practices is key. One of the achievements of this project was to highlight the importance of a comprehensive system for maintaining the road network through RAMS practices. RAMS not only helped in reducing the transportation costs and increasing public perception about the importance of maintenance but was also instrumental in supporting the national economy through better and sustainable connectivity and mobility. Further enhancements in the current RAMS will help the transportation sector manifold in the country. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies - 31 - 111. The Borrower’s ICR, as well as their comments on this document are included in Annex 7. Any corrections and/or other changes that have been suggested have been incorporated where appropriate and where evidence was available to substantiate recommended changes. (b) Cofinanciers. Not applicable. (c) Other partners and stakeholders. Not applicable - 32 - Annex 1. Project Costs and Financing (a) Project Cost by Component (in $, millions equivalent) Appraisal Revised Actual/Fi Percentage Percentage Estimate Estimate nal Components of of Revised (Original) ($, ($, Appraisal Estimate ($, millions) millions) millions) Component 1 of Parent Project: Maintenance and Rehabilitation of National 112.51 179.15 239.00 78.06 49.23 Roads Component 2 of Parent Project: 10.04 10.04 13.07 6.97 2.76 Maintenance and Rehabilitation of Rural Roads Component 3 of Parent Project: 0.60 0.60 0.66 0.42 0.16 Institutional Strengthening and Policy Support Second AF: 81.00 63.06 Rehabilitation of Priority National Roads and – 0.00 27.48 19.00 12.07 Maintenance Finance Total Baseline Cost 123.14 363.87 327.86 85.44 79.64 Unallocated (physical and price contingencies) 20.98 74.08 0 14.56 20.36 Total Project Costs 144.12 363.97 327.86 – – (b) Financing Actual/Latest Appraisal Estimate ($, Percentage of Percentage of Source of Funds Estimate ($, millions) Appraisal Actual/ Latest millions) Borrower 44.12 54.63 12.89 17.64 IDA 298.00 255.00 87.08 82.36 Total 342.22 309.63 – – (c) Project Cost by RSAP-1 and RSAP-1I 2005 2008 Components 2010/11 Appraisal Estimate Annex 1 of PP RSAP-1 Actual/Final (Original) Revised Estimate (Original + First AF) ($, millions) ($, millions) ($, millions) Maintenance and Rehabilitation of National 112.51 179.15 239.00 Roads Maintenance and Rehabilitation of Rural 10.04 10.04 13.07 Roads Institutional Strengthening 0.60 0.60 0.66 and Policy Support Total Baseline Cost 123.14 189.79 252.73 - 33 - Unallocated (physical and 74.08 20.98 0.00 price contingencies) Also includes VAT **Total Project Costs 144.12 263.87 252.73 2011 2013 2015 Components Appraisal Estimate Annex 1 of PP Actual/Final RSAP-1I (Second AF) (Original) Revised Estimate ($, millions) ($, millions) ($, millions) Rehabilitation of Priority 85.50 81.00 63.06 National Highways Maintenance Finance and 14.50 19.00 12.07 Institutional Strengthening **Total Project Costs 100.00 100.00 75.13 ** Total (Original +First 244.12 363.87 327.86 AF+Second AF) (d) Financing per Credit and Share Appraisal Estimate *Actual Latest Source of Funds ($, millions) ($, millions) CR 4138 Borrower 44.12 54.63 IDA 100.00 96.40 CR 4429 Borrower 0.00 0.00 IDA 98.10 83.60 CR 4906 Borrower 0 0 IDA 100.00 75.00 Total Borrower 44.12 54.63 IDA 298.10 255.00 Grand Total 342.22 309.63 Note: *Current value for client connection. - 34 - Annex 2. Outputs by Component Status at Output Indicators Actual Achievement % Closing 1st Revised 2nd Original Baseline Target Revised Target Second AF Target Component 1. Maintenance and Rehabilitation of National Roads RSAP-1 (Parent Project and First AF) National roads resurfacing and 0 620 – – 617 Successfully achieved target for national roads improvement (km) Reduction in IRI for project roads 7.8 2.9 – – 2.8 Successfully achieved RSAP-1I (Second AF) National roads resurfacing and Successfully achieved revised target for national 0 134 – 78 78 improvement (km) roads Reduction in IRI for project roads A6 6.0 2.8 2.33 A2 4.3 – 2.4 1.9 Successfully achieved A3 3.1 2.4 2.0 A6 66 42 40 Reduction of travel time A2 13.8 – 12.5 11.31 Successfully achieved A3 26.5 23.9 23.58 Component 2. Maintenance and Rehabilitation of Rural Roads RSAP-1 (Parent Project and First AF) Rural Roads resurfacing and Achieved. Formally revised target was reduced 0 635 – 157 162.3 improvement (km) and restricted in 2013. Reduction of travel time (%) 10 – – 60 Successfully achieved Component 3. Institutional Strengthening and Policy Support RSAP-1 (Parent Project and First AF) Establishment of and fully operational bodies such as the – – – – – Successfully achieved RMTF, Board Trustee, TAC, and Trust Fund Secretariat 50% of relevant staff in the Successfully achieved Planning and Maintenance of RDA trained in road asset management Training on Asset management, HDM-4, 0 50% – – 67% Maintenance Manual, Contract management, Social and Environment Management, PPP’s in highways - 35 - Rural road Maintenance Successfully prepared and delivered and used. Strategy/Manual Rural Road Database established Successfully established and used for the whole country. Out of 342 Local Authorities to be covered, the road databases and investment land for226 Las have already been prepared RSAP-1I (Second AF) Annual road maintenance program New Approved – – Approved Successfully achieved approved by the RMTF About 60 km of national roads on which periodic maintenance has New 60 – – 64.50 Successfully achieved been carried out Rural road Strategy/Manual Note: A2 =; A3 =; TAC =; - 36 - Annex 3. Economic and Financial Analysis Introduction 1. Economic analysis is performed to assess whether the economic benefits of the Project are higher than its economic cost and whether these benefits are similar to those estimated at the appraisal stage. We accomplish this by first assessing whether the economic internal rate of return (EIRR)8 is greater than 12% and then comparing it to the EIRR estimated at the appraisal stage. Framework Assumptions: 2. The following general assumptions are made when performing the economic analysis:  The analysis is performed for a period of 20 years—the analysis period starts from the construction commencement year since traffic is allowed to use the road during construction.  Project discount rate is assumed to be 12%.  For the conversion rate it is assumed that USD 1 is equivalent to LKR 140.2871  The vehicle classification used in the analysis consists of 10 vehicle. The ten vehicle types considered are medium goods vehicle (MGV), Van (VAN), medium bus (MBU), light goods vehicle (LGV), motorcycle or scooter (MCL), three wheeler (TWL), car (CAR), large bus (LBU), 3 axle truck (HG3), and articulated truck (ART).  To convert financial costs into economic costs a conversion factor of 0.97 is used (based on the 2014 Central Bank data). With vs. Without Project Scenario: 3. The without project case assumes that the current maintenance regime that allows for a gradual deterioration of the road continues. The with-project case entails the actual rehabilitation of the roads and subsequent improvement in road maintenance. It is assumed that there is minimal routine maintenance in both scenarios with a slight improvement in the project case to reflect current government intentions to sustain the conditions of the newly rehabilitated roads. Periodic maintenance works items is assumed to vary with the surface type of the road while routine maintenance works will not. Costs and Benefits: 4. The economic benefits of improved road infrastructure is assumed to stem from two sources: (1) time savings benefits (VoT) and (2) reduction in VoC. For the analysis, traffic volume and growth data for each road was extracted from the relevant Feasibility Study Reports. 8 EIRR is the discount rate that equates the discounted stream of benefits and costs. - 37 - 5. Project costs account for both improvements (road widening) and maintenance (routine and periodic). To account for improvements the whole package is considered as a single section and the cost per km varies for each package. Maintenance (routine and periodic) costs are recurring and apply in both ‘with’ and ‘without’ project with some difference in the work items. These work items depend on the general maintenance policy adopted by the RDA in their normal maintenance works. A key difference in the modeling of the periodic and routine maintenance is that the former is allowed to vary with the surface type of the road while the latter is not. The total capital costs in financial terms for project were Rs. 30,059 Million or $214.3 Million. The breakdown of the cost by contract is presented in table 3.1. Table 3.1: Contract Package Details Contract Final Contract Date of Actual Date of Road Length No. Amount (Rs) commencement Completion WB/RSAP/C Rehabilitation and Improvement on 35.08 km 932,713,752.94 1-Mar-2006 27-Nov-2009 P/01 Ingiriya-Ratnapura Rd WB/RSAP/C Rehabilitation and Improvement on 75.30 km 2,683,861,789.60 15-Mar-2006 8-Dec-2009 P/02 Nittambuwa-Kandy Road WB/RSAP/C Rehabilitation and Improvement on 22.14 km 980,242,714.82 6-Sep-2006 10-Jul-2009 P/03 Bandarawela-Haliela Road WB/RSAP/C Rehabilitation and Improvement on 20.200 km 829,534,965.36 6-Sep-2006 29-Jun-2009 P/04 Bandarawela-Haliela Road Rehabilitation and Improvement on WB/RSAP/C Galle - Deniyaya Road & Deniya 80.09 km 1,918,216,256.91 6-Nov-2006 5-Jan-2010 P/05 Akuressa Rehabilitation and Improvement on WB/RSAP/C Dengama - Mulatiyana Road and 22.98 km 948,393,818.31 21-Mar-2006 6-Jan-2009 P/06 Hakmana - Talahaduwa Road WB/RSAP/C Rehabilitation and Improvement on 24.61 km 971,282,211.38 6-Nov-2006 11-Aug-2009 P/08 Medawachchiya-Punewa Road WB/RSAP/C Rehabilitation and Improvement 69.28 km 1,573,584,156.66 6-Nov-2006 19-Nov-2009 P/09 onWellawaya-Siyambalanduwa Rehabilitation and Improvement on WB/RSAP/C Siyambalanduwa-Ampara-Karathiv 77.83 km 1,763,152,639.85 6-Nov-2006 26-Mar-2010 P/10 Road Rehabilitation and Improvement on WB/RSAP/C Maradankadawala-Jayanthipura 59.90 km 1,897,490,215.49 5-Jan-2007 30-Jun-2009 P/11 Road WB/RSAP/C Rehabilitation and Improvement on 68.66 km 2,286,806,255.81 24-Aug-2007 14-Aug-2009 P/12 Jayanthipura - Trikondimadu Road WB/RSAP/C Rehabilitation and Improvement on 61.37 km 1,768,861,093.87 8-Mar-2007 30-Jun-2009 P/13 Padeniya - Puttlam Road Rehabilitation and Improvements on Ambepussa-Kurunegala- WB/RSAP- Trincomalee Road (A06)from 10.2 km 916,085,524.85 11-Jul-2011 18-Oct-2012 1I /KT/01 157+000 km to 167+280 km & Kantale to Perathuweli (B196) Road Rehabilitation and Improvements on WB/RSAP- Ambepussa-Kurunegala- 10.72 km 867,898,141.30 11-Jul-2011 18-Oct-2012 1I/KT/ 02 Trincomalee Road (A06) from 167+280Km to 178+000Km Rehabilitation and Improvements on WB/RSAP- Ambepussa-Kurunegala- 12.0 km 740,367,191.80 11-Jul-2011 10-Oct-2012 1I/KT/ 03 Trincomalee Road (A06) from 178+000 Km to 190+000Km Rehabilitation and Improvements on WB/RSAP- Ambepussa-Kurunegala- 9.3 km 986,889,478.40 11-Jul-2011 10-Oct-2012 1I/KT/ 04 Trincomalee Road (A06) from 190+000Km to199+300Km Rehabilitation and Improvements on WB/RSAP- Colombo – Galle - Hambanthota 1I/WK/ICB Road (A02) from Maliban Junction to 5.28 km 1,872,510,000.00 7-Jul-2013 31-Aug-2014 01 - Lot 01 Cross Junction (from 13+275 km to 18+480 Km) and AB011 Road from - 38 - Cross Junction to Panadura Bridge (from 0+000Km to 0+080Km) Rehabilitation and Improvements on Colombo – Galle - Hambanthota Road (A02) from Panadura Bridge to WB/RSAP- Nalluruwa Junction (From 25+240 1I/WK/ICB 11.76 km 2,786,250,000.00 7-Jul-2013 30-Apr-2015 km to 30+300 km) and AB011 Road 01 - Lot 02 from Cross Junction to Panadura Bridge (from 0+080Km to 6+780Km) Rehabilitation and Improvements on WB/RSAP- Peliyagoda – Puttlam Road (A03) 1I/WK/ICB 8.4 km 1,701,840,000.00 7-Jul-2013 31-Jan-2015 from Peliyagoda to Mahabage (from 01 - Lot 03 0+600 km to 9+000 Km) the Rehabilitation and Improvements WB/RSAP- on Peliyagoda – Puttlam Road (A03) 1I/WK/ICB 9.3 km 1,633,850,000.00 7-Jul-2013 31-Jan-2015 from Mahabage to Ja-ela (from 01 - Lot 04 9+000 km to 18+310 Km) Analysis 6. Since the various phases of the Project started at different points in time, the economic analysis for each phases is conducted for different time horizons—the economic analysis for RSAP-1 is conducted from 2006 to 2025, the economic analysis for RSAP-1I phase I (sections A2 and A3) is conducted from 2012 to 2031, and the economic analysis for RSAP-1I phase II (section A6) was conducted from 2011-2030. Therefore, we report NPV and EIRR for these three phases separately. 7. Cost benefit analysis indicates that the Project achieved a favorable economic internal rate of return (EIRR)—See table 3.2.  The EIRR for RSAP-1 is 29% and the NPV is $225 million;  The EIRR for RSAP-1I phase I is 44% and the NPV is $275 million; and  The EIRR for RSAP-1I phase II is 24% and the NPV is $32 million. 8. This indicates that the realized benefits of the Project were sufficient to make it a successful for RSAP-1 and the two phases of RSAP-1I. Table 2 presents the breakdown of the NPV by road sections. As expected, the NPV of benefits is positive for all road sections and the EIRRs for each road section varies from 16% to 51%. All EIRRs are well above the 12% threshold. - 39 - Table 3.2: HDM-4 Analysis Summary NPV (in Million) Road Section EIRR Rupees USD RSAP-1: CP01-A008: Ingiriya – Ratnapura 28% 1,631 12 CP02-A001: Nittambuwa–Kandy CP 51% 9,318 66 CP03-A016: Haliela -Bandarawela CP 38% 3,074 22 CP04-B044: Bandarawela –Welimada 21% 633 5 CP05-A017: Galle – Deniyaya –Madampe 26% 2,672 19 CP06-B098: Denagama –Mulatiyana 23% 798 6 CP08-A014: Medawachchiya -NCP Boundary 17% 389 3 CP09-A004: Wellawaya –Siyambalanduwa 33% 3,830 27 CP10-A025: Siyambalanduwa–Ampara 22% 1,558 11 CP11-A011: Maradankadawala–Jayanthipura 16% 581 4 CP12-A011: Jayanthipura –Tirukkondainmadu 21% 1,788 13 CP13-A010: Padeniya- Puttalam 31% 5,241 37 Sub-Total: 29% 31,513 225 RSAP-1I -Phase I (sections A2, A3): CP01-A002: Maliban Junction - Walana Junction 39% 18,128 129 CP02-A003: Nawaloka Junction – Seeduwa 51% 20,465 146 Sub-Total: 44% 38,593 275 RSAP-1I-Phase II (section A6): CP05-A006: 157+000 km to 167+280 km 24% 1,546 11 CP06-A006: 167+280km to 178+000km 21% 773 6 CP07-A006: 178+000 km to 190+000km 21% 628 4 CP08-A006: 190+000km to199+300km 28% 1,530 11 Sub-Total 24% 4,477 32 Comparison to PAD 9. We compare the various phases of the Project before and after project completion (see table 3.3). For the 620 km road section, we compare the economic analysis in the PAD to the economic analysis in the ICR. For the KT Section, A002, and A003, we compare the economic analysis of the restructuring paper to the economic analysis of the ICR. - 40 - Table 3.3. Results of Economic Analysis-Before and After the Project EIRR at NPV at appraisal EIRR at ICR NPV at ICR Length Appraisal ($, millions) (%) ($, millions) (%) 620 km (using different time 31 103 29 225 horizon) 620 km (using same time 31 103 17 24 horizon) Before Restructuring 134 km 51 522 Dropped and replaced with A002, A003 (A006) After Restructuring K-T Section 21 24 24 32 A002 29 101 39 129 A003 47 132 51 146 10. The results of this comparison are presented in table 3.6. For the 620km section, the NPV at the time of appraisal was $103 million and the EIRR was 31 percent. In comparison the NPV after project completion (at ICR) is $225 million and the EIRR is 29 percent. This difference in the NPV values is driven by a difference in the assumptions of the PAD and ICR—the analysis in the PAD is conducted for an eleven year time horizon while the analysis of the ICR is conducted for a twenty year time horizon. If we limit the time horizon of the ICR analysis to eleven years then the resulting NPV is $23.6 million and EIRR is 17 percent9. 11. The estimates of NPV and the EIRR for the K-T section, A002, and A003 increased after project completion. For the KT section, the NPV and EIRR at the time of restructuring was $24 million and 21 percent, respectively—the NPV is $8 million higher and the EIRR is 3 percentage points higher after project completion. For A002, the NPV and EIRR at the time of restructuring was $101 million and 29 percent, respectively—the NPV is $28 million higher and the EIRR is 10 percentage points higher after project completion. For A003, the NPV and EIRR at the time of restructuring was $132 million and 47 percent, respectively—the NPV is $14 million higher and the EIRR is 4 percentage points higher after project completion. For both RSAP-1 and RSAP-1I the benefits from travel time savings are higher than the benefits from vehicle operating cost savings. 9 Ideally, the analysis in the PAD and ICR should be performed over the same time horizon. - 41 - `Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team Members Responsibility/ Names Title Unit Specialty Lending Isabel Chatterton Financial Specialist SASEI Team Leader Amali Rajapaksa Senior Infrastructure Specialist GTIDR Co-Team Lead Rajesh Bahadur Singh Dongol Program Assistant GFMDR Team Support Deepal Fernando Senior Procurement Specialist GGODR Procurement Surendra Govinda Joshi Consultant GTIDR Tran. Spec. Sumith Pilapitiya Lead Environmental Specialist GENDR Environment Swaminathan A K Consultant GTIDR Tran. Spec. Simon Thomas Lead Transport Economist SASDT - HIS Economist Supervision/ICR Amali Rajapaksa Senior Infrastructure Specialist GTIDR Team Leader Ishtiaque Ahmed Transport Specialist SASDT - HIS Transport Zia Al Jalaly Senior Social Development Specialist OPSOR Social Dev. Farahnaz Azoor Program Assistant GFAAR Team Support Darshani De Silva Environmental Specialist GENDR Environment Rajesh Bahadur Singh Dongol Program Assistant GFMDR Team Support Deepal Fernando Senior Procurement Specialist GGODR Procurement Susanne Holste Lead Social Development Specialist GSURR Social Dev. Sashikala Krishani Jeyaraj Program Assistant SACSL Team Support Ashok Kumar Senior Highway Engineer GTIDR Transport Sumith Pilapitiya Lead Environmental Specialist GENDR Environment Zafar Iqbal Raja Senior Highway Engineer GTIDR Transport Amali Rajapaksa Senior Infrastructure Specialist GTIDR Team Leader Eashwary Ramachandran Operations Analyst GENDR Operations Rajesh Rohatgi Senior Transport Specialist GTIDR Transport Sunethra Chandrika Samarakoon Procurement Specialist GGODR Procurement Jiwanka B. Wickramasinghe Senior FM Specialist GGODR Financial Mgt. Bernadeen Enoka Wijegunawardene FM Specialist GGODR Financial Mgt. Samantha Prasada Wijesundera Water and Sanitation Specialist GWADR Wat. & San. Haider Raza Senior Procurement Specialist GGODR Procurement Comfort Onyeje Olatunji Program Assistant GTIDR Team Support - 42 - (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle $, thousands (including Travel No. of staff weeks and Consultant costs) Lending FY04 19.71 49.18 FY05 42.02 141.9 FY06 47.82 176.3 FY07 0.87 4.8 FY08 0.0 0.0 Total 110.42 372.18 Supervision/ICR FY04 – 0.00 FY05 – 0.00 FY06 17.41 65.38 FY07 42.45 101.00 FY08 29.64 91.01 FY09 51.88 110.31 FY10 48.21 128.99 FY11 33.99 113.24 FY12 60.75 174.16 FY13 83.53 173.55 FY14 31.98 64.66 FY15 27.05 115.82 FY16 14.24 60.71 Total 441.13 1198.83 - 43 - Annex 5. Beneficiary Survey Results - 44 - Annex 6. Stakeholder Workshop Report and Results 1. A Bank mission visited Sri Lanka during June 8 –12, 2015 to gather information for the preparation of the ICR. The mission team comprised Radia Benamghar (Transport Specialist/ICR Team Leader), Mirtha Pokorny (ICR Author), Amali Rajapaksa (Senior Infrastructure Specialist/Project Team Leader), Darshani de Silva (Environmental Specialist), Haider Raza (Senior Procurement Specialist), Razaak Ghani (Senior Social Development Specialist), Susrutha Goonasekera (Social Development Specialist), Enoka Wijegunawardene (FM Specialist), Comfort Olatunji (Program Assistant), and Sriyakanthi Sujeeva Roshinie Pieris (Team Assistant). 2. The mission’s objective was to collect information and inputs that will help in drafting the ICR. As part of this mission, the ICR team (a) exchanged views with the implementing agency about the drafting of the completion report; (b) carried out site visits of two roads, A002 and A003; (c) conducted interviews with the relevant government agencies and stakeholders, including the MOH, Ministry of Investment Promotion, RDA, PMU, contractors, communities, and so on; and (d) delivered an ICR workshop along with working group session to draw the lessons learned from the Project. 3. The workshop was attended by a large audience from the various government agencies, PMU, contractors, consultants, and beneficiaries who expressed ideas and provided valuable input for the ICR. The following topics were discussed during the workshop: Design, Construction, and Supervision of Civil Works (Group 1) Design and Quality of Supervision of Civil Works (a) Problems in the design and supervision stage  Lack of proper investigations due to insufficient time allocations (pavement design and changes during construction)  Poor attention to hydrology and drainage matters and related construction issues, lead away drains, and so on (acquisition)  Design issues with vertical alignment with hydrology and access to adjoining land properties  Acquisition issues during construction, public complaints, and general intervention (b) How to address the problems  Identify proper evaluation of the design components  Coordination between design team and stakeholders (c) How to ensure that design and design review are compatible - 45 -  Design a checklist for design and design reviews and strictly follow the same 4. Supervision of work  Problems in controlling quantity and quality of the construction work  Continue receiving inputs (availability) from the design team throughout the Project duration (responsibility concerns)  Adequate staff allocation and time allocation and review of the experience of staff in the relevant field 5. Construction industry  Developments in the local construction industry through project implementation such as the following: o Need for tremendous improvements for staff training and transfer of knowledge to be continued (the discussion centered on whether this has been done adequately) o Improvements to labs and laboratory equipment and continuous review of the standards of the same o Contract administration capacity should be improved Road Maintenance (Group 2) (a) RMTF  Yes It makes a system to identify in priority maintenance program *HDM and it created the situation to government to collect funds from fuel CESS and so on. Because of this the quality of the road was improved.  Yes  Continue Funding - Bank/Donors 50–50% 40–60% 30–70% 20–80% - 46 - (b) Asset management  All decisions with regard to roads are based on engineering and technical information and for that purpose asset management is important  Updated correct/reliable data for management to take decisions  This systems allows to take intervention decisions based on economic value and value for money  Funding for data collection to be established (c) Rural roads  Appropriate data-based system has to be developed for road information for them to take decisions on interventions  Make allocations through RMTF (with the increased capacity of RMTF) Road Safety (Group 3)  Road safety corridor was selected based on accident data statistics  Police was very weak in coordination during construction  Utility shifting agencies were weak in shifting utilities on time  Utility agencies are not able to provide existing underground utility details  Pilot to be improved (junction mirrors can be provided)  Contract period should be selected carefully  Both side fencing can be implemented  Roadside parking should be restricted by providing identified parking area Environmental and social safeguards (Group 4) 1. GoSL/bank policies  LAQ  NEA  L/Government 2. NEA - 47 -  LAA  NIRP (2001)  Environmental and Social Safeguards Manual 3. LAD—v/Land Donation  GRM  E&S Committee at Project Level  GRC – D.S. Level 4. Voluntary land donation should be avoided - 48 - Annex 7: Part 1 -- Borrower's ICR 1. Introduction 1. In 2004, the GoSL sought the World Bank’s re-engagement in the road sector after a lapse of over a decade, to help address the critical state of the country's road network due to a huge backlog of road maintenance at all levels – national, provincial and rural. Contributing to (and exacerbating) the disrepair and deterioration of roads was the increasing traffic congestion in the national roads, which compelled the government to increase road capacity – a costly process requiring land acquisition – at the expense of the maintenance of all classes of roads. In addition, Sri Lanka’s violent civil conflict, experienced since the early 1980s, took a heavy toll on the government’s treasury. Sri Lanka needed considerable investment to bring its road system to a level of performance consistent with its DOs. 2. The GoSL’s development strategy highlighted the urgent need to develop its roads and increase connectivity and interaction between the provinces for both social integration and economic development. This coincided with the World Bank’s assistance strategy that supported investment in roads as an essential infrastructure development area for pro-poor growth linked to shared prosperity. Both strategies emphasized the importance of developing private industry for road construction and maintenance. By disbanding the state-owned Road Construction Development Company (RCDC) in 2003, the GoSL demonstrated that it was serious about revitalizing the road sector by enabling the participation of the private construction industry. The RDA’s reliance on the RCDC for most road works funded by the government had inhibited the growth of Sri Lanka’s domestic road construction capacity. 3. Thus, the GoSL was committed to institutional and policy reforms to meet the needs and targets of the road sector. In November 2004, the Bank, JICA, and the ADB decided to coordinate their activities in the road sector within the context of the Reform Framework. The GoSL began addressing Sri Lanka’s road rehabilitation needs for the first time through the RSAP. The project was designed to focus on the “maintenance of existing assets”. The maintenance focus was supported by the rehabilitation of national roads, establishing a sustainable arrangement for continuous maintenance as well as carrying out the preparatory work necessary for the rehabilitation of rural roads. This was supported initially through an International Development Association (IDA) credit of $100 million. The roads rehabilitated in this original phase provided the much-needed connectivity to all of the provinces, including the north and the east. Additional funding of $98 million in 2008 and subsequently another $100 million in 2011 was channelled to improve the road conditions in the urban sections in the high traffic areas of the western province. The project assisted Sri Lanka to increase the percentage of national roads in good and fair condition from 48 percent to 65 percent. 4. With regard to rural roads, although ADB is providing support for rural infrastructure including rural roads in the southern, north central, north western, and central provinces, this and overall donor assistance to this sector, had been modest compared to the needs. For this reason, the Project prepared the necessary framework and built the required capacity for the rural road sector to absorb funding for their rehabilitation needs by establishing a rural road strategy, preparing a comprehensive database and providing training to the Pradeshiya Sabhas (provincial authorities) in related areas, including the prioritization and selection of investments. 2. Assessment of Project Objectives, Design, Implementation Experience and Outcomes Project Objective: RSAP’s DO was to lower transportation costs through the sustainable delivery of an efficient national road system that serves the needs of road users and the Sri Lankan public at large. 5. Project Design: The project’s design had to take into account the country’s huge backlog of road maintenance – a situation compounded by not just financial constraints but also inadequate institutional capacity in the road sector – and the urgent need to improve connectivity between provinces by rehabilitating the priority - 49 - corridors of the national network. The RMTF was established to ensure a sustainable mechanism for continuous funding for the maintenance of road assets and to prevent the recurrence of the deterioration of the road network. 6. To achieve the increased workload from road rehabilitation, the private contracting industry needed to be co-opted and provided work. To support this, the RCDC was closed down and hands-on support provided to improve the capacity of contractors in a well-coordinated effort by the three main road sector donors. Moreover, as the rehabilitation of the rural roads was also a priority concern and the capacity to achieve it insufficient, the Project laid the framework for developing institutional capabilities for road infrastructure rehabilitation and maintenance within Sri Lanka’s decentralized governance system 7. RSAP did not simply focus on the rehabilitation of roads. It emphasized the improvement of development effectiveness by identifying and catalysing the changes the sector needs to make to achieve the required efficiencies in the longer term. RSAP supported RDA to establish a fully-fledged asset management system to inventorize, prioritize, monitor and manage its network with a view to increasing the organisation’s ability to provide road users with a level of service that they would demand. Concerning Rural Roads, the Project began its process of addressing maintenance needs of rural roads by incorporating a small pilot component. 8. RSAP’s national roads component covered maintenance, rehabilitation and minor upgrading in eight provinces10. It resurfaced and improved 685.475km of Class A and B roads; provided technical assistance for the selection, design and construction supervision of the civil works; building capacity within the RDA; and enabled the development of the local construction industry. The rural roads component rehabilitated 162.23km of rural roads in nine local government areas, whose selection was based on the recently completed poverty map of Sri Lanka; built capacity within local authorities and of local contractors for road rehabilitation and maintenance; and created a comprehensive rural roads database for the entire country that would support more rational planning and prioritising in the development of rural roads, which represents nearly 75 percent of the country’s total road network. The institutional strengthening and policy support component cut across the first two components, providing technical support for institutional changes and the adoption of more efficient mechanisms and systems for improved road maintenance. RSAP was equally a capacity building project as it was one of infrastructure development. 9. Implementation experience: At the outset, a pressing challenge was to complete the procurements for the RSAP road packages within a strict timeline due to the urgency of achieving the Project’s objectives. Delays in the implementation of RSAP would have caused serious bottlenecks and holdups in the GoSL’s socio- economic recovery and development process. The mobilization of consultants and the invitation and evaluation of bids had to be done in quick succession. The Bank conducted training sessions for government officials on how to select consultants following the Bank procedures. Training was also carried out to introduce the RSAP perspective to contractors. The Bank was also responsive to the GoSL’ s request to use the standard bidding document of the Institute of Construction Training and Development (ICTAD), as the contractors were familiar with it. This helped move the procurement process. 10. Due to the relatively small number of contracts previously offered to the private sector, the local construction industry had limited opportunities to develop. This lack of experience in turn resulted in most companies being unqualified to meet the World Bank’s requirements for bidding. As a solution to this impasse, RSAP encouraged a joint venture facility that enabled companies to join forces and combine capacities. This would become one of the momentous strategies to be adopted by the Project. Another strategy adopted to support the local construction industry was the division of the Kantale - Trincomalee highway into four equal segments 10 However, the Northern Province could not be included as it was not accessible for a project of this nature due to the conflict. - 50 - to enable smaller companies to bid for the contracts. Local contractors recognize RSAP as the Project that was key to the development of the local construction industry. 11. The National Road Sector Master Plan, which was finalized in 2007, required the widening of roads from 5.5m to 6.2m. This called for changes in design after the award of the civil works contracts – causing delays in implementation as well as the escalation of costs and sometimes questioning the capacity of the contractors to undertake the increased workload. The project experienced variations, which were 31 percent of the original awarded price. The design changes made to accommodate the widening of the road as well as designs of culverts and drains not being identified up-front have contributed to this issue. This situation arose due to the appropriate design requirements regarding culverts and drains not being identified at the design stage. 12. The Grievance Redress System that the Project put in place made it possible for the public to resolve their problems quickly, transparently and out of court. The public also had the benefit of a specially allocated hotline to access project management with concerns and complaints. If people do not agree with the solution offered, a grievance redress committee (GRC) is formed so that they could make their case. If there was no agreement, people could take legal action. RSAP received over 4000 complaints; all but two were resolved on site. More than 60 percent of the public complaints or requests were related to house access, 25 percent to road access, and 10 percent to retaining and parapet walls. People often referred to the Bank’s safeguard policies to protect their interests. 13. RSAP also took practical measures to improve coordination on road development between national agencies. For example, the RDA and the NWSDB have frequently been accused of lack of coordination in project planning, where the NWSDB would dig up the roads to lay water pipelines as soon as a road was rehabilitated. The road works planned in the urban road sections within RSAP, with the laying of water pipelines in coordination with NWSDB, has started a new working relationship that would avoid such waste of resources. 14. A vital initiative supported by RSAP was the RMTF, set up to ensure the sustainable and transparent allocation of resources and funds as well as to carry out efficient monitoring of maintenance. The financing of maintenance works by the World Bank was intended to help the RDA to build capacity to optimize the selection of roads for maintenance and align its processes toward minimizing costs. An innovative concept that the RSAP- 1ntroduced as a pilot initiative on the Kantale - Trincomalee road was Third Party Monitoring, which is essentially a system of benefits monitoring from the perspective of users and stakeholders. This would be a means of gaining their views regarding the road construction so that they could influence the design and implementation of the Project. The long-term objective was to make benefit monitoring by project beneficiaries an integral part of road sector monitoring processes. 15. A five-pillar approach formulated by the UN on road safety was adopted to design a 17km demonstration road safety corridor from Ratmalana to Nalluruwa – a stretch of road with a high level of road accidents. A committee comprising representatives from the MOH, the RDA, the Traffic Police, the Ministry of Health, the four local authorities on the corridor (Dehiwala/Mt. Lavinia, Moratuwa, Panadura and Kalutara), the University of Moratuwa, the MOT and the National Council for Road Safety (NCRS) designed and implemented the demonstration corridor that strengthened some of the pillars identified, such as improving infrastructure that is safe, improving road user behavior, and commencing a dialogue toward strengthening post-crash care services. The launch of the new safety corridor saw the parallel launch of an awareness campaign using multimedia including street drama and animated road signs to promote good pedestrian practice. 16. The Project gave rise to a new form of social responsibility among contractors. In all construction packages of the national roads components and in some in the rural roads component, contractors funded and carried out several activities, addressing the needs of communities. These included repairs to access roads, construction and renovation of buildings and providing for the needs of service providers such as police stations, schools and health clinics. These actions of contractors helped create goodwill between the contractors and the community and this, in turn, had a positive impact on the relationship between the community and the Project, especially in managing conflicts and disputes. - 51 - 17. In the area of technology too, RSAP made some clear advances. It was with this project that road construction in Sri Lanka was updated from the use of macadam and double bituminous surface treatment (DBST) to asphalt, which provides smoother surfaces as well as the ability to better withstand wear and tear. Although the initial investment was higher, the considerable savings on maintenance costs meant that the switch to asphalt, based on sound economic analysis, proved beneficial. Thus, the introduction of asphalt for the national roads resulted not only in improved quality but also in long-term financial benefits. This required contractors to purchase asphalt plants, which RSAP facilitated with an advance payment. By December 2014, the Central Environmental Authority had granted 140 Environmental Protection Licences for asphalt plants, of which RDA owns and operates ten. 18. The RSAP-1I roads pass through the most urbanized and congested areas in the country. Finished road levels have reached their highest levels in relation to adjacent properties. Overlay with milling will become increasingly necessary in future pavement strengthening projects. Considering the Project requirement and future benefit to the RDA, a Milling Machine was procured under RSAP and used for the Project. In low lying areas where reconstruction is necessary, high water table has limited excavation depth. This precludes the possibility of increasing the pavement thickness by excavating below existing road level and was a problem when a thick overlay is required. Alternatively, for the same pavement life and strength, pavement thickness reduced significantly by reinforcing with paving fabric (at AC level) and geo-grid (at DGAB level) and geotextile (at sub-base level). 19. RSAP trained all project engineers on environmental safeguards, in a one-year process done in modules, in collaboration with the University of Peradeniya. The Southern Transport Development Project – the construction of the southern expressway – was used as the case study to understand what went wrong, what went right; and the training process also helped the engineers to see the true value of an Environment Impact. The RDA and the construction companies improved their proficiency to implement safeguards and manage safeguard issues while putting in place the systems and procedures that were needed to institutionalize these practices. 20. The RDA had over a period suffered from external consultants carrying out designs with the risk of that design requiring additional quantities, or technical errors being borne by the government. RSAP encouraged the designs to be carried out through RDA's design units established with the re-engineering of the RDA. This was the first opportunity that RDA had of implementing its own designs under an externally funded project. A similar initiative was carried out for building supervision capacity, where a pilot project was implemented to supervise the Kantale - Trincomalee section by a unit within the RDA. RDA is also increasingly carrying out feasibility studies and economic analysis of its own projects. The RSAP helped improve capacity in construction management, planning, project formulation, construction design, acquisitions, construction supervision, and maintenance. 21. Project Outcomes: PDOs has two components they are the lowering transportation cost and sustainable delivery of an efficient National Road system. Lowering transportation cost has been achieved through: Outcome indicators on the National Road Networks via reduction in VoC, reduction in average network roughness, reduction in network poor and bad condition roads, number of kilometres rehabilitated and intermediated indicators on the National Road Networks was achieved via: reducing Travel Time, Annual Road Maintenance Program and User Satisfaction Survey. 22. The road sections for upgrading and rehabilitations were selected based on the NRMP (2007 – 2017) and the VoC calculation proposed under “target” has been based on the NRMP. The majority of the candidate roads identified for upgrading in the NRMP were in the western province that carry a high volume of traffic. The level of traffic has a high impact on the VoC. However since the cessation of the conflict, to address socio economic issues, more funds were channeled in to the eastern and Northern provinces that were starved of any investment over a long period of time due to the conflict and accessibility issues. These investments were not featured in the NRMP due to these areas not being accessible at the time of preparing the NRMP. - 52 - Table 7.1. Results Framework and Monitoring Indicators Outcom Achieve ment of Target Values 2006 2007 2008 2009 2010 2011 2012 2013 2014 e Outcome Indicators Documents Reference Achieved Achieved Achieved Achieved Achieved Achieved Achieved Achieved Achieved Baseline Target Target Target Target Target Target Target Target Target 23.9 2005 0.0 0.60 1.50 2.70 3.80 (LKR per 15.37 14.91 14.83 14.32 14.21 (PAD) 0% % % % % km) Reduction in Not 2011 VoC 14.6 Achieved (Projec (LKR per 13.6 14.21 12.5 14.13 11.5 14.03 10.5 13.88 t km) Paper) 2005 9.5 Reduction in 9.5 6.8 9.31 6.7 9.02 6.5 8.66 6 8.39 6.2 (PAD) 3 average 2011 network IRI Achieved (Projec from 9.5 IRI to 6.2 6.1 6.2 6.1 6.1 6 5.9 5.9 5.9 t 8.4 IRI Paper) Reduction in 2005 48 52 48 48 47 43 44 38 38 35 40 network in poor (PAD) and bad 2011 Achieved condition from (Projec 38 37 40 36 35 36 34 35 33 52% to 35% t (in percentage) Paper) Component 1: To be About 620km of – undertak – – 150 – 300 – 450 – 620 620 – Achieved National roads en YR 1 resurfaced Reduction in network IRI for – 7.8 8.6 6.93 7.9 – 4.5 – 2.8 4.03 2.9 2.8 – Achieved RSAP roads from 7.8 TO 2.9 50% of relevant staff in the Planning and Maintenance – 0% 0% – 10% – 30% – 40% – 50% 67% – Achieved Divisions of RDA trained in road assets management To be Component 2: – undertak – – 50 – 250 – 450 50.03 635 101.38 – 162 – About 635km of en YR 1 Achieved Rural Roads improved Restructured Reduced Travel Time for Rural – 0 0 – 0 – 3 – 3 – 4 60 – Achieved Roads (%) Component 3: Routine and Periodic maintenance funding 30 Achieve – 13 28.24 34.8 30.82 38.3 28.65 42.2 26.53 46.3 56.54 45.22 70.55 43.10 46.72 46.48 39.89 50.55 28.53 increased to .0 d $46.3 million (constant 2005 prices) (in $, millions) - 53 - km of roads 80.15 Achieve – – – – – – – – – – – – 0 – 41 – 80 – – rehabilitated 5 d P h Reductio a Achieve n in – 6 2.8 2.33 – – – – s d Network e Roughnes 1 s IRI for A Achieve project – 4.3 – – – – 2.4 1.9 2 d Roads A Achieve – 3.1 – – – – 2.4 2 3 d – P h a Achieve Reduced – 66 42 40 – – – – s d Travel e Time on 1 project A Achieve roads – 13.8 – – – – 12.5 11.3 2 d A Achieve – 26.5 – – – – 23.9 23.5 3 d P h a Improved – – – – – – – – – – – – – – – – – – – 80.89 – s level of e satisfacti 1 on of A road – – – – – – – – – – – – – – – – – – – 2 users Achieve No 38.3 68.15 A d – collecte – – – – – – – – – – – – – – – – 3 d Intermediate Result (Component 2): Road Maintenance Trust Fund in Operation and road Maintenance 3-year program implemented About 60km of National roads on which Achieve periodic – 0 0 – 20 – 45 – 60 64.5 d maintenance has been – carried out Annual road Maintenance New Appr Appr Appr Appr program is – – – – – – Baseline oved oved oved oved approved by RMTF Parent Project Component 2 Number of km Achieve Rural Roads – – – – – – – – – 157 162 d improved - 54 - 23. Over the period from 2007–2012 alone, the GoSL has in effect achieved 109 percent of the NRMP. Even though a considerable length of roads in the northern, eastern, and southern provinces, which carry a very low traffic volume were rehabilitated, this was not adequate to compensate for the loss of VoC from the high traffic volume of western province roads. This has resulted in falling short of the VoC targets forecasted. 24. The Maintenance funds disbursed lag behind the target value when considering the reported figures which are only for the periodic and routine maintenance of roads and bridges, and hence reflect as not achieving the target. 25. However there were maintenance works carried out for the widening and improvement of roads, which were carried out with funds from outside the RMTF, such as from the GoSL, ADB, JICA and other Donors. These cost were not taken into the reported routine and periodic maintenance expenditure. Table 2 (below) clearly shows that with the cumulative expenditure for both maintenance of roads and bridges and for road widening results in achievement of the target for maintenance expenditure. Table 7.2. Maintenance Expenditure 2006 2007 2008 2009 2010 2011 2012 2013 2014 Achieved Achieved Achieved Achieved Achieved Achieved Achieved Achieved Achieved Target Target Target Target Target Target Target Target Target Maintenance of 30. 26. 46.7 28.5 Roads and Bridges - 28.24 28.65 56.54 70.55 39.89 82 53 2 3 $, million (Table 1) 38. 42. 46. 45. 43. 46.4 50. 30.00 34.8 Widening and 3 2 3 22 10 8 55 41. 47. 130. 171.2 110. Improvement of 28.88 55.87 78.51 134.7 41 84 68 5 48 Roads ($, million) 72. 74. 135.0 205.2 177. 211.1 139. Total ($, million) 57.12 84.52 23 37 5 5 4 4 01 26. Although initially 635 km of rural roads were identified, subsequent restructuring of the Project saw this reduced to 157 km, with 162 km of roads being rehabilitated, and hence the target was achieved. 27. The second component which is the sustainable delivery of an efficient National Road System was achieved via the establishment of RMTF and channelling of funds into road rehabilitation and maintenance, closure of RCDC and establishing capacity of contractors, establishing efficient cooperation amongst utility providers to prevent damage to the investment toward the latter part of the Project, and ensuring that all utility relocation costs are financed through the Project, and achieving environment and social sustainability through changing the culture of RDA and contractors to comply and follow environmental safeguards, as well as effective implementation of Grievance Redress Committee and third party monitoring. 28. In addition to the rehabilitation of 162 km of rural roads and the building of local capacity in the sector, RSAP’s rural roads pilot project produced a rural roads database and along with it a rural roads development strategy. The data generated includes information such as road condition, vulnerability of roads to flooding, number of families with direct access, percentage of Samurdhi (government welfare) recipients living along the roads, and the main use of the roads, for example, whether used mainly for industrial or agricultural transport. The pilot project provided a good indication of the benefits arising from simple, but well-structured road interventions that enhanced both mobility and access to services and markets. - 55 - 29. In the Rural Roads Pilot Project, the following key results11 were observed: The percentage of families with immediate access to an “All Weather Road” rose from 48.3 percent to 70.9 percent. Agricultural sales points increased by 160 percent and service providers of different sorts – ranging from communications centres to repair workshops – increased by 143 percent. In the survey areas, 14 percent of students moved from their poorly resourced schools and gained admission to schools located in towns – with better facilities and better quality teachers. Moreover, school attendance of both students and teachers showed a definite improvement. Student attendance increased by an average of 45 percent - ranging from 40 percent to 60 percent with seasonal fluctuations. A competitive market environment emerged, with small enterprises diversifying their range of products and increasing annual turnover. However, some small businesses who found it difficult to adjust to the changing environment, became nonviable and closed down. Improved roads also created a competitive environment for marketing agriculture produce, with local producers better able to negotiate for favorable prices. Improved labour mobility and employment opportunities for people saw household incomes taking an upward turn. Road development encouraged new settlements – leading to increased land values. Due to the reduction of Transport Cost, Traffic volume of trishaws, vans/cars, busses, and lorries have increased significantly. 30. New systems and practices: RSAP recognized that environmental and social risk management requires appropriate planning, implementation and monitoring, even for small works. The GoSL regulations stipulate an Environmental Impact Assessment for the construction of new roads, but this does not apply for the rehabilitation of roads, which does not generally call for land acquisitions. However, the Bank requires that its environmental safeguard policies be put in place to ensure long-term sustainability, even when government regulations didn't call for them. 3. Factors Affecting Implementation 31. The implementation of RSAP was somewhat constrained due to the country’s security situation. The project had to contend with a failed ceasefire and renewed hostility soon after it commenced work. However, with the cessation of violence in 2009, implementation took place at a steadier pace. A summary of the other factors affecting implementation are given below. 32. Delays in approvals: Road contractors had to deal with delays caused by the time taken by both the Central Environment Authority and the Geological Survey and Mines Bureau (GSMB) to issue approvals and licenses. 33. Estimation of Contract periods: The original contract periods had been understated and the contracts neither fully appreciated nor reflected the work involved in difficult terrains, nor the additional work due to variations not fully reflecting the time for delays in environmental clearances, which later required additional time for expansion of the scope of work. 34. Sufficient time for completion of designs: Some of the shortcomings of the designs arose as a result of the short time periods allowed for the designs due to the urgency of the Project. 35. Involvement of Provincial Staff: Effective engagement and interest of the provincial staff did not take place at the design stage. 36. Spread of roads in the Project and staff of the supervision consultant: As the RSAP was the first project for rehabilitation, the priority roads across the entire nation were included and the geographical spread made it more difficult to supervise all of these sections as the RDA themselves were not accustomed to 11 Rural Accessibility Survey - Rural Roads Pilot Component, Ministry of Local Government and Provincial Councils. 2012 - 56 - supervision at such a scale. This was made especially difficult because there were multiple donors involved in the rehabilitation of national roads. 37. Unavailability of bitumen: The project has suffered from shortages in the availability of bitumen due to the high demand created by the road-works. The monopoly supplier, the Ceylon Petroleum Corporation, in the past was not able to match demand with supply. 38. The isolation of the PMU: The PMU was earlier isolated due to limited participation from other units of the RDA – a situation somewhat aggravated by the differences in the incentives structure that applied to PMUs within the government systems. The earlier interventions of the Project suffered teething problems as a result of this separation. 39. These areas of concern and bottlenecks were addressed as the Project progressed and resulted in improvements in the design of subsequent project phases/credits. 4. Transition Arrangements for Sustainability of Project Achievements 40. The “culture change” that RSAP catalyzed that ranged from the improved practice of safeguards and an effective grievance redress system to better efficiency in all aspects of project design and implementation will be key to the sustainability of its outcomes. The setting up of the RMTF and maintenance systems, the building of capacity within the RDA, including the capacity to manage contracts, and the development of the private contracting industry, are key factors that influence sustainability. A Road User Satisfaction Survey carried out toward the end of the Project confirmed that the general public perceived the RDA’s role current and changing role positively. 41. The RDA has for decades been compelled to carry out ad hoc maintenance based on the availability of finance, putting the already rehabilitated network at risk. The newly energized RMTF will provide a sustainable flow of funding for the rehabilitation of roads. Moreover, the RDA’s transitioning from a project implementation agency to an asset management agency, where the planning, the interventions and managing the country’s road assets is done over a longer time horizon, will make a significant difference both in cost-effectiveness and sustainability. It will also allow the private contractors to attend to required interventions on a timely basis, using a Design-Build-Maintain-Operate and Transfer Model. Due to the sharing of risk between the RDA and the contracting entity, some of the issues previously encountered by RDA can be mitigated. 42. The construction contracts currently in implementation in Sri Lanka are based on payment being made on the amount of materials used; there is little incentive for contractors to save costs. As a result, the quantity variations and resulting contract payments are high. The government is therefore unable to have certainty in expenditure and need to secure funds beyond the budgeted amount. Moreover, if something goes wrong with the design, that risk must also be borne by the government. The Bank is assisting the RDA to adopt a contract format that is more beneficial for the country as a whole, which will provide the government price certainty, and share risks between the government and the contractor. This will also encourage innovation by providing incentives to the contractor to save costs, ensuring value for money. 43. A road-specific EMP was developed for each segment of road rehabilitation. The EMP identified issues based on various activities that will happen within the contract – and came up with mitigation measures. The contractors could then decide which final solution they wanted to employ and bid for the cost of the work involved. The World Bank’s strategy of building capacity in the compliance of safeguards changed the mind-set of the road agency as well as of the contractors. Today, many contractors are adopting these environmental and safety standards, as normal procedure, even outside RSAP. - 57 - 5. Risks for Sustainability 44. One of the continuing risks for sustainability is that there is still no independent source of finance for road rehabilitation and maintenance. The government still depends on a budget allocation. Another crucial need is a revision of the National Roads Master Plan so that it reflects the current context better – incorporating more updated information on crucial aspects like increasing road congestion. The Plan would also need to outline a strategy and an investment plan, clearly outlining its view of what needs to happen and where the funding is coming from. RMTF was established in 2006 with the intention of improving adequacy, reliability, and consistency of maintenance financing for all types of roads. However these objective yet not been reflected in annual maintenance allocations. The government facilitates finance for RMTF through domestic sources. Since 2012, the World Bank co-financed (50 percent share) for identified twelve road projects through RMTF. The contribution of each party was equal to $10 million and these project roads were rehabilitated during 2012 to 2015. 6. Lessons Learned 45. RSAP was of course not an entirely smooth journey. The challenges experienced have been carefully documented so that the lessons learned and recommendations made can be taken into account in future projects. During the long period of implementation of RSAP much of the learning has already been taken into account and changes reflected in the recent investments. 46. The lessons learned include the following: the need for appropriate guidelines to enable the speedy issue of licenses; the importance of identifying all of the parameters at the beginning of project implementation to prevent recourse to avoidable design variations that are time and cost intensive; the need to stipulate cross sections at shorter intervals when preparing detailed engineering designs; the importance of looking at the benefits of using ‘Design-Build’ and ‘Fixed-Price Lump Sum’ contracts to prevent avoidable physical variations during project implementation; the importance of not comprising on the time allocated for the designs; the need to effectively engage provincial staff at the design stage to prevent some of the variation orders that resulted from provincial requirements reaching the Project several years after its commencement; the need to carry out community consultations not only before the implementation of the contracts but also during implementation, as people are more likely to become engaged when construction has begun. 47. Involvement of Communities: Consultations were carried out at the inception of the Project to fully understand the requirements of communities and to incorporate these needs within the Project at the design stage. Experience has however shown that communities effectively engage in such discussions predominantly during implementation. 48. Co-ordination between involved Agencies: where hitherto there had been a lack of co-ordination and co-operation between Utility providers, the RSAP fostered a culture of dialogue between concerned Agencies, so that the road rehabilitation could be undertaken with minimum disruption 49. Some of the key lessons learned which was reflected in this project are: Cost overruns – these have occurred in most of the road contracts due to various reasons. Delays in the contract completion – this occurred due to poor resource planning, poor construction methodologies and so on. Lack of attention to maintenance in rural roads. Bank’s Performance 50. The GoSL saw the Bank's role as going beyond that of a donor to one of an advisor and facilitator. The systems the Bank helped put in place for project implementation and monitoring will have a long-term impact on the efficiency of Sri Lanka's road sector. Improving capacity within the RDA, the agency responsible for steering the Project, was a process that began from the preparatory stages, reaching several significant milestones as the Project progressed. - 58 - 51. At the beginning of the Project, a pressing challenge was to complete the procurements for the RSAP road packages within a strict timeline. The mobilization of consultants, the invitation of bids, and the evaluation of bids had to be done in quick succession. The Bank collaborated with the RSAP to ensure that contracts were awarded in a timely manner. When it comes to donor-funded projects the criticism that the GoSL often gets is that the start-up time is very long. The GoSL’s previous experience is received, which is a considerable waste of time. But in this particular project, an advance payment was received from the intended loan to begin the design work. Later, it was absorbed into the main loan. This saved a lot of time. With RSAP we were able to design various subprojects as we went along, ensuring the smooth implementation of the Project. 52. The local office responded to urgent project requirements quickly without having to always wait for advice or the go ahead from Washington. The Bank also consistently responded to issues within the industry in a manner that was sensitive to the situation in the country, without adopting a bureaucratic one-size-fits-all approach, while at the same time ensuring that the solutions were appropriate and had implications for sustainability. The Bank’s close partnership with Sri Lanka’s other main donors in the road sector – ADB and JICA – with whom they had regular consultations also at the field level, helped avoid duplication of work and ensured that there was no conflict. 8. Conclusions 53. The main achievement of RSAP-1s that it has enabled the road sector in Sri Lanka to develop its all- round capacity for progress and sustainability. The World Bank’s close collaboration with the RDA generated innovative and practical solutions to address a range of challenges. Crucially, most of these approaches, based on clearly defined principles, are being incorporated and institutionalized to enable the sector's long-term development. The extent of RSAP’s achievements can be attributed largely to the World Bank’s long-term strategic partnership with the GoSL in the country’s continuing efforts to modernize and streamline its road sector. Annex 7: Part 2 -- Borrower Comments on Final Implementation Completion and Results (ICR) Report 54. This has reference to the letter signed by DKR Swarna, Addl. Director General (Projects) For Director General Road Development Authority dated 20 April 2016 on above subject. Our comments on the finding of the report are given below. 55. Page iii - Indicator 1(b) - We also agree that use of Network level VoC as PDO indicator is not suitable as the project interventions are very limited to reflect at network level as indicated in the report. And also without having the baseline data at approval level calculation of achievement is not reasonable 56. Indicator 4 (a) and 4(b) - Road Maintenance Expenditure: Annual maintenance allocation for national roads are provided in the Annual budget Estimate as a maintenance item under the name of "Maintenance of Roads & Bridges (Road Maintenance Fund)". But there other items in the budget estimate which will be used for maintenance activities such as periodic maintenance which has not been reflected in the value as it is difficult to separate out from other activities at head office level. As such total expenditure on maintenance of National roads is more than the allocation indicated in the item. It is to be noted here that the original maintenance allocation in the year 2010 was Rs. 4200 mn (US $ 37.1) as per Budget Estimate and actual expenditure under the maintenance item in the year 2010 was Rs6,342.0 (US $56.0mn) (at the rate of 1 US $=Rs.113 at 2010). 57. Page 11- Para 32 - Factors generally subject to Government control • Unrealistic Estimates: The Estimates were prepared by International consultant using available data and the prices at the time of preparation of project. Due to the rapid increase of construction activity all price went up due to the reasons indicated in the report. As such it is out of control of government. As such it is better not to consider this as a government control item. However the bid price increase to a level of 6.8% is normal. - 59 - • Inadequate initial Engineering design: The design of the Original contract package was carried out by international consultant with the traffic and safety assessment by their experts. It is to be noted here that RDA use the "Geometric Design Standards of Roads developed in 1998" for all RDA design. It require minimum of 6.2 m for two lane road. However these design done by international consultant has not gone through the standards. This has to be avoided in future projects. • Unavailability of bitumen: There was a shortage of bitumen time to time due to high demand. If contractors were programmed their works it could have been reduce the impact due this situation. RDA wanted to import bitumen through Maganeguma to control the quality of the bitumen supply. • Time overrun: All the packages of parent project was design by international consultant and also supervised by international consultant. As such the consultant should held responsible for the wrong design. Even if the time allocated for the design is short the consultant should deploy adequate staff for realistic design. This should be reflected in future design activities. 58. Page 12 • Culverts and drains were not properly identified: Design of Parent project roads were done by international consultant. After realization of this situation it was corrected by carrying out the design in- house by RDA for roads under the RSAP-1I • Delay in obtaining Environmental protection/ Explosive permits: Due to the situation of the country at that time issue of permit to use of explosive had to go through several screening process. However it can be a delay of the contractor also not following correct procedure. • Lack of coordination with utility agencies: The name of NWSDB should be corrected as "National Water Supply and Drainage Board". • Frequent turnover of the PMU staff: All changes of Project Directors were done at the latter part of the project. But there were not much issues such as cost overrun in that period. 59. Page 13 - Bullet 12 - ( Para 2) - This is entirely the mismanagement of international consulting firm who carried out the supervision work and they should have control of getting qualified staff not the individual local consultant. If local consultant is not performing team leader of the international consultant could change the consultant. 60. Page 16 para 44 - There was a delay in HDM 4 data evaluation due to the migration of old data to new system which has been upgraded through the project. 61. Page 19 para 54 - This is good as a pilot project to identify issues of implementation of DBMOT method. It is required to have very good data bases as well as the good understanding of behavior of road especially pavement with the situation of the road in terms of traffic and other parameters. As per the standard practices of selecting payment methods for road works, Lump Sum type contracts are suitable only for the Projects, where uncertainties for change of scope during implementation is minimal. With the lack of knowledge of the behavior of road, the bidder will tends to price high to cover up the risk. Therefore, suitability & cost effectiveness of this method will be known only with the receiving of bids. 62. Page 22 Para 64 - Reduction in VoC - It has been indicated here that with the significant reduction of IRI of project roads, only modest reduction of VoC was achieved. It has to be noted here that assuming of increase of speed to have high fuel consumption is not reasonable as these roads are operating at low level speeds. With the improvement of the roads there can be increase of traffic volume on these roads due to attraction. This will keep the speed at same range before the intervention or reduce the speed below the project level resulting of modest reduction of VoC. 63. Page 23 Para 69 - By 2015 GoSL had spent LKR 370 billion to be corrected as "By 2015 GoSL had spent LKR 187.64 billion on National roads" 64. Page 26- Para 75 - Economic analysis period of using 11 years is not reasonable. As indicated in the para it should be same and more appropriately it should be 20 years. - 60 - 65. Page 34 - Para 104 - In this section it is stated that the supervision consultant's performance is non- satisfactory, produced poor designs and followed poor engineering practices and it affected performance of all 12 civil works contracts. It is the clear evidence of poor management of the project by lead consulting firm, but not due to poor performance of few individual local consultants, as highlighted in this report. 66. Page 35 - Para 109 - A dedicated project management consultant - It is to be noted here that project management is responsibility of Team Leaders and Resident Engineers of the supervising consultancy firm. It is necessary to consider their experience in selecting of these staff in future projects. 67. Page 35 - Para 110 - Quality of engineering design, construction supervision should ensure defect free- work - This is very good lesson learned from the project. Add to this most of the issues were due to the poor performance of the international consultant selected for the project in all phases of the project implementation. It is better to revisit the selection process of these consultant. It is proposed to adopt following process in future project implementation a. All the designs procedures, adopted by the Consulting firms shall be reviewed by RDA specialized divisions. This requirement should be made mandatory and the PMU should submit them to RDA for review. If the present capacity of those divisions are not sufficient to take action to mobilize more resources b. The consulting firms, engaged in RDA projects, should follow only RDA design manuals or any other standards, approved by RDA. All the designs manuals shall be updated, if required. c. Prepare Contract Administration manual and update already available manuals and make it mandatory for using them by the Consulting firms and PMUs. d. Standardized Contract documents and make mandatory for using them by all PMUs. e. Formulate Technical Audit Division for updating standardized documents and Manuals, using experience of past and ongoing projects to avoid/minimize problems and to improve performance. f. Ministry or RDA to form separate section under Director (Quality Assurance) to carry out random quality control testing in all project. 68. Page 35 - Para 111 -Adopting sustainable funding sources for the RMTF - RMTF is a good mechanism for the sustainable maintenance of road network and transparency in maintenance fund usage. Management of road sector agencies need to provide high level of commitment for the functioning of RMTF. - 61 - Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders 62 Annex 9. List of Supporting Documents 1. The World Bank (2005) Project Appraisal Document on a Proposed Grant in the Amount of US$100 million equivalent to the Democratic Socialist Republic of Sri Lanka for a Road Sector Assistance Project. 2. The World Bank (2008) Project Paper on a Proposed First Additional Financing Credit in the Amount of US$98.1 million equivalent to the Democratic Socialist Republic of Sri Lanka for a Road Sector Assistance Project. 3. The World Bank (2011) Project Paper on a Proposed Second Additional Financing Credit in the Amount of US$100 million equivalent to the Democratic Socialist Republic of Sri Lanka for a Road Sector Assistance Project. 4. Restructuring Papers, 2008, 2011, 2013, 2014, 2015. 5. QAG Report Quality at Entry, 2008 6. Evaluation of Implementation of Environmental and Social Management Framework of the Road Sector Assistance Project., 2012 7. Rural Accessibility Survey – Rural Road Pilot Component, 2012 8. Road User Satisfaction Survey 9. Aide Memoires for Project Supervision missions 2005-2015. 10. Project Implementation Status Report, Project Files 2005-2015 63 Annex 10. QAG Report QUALITY ASSESSMENT OF LENDING PORTFOLIO (QALP-1) Summary Assessment Sheet PROJECT Sri Lanka - Road COUNTRY Sri Lanka Sector Assistance REGION SAR SECTOR Transport TASK TEAM LEADER Amali Rajapaksa SECTOR MANAGER Michel Audige COUNTRY DIRECTOR Naoko Ishii LN/CR AMOUNT ($MIL) 100 LN/CR NUMBER 41380 TTL (Please indicate if HQ Headquarters PROJECT ID P086411 or Field-based) DATE APPROVED 12/15/2005 DATE EFFECTIVE 03/16/2006 MTR DATE (Planned) 01/12/2009 MTR DATE (Actual) Not Assigned REVISED CLOSING ORIGINAL CLOSING DATE 09/30/2011 09/30/2011 DATE PREVIOUS QAG FRAGILE STATE (Yes/No) No None ASSESSMENT(s) Roads & highways (90%) SECTORAL COVERAGE Central govt. admin (6%) Sub-natl. govt. admin (4%) SPVN BUDGET $000 (incl. TF): FY06 FY07 FY08 ACTUAL: 65.4 100.6 91 DO/IP RATING: FY07 FY08 DO S DO S ISR: IP S IP S EA CATEGORY Partial Assessment ASSESSMENT DATE 10/10/2008 LIKELIHOOD OF ACHIEVING DOs: 2 Guidance: 1= Highly Likely; 2 = Likely; 3 = Moderately Likely; 4 = Moderately Unlikely; 5 = Unlikely; 6 = Highly Unlikely. 64 Amali Rajapaksa; Isabel Chatterton; Sumith Pilapitiya; Darshani De Silva; Sunethra Chandrika Samarakoon; Jiwanka B. Wickramasinghe; Samantha PERSON(s) INTERVIEWED Prasada Wijesundera; Deepal Fernando; Zia Al Jalaly; Zafar Iqbal Raja; Ishtiaque Ahmed; Ashok Kumar MODERATOR Peter Ludwig OBSERVER(S) PANEL MEMBER(S) Robin C. Carruthers; Jacques Bure; Arie Chupak SPECIALIZED REVIEWERS: 1. Social Santhadevi Meenakshy 2. Environment John Buursink 3. Procurement Jorge Claro QUALITY ASSESSMENT OF LENDING PORTFOLIO (QALP-1) COUNTRY: Sri Lanka PROJECT TITLE: Sri Lanka - Road Sector Assistance QALP-1 Assessment: Likelihood of Achieving Development Objectives List of DOs: Ratings: PDO The objective of the Project is to lower transportation costs through sustainable delivery of an efficient national road system that serves the needs of road users and the Sri Lankan public at large. DO1 DO2 DO3 DO4 DO5 2 Other (if Applicable) Likelihood of Achieving DOs Comments: The chosen indicators will not show whether an efficient national road network has been delivered on a sustainable basis, and even if it has, whether this is likely to contribute to economic (implicit in the phrase “needs of road users”) and social development (implicit in the phrase “Sri Lankan public at large”). But if the civil works are implemented together with those of the other funding agencies in the sector, the indicator values will be achieved. Overall Comments A. Quality of Design 1. The original physical component of the Project (improvement to national roads) was designed to meet a specific government request. A rather different physical component (a pilot exercise in upgrading rural roads) was later added to meet a request from the next government. Adding the additional component was a justifiable risk, given the less acceptable alternatives, and in the event, has turned out well. The institutional component relied on a component in a parallel project being implemented by the ADB. This was a marginally acceptable risk, given that ADB had assumed the role of lead agency in the sector, but has not turned out so well. 2. The DOs are closely related to the first and third project components, and the outcome indicators were crafted to meet some subtle differences in how the objectives might be achieved. 65 It was a justifiable simplification to use Vehicle Operating Costs (VoC) and not truck tariffs as the indicator of transportation costs, given the circumstances of Sri Lanka and the difficulties that would have been encountered in trying to estimate tariffs. But in future road development projects in Sri Lanka, as elsewhere, more attention should be paid to the structure of the trucking industry and the reasonableness of the implicit assumption in the DO that reductions in VoC will be passed on to users. It would not be essential to go to the next stage (as suggested at the interview) and link changes in truck tariffs to increased economic and social activity, as such impacts take time to become apparent whereas the lower tariffs should be observable more quickly. 3. The method of selecting sections of road to be improved in the national roads component included a thoughtful balance of technical and social considerations, as well as taking account of the roads being improved under the parallel ADB roads project. More attention should have been given to the design standards used, as they were subsequently found to need improving to achieve an acceptable level of road safety. The design concept for the rural roads component, of including pilot projects for different types of rural road improvement, was a justifiable compromise between rejecting the new government request to include a rural roads component in the Project and fully accepting such a component as an equally significant component as that for national roads. 4. The institutional component aimed at the creation of a road fund for road maintenance, a Board of Trustees to manage the fund, and a Technical Advisory Committee to provide technical support. This was an ambitious agenda, particularly taking account that the existing road agency that would have to be restructured was also implementing the civil works component of the Project and similar projects of other financing agencies. Under these circumstances the use of a PIU (or multiple PIUs) was more justifiable than would otherwise have been the case, since this insulated management of the civil works from much of the institutional disruption. B. Quality of Implementation 5. Given the continued civil unrest in Sri Lanka, the Government’s attention to the Project has been strong. When it became clear that the increases in unit costs for civil works could result in some of them not being completed, the Government took appropriate action to request additional Bank funding. However, some of the additional funding was to meet a change in design standards between contracting the civil works and completing them. While having a plausible justification in meeting higher safety standards, more consideration could have been given to completing the existing contracts under the original design standards, since they could not have been so unsafe as to be unacceptable. The response to rapidly escalating bitumen costs was to allow the contractors to import bitumen as an exception to the Government policy of having it provided by a state monopolist. Unfortunately this response was too little and too late, as the set up logistics costs for individual contractors to importing relatively small quantities of bitumen has not resulted in significant cost reductions, and this could have been foreseen. A quicker, more innovative and comprehensive response was needed. 6. There have been delays in contracting the consultant that has the task of designing the new road maintenance funding system, a delay that is compatible with the current Government’s reluctance to pursue institutional changes. Whatever the reason, the slow progress jeopardizes the chances of achieving the DO objective of delivery of an efficient national road system. 66 7. The Government has been responsive in purchasing new equipment that will facilitate more measurements of the IRI (the values of which constitute one of the outcome indicators) and has been diligent in trying to understand how the previous government arrived at the baseline values. They have also been making estimates of vehicle operating costs (another outcome indicator). C. Quality of Bank Supervision 8. The Bank (both the task team and management) has been responsive in addressing the rapid escalation of contract costs and the Government’s proposed solution for dealing with the cost overruns. While there is little evidence of collusion between contractors in seeking price variations based on unit cost increases, more thorough investigations could have been made, together with other road funding agencies, to verify that this has not happened. Experience from other countries indicates that when unit costs increase rapidly, some contractors take advantage of the situation to increase their invoices even more. The Bank has been diligent in maintaining frequent contact with other agencies active in the road sector, particularly the ADB as the lead agency. These contacts have not been as productive as they might have been, for example in devising a scheme to reduce the impact of increases in bitumen costs, or reappraising the implementation of the design change to on-going contracts. The Bank has given much attention to the interests of residents close to the roads being improved and to efforts to minimize construction impacts on them. However, in agreeing to change the design standard of rural roads from gravel surface to a more expensive but less environmentally intrusive paved surface, perhaps more attention should have been given to the neighbors and users of roads that cannot now be improved as more funding has been given to on-going projects. The decision to adopt this more expensive design concept will have large scale implications for the whole rural roads program. D. Panel Recommendations 9. There is a high probability that the civil works components of the Project will be completed more or less on time, but well over the original budget but probably within the modified budget (taking account of the additional Bank funding and that the worst of the unit cost escalations might now be past). So in the second half of the implementation period more attention can be given to the third project component, Institutional Strengthening and Policy Advice, that addresses the part of the DO related to the delivery of an efficient national road system. This system will require a secure funding for new roads as well as road maintenance and a national road agency that has effective incentives and access to human and financial resources to address them. The performance of the ADB in progressing this agenda has not met the expectations indicated in the PAD and implicit in the DO of the Project. The World Bank will have to take a more proactive role if this agenda is to be met. 10. One way that this could be done is through the rural roads program where the role of the World Bank is stronger than it is for national roads. Given that the new CAS puts a strong emphasis on rural roads and their development role, the pilot projects included in the on-going project could provide a wealth of useful information and support for policy initiatives related to broader issues of funding for roads and the institutional arrangements for building and maintaining them. This relates particularly to design standards and the trade-offs between constructing fewer but better roads or more but less costly roads that are perhaps more difficult to maintain. 67 11. The strategy of designing road maintenance funds based on models used in other countries might no longer be sustainable. While road maintenance needs could often be met from a surcharge on transport fuels of the order of U$0.05 to U$0.09 per liter, with the increase in costs of road maintenance it is probable that this level of surcharge would have to at least double, and increase even more if the global economic downturn results in a slower rate of growth of traffic (road maintenance needs are largely fixed and only decrease slowly in relation to decreases in traffic). There is conflicting evidence as to whether it is the absolute level of the surcharge (an increase of about U$0.05 per liter) that limits the feasibility of an increase, or its share of the total cost of fuel (if the surcharge doubles while the cost of fuel increases three fold or more, the impact of the surcharge might be less than expected). The impacts of the changing global economic context will need to figure prominently in the assessment of an appropriate road maintenance funding mechanism. 12. The Technical Audit has an important role in the supervision of this project. Emphasis should be given in future Aide Memoires to the findings of the audits. 13. The review has noted some apparent inconsistencies in the physical and financial progress of the Project: physical progress is estimated at about 50%, disbursements also estimated at 50%, but contractual costs including escalation factors have been about 50% higher than indicated in the original contracts. These three facts cannot all pertain at the same time. Preparation for the Mid- term review provides a good opportunity for the percentage progress figures to be shown consistent with the cost escalation estimates. E. Any Systemic Lessons 14. The DOs and the indicators used to measure them do not give any indication as to whether economic or social development will result from the Project. The projected values of at least two of the indicators (VoC and IRI values) will be achieved if the Project civil works are implemented, and so are not in themselves indicators as to whether any development will result from the Project. Great care needs to be taken in defining DOs that are not so ambitious that even if they are achieved, such achievement cannot reasonably be attributed to the Project, while at the same time not making them so trivial that so long as the civil works are completed, the indicators will show that the DOs have been achieved. 15. Institutional changes invariably take much longer than expected, so some provision for a slower implementation (a “Plan B”) should be made. This could be an addition to the “Measures taken to mitigate risks.” It would indicate what measures should be taken should any of the risks be realized; it is not enough just to show what has been done to reduce them. 68 Rejoinder from the Region QUALITY ASSESSMENT OF LENDING PORTFOLIO (QALP-1) Guidance Questionnaire Summary Assessment Sheet Assessment Rating 1 = Highly Satisfactory/Likely 2 = Satisfactory/Likely 3 = Moderately Satisfactory/Likely 4 = Moderately Unsatisfactory/Unlikely 5 = Unsatisfactory/Unlikely 6 = Highly Unsatisfactory/Unlikely NA = Not Applicable QALP-1 Assessment Quality of Design 2 Quality of Implementation 2 Quality of Bank Supervision 2 Focus on Development Effectiveness 3 Fiduciary/Safeguard Aspects 2 Supervision Inputs and Processes 2 Candor and Realism of ISRs 2 QALP-1 ASSESSMENT: Likelihood of Achieving DOs 2 Likelihood of achieving DOs is rated on the same six-point scale as other ratings. In arriving at a final judgment on outcomes, panelists should exercise their judgment and take into account quality of design, implementation and Bank supervision, as well as country conditions and known risks. QUALITY ASSESSMENT OF LENDING PORTFOLIO (QALP-1) Guidance Questionnaire Context Section A. Quality of Design 2 (i) Strategic relevance and appropriateness of DOs? Comments: The project design did more or less what the 2 Government wanted. The DOs expressed exactly what they needed, and they are still highly relevant. 69 (ii) Appropriateness of project design, complexity, and approach, given the country’s past track record and absorptive capacity? 2 Comments: The original fears that the pilot rural road projects might detract from the focus on the national roads component were not realized. (iii) Adequate attention to technical, financial and economic aspects? Comments: The selection of road sections to be upgraded made sensible use of the Highway Development Model 4(HDM4), as did the overall evaluation. However, the technical design 3 should have looked more at the safety implications of using a relatively narrow road width of 5m, since it was later found desirable to change it to 6.2m. (iv) Quality of the Results Framework to achieve the DOs, including realism of causal links between the Project’s inputs, activities, outputs, and intended outcomes? 2 Comments: Given the development objectives, the outcome indicators and causal links are adequate. (v) Quality of arrangements for monitoring and evaluation (including clarity and precision of the 2 performance indicators to assess progress and outcomes and availability of baseline data)? 16. Comments: The Roads Development Authority (RDA) is measuring in regular intervals the IRI, VoC and the conditions of the network. However they have gotten the initial baseline values wrong for the IRI, through overuse of extrapolation. The task team is taking adequate precautions to ensure that the subsequent measurements are more reliable. (vi) Adequacy of measures incorporated in project design to address policy constraints? Comments: RDA was being redesigned, therefore selection of PIU for project implementation was justifiable. The ADB is taking the lead on RDA restructuring and Road Fund. While at the time it was correct not to question the leadership of ADB as the Bank 2 was absent from the road sector for 10 years, the subsequent failure to achieve the expected changes indicates a more proactive response from the World Bank will be needed to achieve the DOs. (vii) Level of Borrower ownership, commitment? Comments: The Borrower has demonstrated strong ownership throughout project execution, 2 exemplified by the requests to include rural road pilot studies and for additional funding when it was necessary. (viii) Extent of integration and quality of arrangements for governance and anti-corruption in project design? Comments: The use of third party technical audits is an excellent means to keep the implementing 2 agency, contractors and consultants on their toes with respect to the quality of civil works executed. Also see comments under FM (xi) and procurement (xii) below, which are both rated satisfactory. (ix) Extent of integration and quality of social development aspects in project design? 2 17. Comments: Development objective of the Project provides opportunity to address vulnerability due to poverty, isolation, and gender. 70 18. The project design adequately evidences that measures were taken to promote positive social impacts result from the Project implementation. As part of project preparation, social impact assessments have been completed and social framework was generated though the Project does not trigger OP 4.12. The framework proposes continued consultations with the stakeholders during every stage of the Project to ensure social safeguard compliance is maintained throughout the Project cycle, and provides the principles and the methodology for evolving a grievance redress mechanism. It goes to state that a Grievance Redress Committee (GRC) will be established with critical roles for civil society and concerned stakeholder. 19. With regard to poverty, project covers the poorest segments of the country including provinces close to conflict areas and proposes labor intensive interventions where appropriate. 20. Stakeholder consultations were organized in a manner that ensured gender equity. The social impact assessment report proposes ways to integrate gender dimension within the earth works. The project design would have benefitted by establishing criteria for gender focus in: (i) earth works; (ii) proposed stakeholder dialogue during implementation; and (iii) institution of safety measures in congested areas such as wide sidewalks and proper lighting of sidewalks to enable safe mobility of poor female road users. (x) Extent of integration and quality of environmental aspects in project design? Comments: Environmental aspects of this “Category B - Partial Assessment” project have been fully integrated into project design. Sufficient attention was given to the financial and human 2 resources required to monitor and supervise the multiple environmental issues of national and rural roads during project implementation. Environmental management is well integrated into project design and implementation. (xi) Extent of integration and quality of FM aspects in project design? Comments: Adequate actions to ensure the Project’s FM arrangements to properly record and control project’s financial activities were in place before Loan effectiveness. Adequate attention 2 was also given to mitigate the potential risk of fraud and corruption. Thus, required FM arrangements were in place at an early stage of project implementation. (xii) Extent of integration and quality of procurement aspects in project design? 2 21. Comments: Governance and anti-corruption: As reported in other Sri Lanka projects evaluated, the country does not appear to have a high incidence of corruption although the CPAR did warn about institutional weaknesses, the absence of appropriate procurement controls including the absence of a formal system for performing procurement audits, and a weak capacity to implement the procurement laws and regulations, rating the risk of fraud and corruption as Medium. The same year of the release of the CPAR, Transparency International ranked Sri Lanka as No 67 in a survey of 145 countries, which coincides with the medium classification of the CPAR. 22. The overall procurement risk in this particular project is also classified as Medium although the CPAR is not mentioned. The procurement arrangements in the PAD are very good and consistent with a project of this nature, with stringent controls, ex-ante review of procurement actions and very frequent procurement review missions that include field visits, on-site inspections and post-reviews of procurement actions. In addition, like most major road projects, it has the usual independent consulting engineering supervision. 71 23. Procurement aspects: The procurement aspects at project design were very well laid out, properly organized and straightforward. In fact there was substantial procurement readiness at project inception with an 18 month procurement plan agreed upon from the start. (xiii) Quality of institutional framework for the Project? Comments: Using several PIUs to avoid problems with RDA, which was in the restructuring 2 process was a necessary precaution to ensure implementation of the physical components of the Project. (xiv) Quality of risk assessment and management? 4 24. Comments: While risks were described in the PAD, there were no mitigation measures mentioned. While perhaps not obligatory at the time, it would have been a desirable addition. 1. Quality of Implementation 2 QUALITY ASSESSMENT OF LENDING PORTFOLIO (QALP-1) Table 1. Risk Flags and DO/IP Status COUNTRY: Sri Lanka REGION: SAR PROJECT ID: P086411 PROJECT TITLE: Sri Lanka - Road Sector Assistance DO/IP/Risk Factor FY07 FY08 ISR Panel ISR Panel DO/GO S S S S IP S S S S Project Components Maintenance and Rehabilitation of S S S S National Roads Maintenance and Rehabilitation of Rural S S S MS Roads Sustainable Maintenance Financing S S S S Arrangements Discretionary Flags Counterpart Funds No No No No Financial Management No No No No Legal Covenants No No No No Monitoring and Evaluation No No No No Project Management No No No No Procurement No No No No Safeguards No No No No Sub-Total 0 0 0 0 1/ Exogenous Flags Disbursements Delays No No No No 72 Effectiveness Delays No No No No Country Environment No No No No Country Record Yes Yes Yes Yes Long-Term Risk Flag No No No No Sub-Total 1 1 1 1 Total Flags 1 1 1 1 Golden Flag No No No No 2/ Overall Risk Status Low-Risk Low-Risk Low-Risk Low-Risk 25. Comments 1/ Exogenous flags are system generated flags and have the same ratings for both ISR and Panel. 2/ APP (DO) - DO rated below the line. APP (IP) - IP rated below the line. APP (DO/IP) - Both DO and IP rated below the line. Potential - Non-problem projects with 3 or more flags and without a golden flag. 2-flag - Non-problem projects with only 2 flags and without a golden flag. Low risk - Non-problem projects with 0 or 1 flag and without a golden flag. 73 Annex 11. Map 74