NSP SUBPROJECT SUSTAINABILITY MARCH, 2013 This publication was produced by Altai Consulting for review by the World Bank, the MRRD and the IDLG. The authors views expressed in this publication do not necessarily reflect the views of the World Bank, the MRRD or the IDLG. ACKNOWLEDGEMENTS This report was prepared by Eric Davin, Aschkan Abdul-Malek and Adrian Carriere (Altai Consulting). Field research teams in Badakhshan, Balkh, Bamiyan and Herat were managed by Atiqullah Sahibzada (Noma Consulting) and included Shingul Kaliwal, Eng. Homayun Mohmand, Eng. Wadood Mohmand and Abdul-Saboor Qaderi. Technical assessments in Badakhshan, Balkh, Bamyan, Herat, and Nangarhar were conducted by Eng. Omar Noori, Eng. Rahimullah Zahid, and Eng. Zamangul Ahmadi. We gratefully thank Ladissy Ichengula, Makiko Watanabe, Richard Hogg and Naila Ahmed (World Bank) for their valuable inputs and assistance. We are also indebted to Abdul Rahman Ayubi, Brigitta Bode, Jovitta Thomas (MRRD-NSP), Nader Yama, Sibghat Khan and Abdul Basir Saber (IDLG) as well as to all the Afghan government representatives, policy advisers, NGO workers, academic researchers and community members who shared with us their views on the various themes that this study attempts to cover. i TABLE OF CONTENTS List of Figures.................................................................................................................................... iii List of Tables ...................................................................................................................................... v List of Acronyms ............................................................................................................................... vi Executive Summary............................................................................................................................1 Introduction and Methodology ..........................................................................................................5 Introduction .............................................................................................................................................. 5 Approach ................................................................................................................................................... 5 Methodology............................................................................................................................................. 7 Subproject Status and O&M ............................................................................................................. 17 Subproject Construction Process ............................................................................................................ 17 Contributions to O&M ............................................................................................................................ 18 Overview of Condition and Maintenance ............................................................................................... 22 Subproject Specific Condition and Maintenance Activities .................................................................... 25 Variables Influencing Subproject Condition and O&M ....................................................................... 45 Variables Influencing Subproject Condition ........................................................................................... 46 Variables Influencing O&M ..................................................................................................................... 61 Subproject O&M Cost and Funding Gap ............................................................................................ 75 Introduction ............................................................................................................................................ 75 Estimated Funding Gap by Scenario ....................................................................................................... 77 Conclusions and Recommendations.................................................................................................. 83 Key Findings ............................................................................................................................................ 83 Conclusions and Strategy Recommendations......................................................................................... 86 ii LIST OF FIGURES Figure 1 - Analysis Framework ...................................................................................................................... 7 Figure 2 - Geographical Scope of Study ........................................................................................................ 8 Figure 3 - How Many Community Members Contribute Funding to Maintenance? .................................. 19 Figure 4 - Who Organised Monetary Contributions? ................................................................................. 19 Figure 5 - Yearly O&M Contributions per Village (AFN) by Subproject Type.............................................. 19 Figure 6 - Yearly O&M Contributions per Village (AFN) by District ............................................................ 20 Figure 7 - How Many Community Members Commit Time/Labour to Maintenance? ............................... 20 Figure 8 - Who Organises Time/Labour Contributions? ............................................................................. 20 Figure 9 - How Many Community Members Commit Materials/Resources to Maintenance? .................. 21 Figure 10 - Who Organises Material/Resource Contributions? .................................................................. 21 Figure 11 - Most Frequent Cause of Poor Key Component Condition ....................................................... 24 Figure 12 - Maintenance Performed: Overall ............................................................................................. 25 Figure 13 - Assessed Canal Locations.......................................................................................................... 26 Figure 14 - Assessed Culvert Locations ....................................................................................................... 28 Figure 15 - Assessed Karez Locations.......................................................................................................... 30 Figure 16 - Assessed MHP Locations........................................................................................................... 32 Figure 17 - MHP Components in Deh Payan Naw and Chapchi Bala .......................................................... 32 Figure 18 - MHP Components in Deh Payan Naw and Khairabad .............................................................. 34 Figure 19 - Examples of Protection Walls from NSP Engineering Manual .................................................. 34 Figure 20 - Assessed Protection Wall Locations ......................................................................................... 35 Figure 21 - Assessed Shallow Well Locations .............................................................................................. 37 Figure 22 - Assessed Solar Panel Locations................................................................................................. 39 Figure 23 - Assessed Tertiary Road Locations ............................................................................................. 41 Figure 24 - Assessed WSN Locations........................................................................................................... 43 Figure 25 - Cause of Non-Functionality in Subprojects............................................................................... 45 Figure 26 - Institutional and Structural Variables ....................................................................................... 46 Figure 27 - Project Status: Was an O&M Committee Formed? .................................................................. 47 Figure 28 - Project Status: When was the O&M Committee Formed by Construction Quality and by Project Status? ............................................................................................................................................ 47 Figure 29 - Project Status: Who Organised Labour Contributions? ............................................................ 48 Figure 30 - Who Organised Material Contributions? .................................................................................. 48 Figure 31 - Who Organised Monetary Contributions? ............................................................................... 48 Figure 32 - Project Status: Did CDC Members Provide Input During Design Phase? .................................. 49 Figure 33 - Project Status: Were CDC Members Involved in Construction Monitoring and Audit of Funds? .................................................................................................................................................................... 49 Figure 34 - Project Status: Did the CDC Receive O&M Training? ................................................................ 49 Figure 35 - On a Scale of 1-10 (1 very dissatisfied 10 very satisfied) How Satisfied are CDC Members with O&M Training? ............................................................................................................................................ 49 Figure 36 - Project Status: Are CDC Members Familiar with the O&M Plan?............................................. 50 iii Figure 37 - Project Status: Do CDC Members have Experience in Construction/Development Projects? . 51 Figure 38 - Quality of Construction: Do CDC Members have Experience in Construction/Development Projects?...................................................................................................................................................... 51 Figure 39 - Project Status: Do CDC Members have Experience Maintaining Similar Projects? .................. 51 Figure 40 - Project Status: Do you Feel Community Members have the Capacity and Technical Knowledge to Maintain the Subproject? ....................................................................................................................... 52 Figure 41 - Project Status: Does the Project Respond to an Actual Community Need? ............................. 53 Figure 42 - Percentage of Subprojects that Respond to an Actual Community Need by Subproject Type 53 Figure 43 - Subproject Status by Average Cost of Construction (AFN) ....................................................... 53 Figure 44 - Subproject Status by Cost of O&M ........................................................................................... 54 Figure 45 - Subproject Status by Technical Complexity .............................................................................. 55 Figure 46 - Project Status: Was the Subproject Constructed Properly and with Good Materials? ............ 56 Figure 47 - Project Status: Does the Project Design Match the Completed Subproject?........................... 56 Figure 48 - Subproject Status: Community Economic Rating ..................................................................... 57 Figure 49 - Project Status: Number of Households..................................................................................... 58 Figure 50 - Project Status: Average Number of Projects ............................................................................ 59 Figure 51 - Main Causes of Poor Maintenance ........................................................................................... 61 Figure 52 - O&M: Has an O&M Committee Been Formed? ........................................................................ 62 Figure 53 - O&M: When was the O&M Committee Formed? .................................................................... 62 Figure 54 - O&M: Has an O&M Officer Been Appointed? .......................................................................... 62 Figure 55 - O&M: Is there a Regular System of Collecting User Fees? ....................................................... 63 Figure 56 - O&M: Who Organises Monetary Contributions? ..................................................................... 64 Figure 57 - O&M: Who Organises Time/Labour Contributions? ................................................................ 64 Figure 58: O&M: Who Organises Material/Resource Contributions? ........................................................ 64 Figure 59 - O&M: Did CDC Members Receives some form of Training from NSP/FP Staff? ...................... 65 Figure 60: O&M: Are CDC Members Familiar with the O&M Plan? ........................................................... 65 Figure 61 - O&M: Community O&M Monetary Investment (AFN) ............................................................. 66 Figure 62 - O&M: Does the Project Respond to an Actual Community Need? .......................................... 67 Figure 63 - O&M: Is there a Regular System of Collecting User Fees? ....................................................... 68 Figure 64 - What are the Main Impediments to Contributing Funds? ....................................................... 68 Figure 65 - What are the Main Impediments to Committing Materials/Resources? ................................. 68 Figure 66 - What are the Main Impediments to Committing Time/Labour?.............................................. 69 Figure 67 - O&M: Average Number of Households .................................................................................... 69 Figure 68 - O&M: Average Number of Projects .......................................................................................... 70 Figure 69 - O&M: Ethnicity.......................................................................................................................... 71 Figure 70 - O&M: NSP Phase ....................................................................................................................... 71 Figure 71 - O&M: Technical Complexity of Subproject ............................................................................... 73 Figure 72 - O&M: O&M Cost of Subproject ................................................................................................ 74 Figure 73 - O&M: What is the Initial Construction Quality? ....................................................................... 74 Figure 74 - Institutional and Structural Variables ....................................................................................... 83 Figure 75 - Subproject Implementation Timeline ....................................................................................... 87 iv LIST OF TABLES Table 1 - District Profiles: Badakhshan ......................................................................................................... 9 Table 2 - District Profiles: Balkh .................................................................................................................. 10 Table 3 - District Profiles: Bamiyan ............................................................................................................. 11 Table 4 - District Profiles: Herat .................................................................................................................. 12 Table 5 - District Profiles: Nangarhar .......................................................................................................... 13 Table 6 - Sample Size and Locations: FGDs and KIIs ................................................................................... 14 Table 7 - Sample Size and Locations: Technical Assessments .................................................................... 15 Table 8- Preferred O&M Contributions (CDCs) ........................................................................................... 22 Table 9 - Subproject Status by Subproject Type ......................................................................................... 23 Table 10 - Canal Key Component Condition (%) ......................................................................................... 26 Table 11 - Maintenance Activities: Canals .................................................................................................. 27 Table 12 - Culvert Key Component Condition (%) ...................................................................................... 28 Table 13 - Maintenance Activities: Culverts............................................................................................... 29 Table 14 - Karez Key Component Condition (%) ......................................................................................... 30 Table 15 - Maintenance Activities: Karezs .................................................................................................. 31 Table 16 - MHP Key Component Condition (%) .......................................................................................... 33 Table 17 - Maintenance Activities: Micro-Hydro Plants ............................................................................. 33 Table 18 - Protection Wall Key Component Condition (%) ......................................................................... 35 Table 19 - Maintenance Activities: Protection Walls .................................................................................. 36 Table 20 - Shallow Well Key Component Condition (%) ............................................................................. 37 Table 21 - Maintenance Activities: Shallow Wells ...................................................................................... 38 Table 22 - Solar Panel Key Component Condition (%) ................................................................................ 39 Table 23 - Maintenance Activities: Solar Panels ......................................................................................... 40 Table 24 - Tertiary Road Key Component Condition (%) ............................................................................ 41 Table 25 - Maintenance Activities: Tertiary Roads ..................................................................................... 42 Table 26 - Water Supply Network Key Component Condition (%) ............................................................. 44 Table 27 - Maintenance Activities: Water Supply Networks ...................................................................... 44 Table 28 - Project Status by Subproject Type ............................................................................................. 55 Table 29 - Maintenance Performed by District........................................................................................... 72 Table 30 - O&M Funding Gap by Scenario (all NSP subprojects, USD per year) ......................................... 77 Table 31 - Average Community Investment (AFN) per Subproject and District ......................................... 78 Table 32 - O&M Cost: Scenario 1 ................................................................................................................ 79 Table 33 - O&M Costs: Scenario 1 (All NSP Subprojects) ............................................................................ 80 Table 34 - O&M Cost: Scenario 2 ................................................................................................................ 80 Table 35 - O&M Cost: Scenario 2 (All NSP Subprojects) ............................................................................. 81 Table 36 - O&M Cost: Scenario 3 ................................................................................................................ 82 Table 37 - O&M Costs: Scenario 3 (All NSP Subprojects) ............................................................................ 82 Table 38 - Key Findings: Institutional and Structural Variables .................................................................. 84 Table 39 - Key Findings: Subproject O&M Cost .......................................................................................... 86 v LIST OF ACRONYMS AFN Afghani (currency) CDC Community Development Council CDD Community Driven Development DC District Center FP Facilitating Partner MHP Micro-Hydro Plant MoEW Ministry of Energy and Water MoF Ministry of Finance MPW Ministry of Public Works MRRD Ministry of Rural Rehabilitation and Development NGO Non-Governmental Organisation NSP National Solidarity Program O&M Operations and Maintenance PC Provincial Center PMU Provincial Management Unit PVC Polyvinyl Chloride (Safety Tape Used by Electricians) RCC Reinforced Cement Concrete USD United States Dollar WSN Water Supply Network vi EXECUTIVE SUMMARY BACKGROUND, OBJECTIVES AND METHODOLOGY In 2003, the Ministry of Rural Rehabilitation and Development (MRRD), with the support of a broad range of donor agencies, initiated the National Solidarity Program (NSP). NSP was a Community Driven Development (CDD) programme designed (i) to provide public services and development opportunities to rural communities and (ii) to form local representative institutions called Community Development Councils (CDCs). Between 2003 and 2010, NSP implemented activities in approximately 23,000 communities throughout Afghanistan, including a large number of infrastructure subprojects. The objective of this report is to present an overview of the condition of assessed subprojects and maintenance activities performed, an analysis of variables affecting both subproject condition and O&M, and a yearly cost estimate for all NSP subprojects nationwide. To achieve these objectives, field teams consisting of engineers and interviewers assessed the condition and maintenance of subprojects through a series of technical assessments, focus group discussions with community and CDC members, and key informant interviews with Facilitating Partner (FP) staff. In total, 100 subprojects were assessed in 10 districts of 5 provinces across Afghanistan, with sampling methodology providing an effective cross-section of varying geographical and ethnic community types. KEY FINDINGS Condition and Maintenance Subproject condition and the performance of effective maintenance activities were mixed across the sample group, as is illustrated by the chart below. While more than a quarter of assessed and completed subprojects were functional and in good condition, 56% were not in good condition, and 14% were not functional at all. Similarly and also illustrated by the chart below, adequate maintenance was performed on 24% of subprojects requiring maintenance, inadequate maintenance on 17%, while 30% of subprojects assessed had no maintenance performed at all and 29% of subprojects required no maintenance to date. Subproject Condition Maintenance Activities 1 Variables Affecting Condition and Maintenance In order to isolate causes of good subproject condition and effective maintenance, a series of institutional and structural variables were examined. Institutional variables focus on the organisational structure and the technical capacity of CDCs, as well as the level of community investment, such as the creation of O&M Committees, CDC involvement in project construction and design, capacity for organising O&M activities, technical experience of CDC members, whether the project responds to an actual community need, and community contributions to the project. Structural variables focus on village and subproject level dynamics, such as security, poverty and access, as well as initial subproject construction and O&M cost, technical complexity of the subproject and quality of design and construction. Throughout the analysis, aspects of all three institutional variables were found to have an impact on both subproject condition and maintenance activities. Organisational structures such as O&M Committees, O&M Officers, fee collection schemes, and O&M plans were more likely to keep subprojects in good condition through effective maintenance due to increased institutional capacity and dedication to O&M. Additionally, CDCs with a higher level of technical capacity were more likely to ensure that subprojects were built properly using high quality materials. Finally, CDCs with subprojects that encouraged community investment either by responding inclusively to a community need, electing a transparent and accountable CDC, effectively managing community contributions, and ensuring community subproject awareness were more likely to maintain subprojects in good condition. Structural variables, though often relevant, had less demonstrable impact on subprojects than institutional variables. Communities that were ethnically homogeneous generally kept subprojects better maintained than heterogeneous communities, most likely due to the ease of organising maintenance in these communities. Nevertheless, the economic situation, the average number of subprojects, and the size of a community had no demonstrable impact on subprojects. On the other hand, more expensive subprojects, which were often also more technically complex, were more likely to be maintained as communities are required to invest a minimum of 10% of subproject construction cost. Also, if the initial construction quality of the subproject is high, the subproject is more likely to be maintained by community members. Finally, subprojects implemented under NSP II have a higher likelihood of effective maintenance than those implemented under NSP I, most likely due to increased attention paid to planning, organising, and mobilising maintenance at the community level. Subproject O&M Cost While many communities could afford to maintain low-cost subprojects such as canals, karezes, culverts, and protection walls, some of the more technically complex and expensive to maintain subprojects were found to be too expensive for certain communities to maintain without outside funding and expertise. Notably, water supply networks and tertiary roads, due to replacement costs for the often extensive pipe networks of the former and the labour and material cost of the latter, are outside of most communities’ capacity for maintenance. In addition, estimated replacement costs for key components of some micro-hydro power plants and solar panels were deemed too high for certain communities. 2 Indeed, if communities need to find paid labour for maintenance of some of the low-cost subprojects, these projects become fiscally untenable as well. CONCLUSIONS AND RECOMMENDATIONS The communities each exhibited different levels of organisational, technical, and financial capacity to implement and maintain subprojects independently, with some requiring no further involvement from the FP, other than distributing the grant for subproject construction, while some are in dire need of assistance. In practise, based on the key findings above, the majority of communities reside somewhere in between those extremes, and would benefit from some level FP assistance to improve their capacity to effectively steward sub-project assets. Moreover, very rarely are communities able to collect enough resources to cover the cost of O&M. Thus, as most communities will likely require some level of assistance to harmonise their respective capacity with that set forth in the NSP Operations Manual, the role of the FP becomes crucial in ensuring that institutions and practices tailored to the particular reality of each community are created, including adequate and ongoing training and comprehensive planning to prepare for O&M activities. Indeed, the FP must be involved both before and during subproject implementation to ensure that essential mechanisms are put into place, and also continue to follow-up with communities who require additional guidance in terms of organising labour, collecting resources, or planning for short and long- term maintenance.The following recommendations thus focus on activities to be performed during the two stages of subproject implementation, as well as the inclusion of a third stage post-project completion. During the pre-project planning phase, an initial in-depth assessment by the FP focusing on potential contributions to O&M and the technical capacity of community members, amongst other variables, should be conducted. Future activities should be based on this initial assessment. Communities should be made aware at an early stage of the necessary labour, material and monetary contributions required for the possible subprojects, which should only be selected if communities are visibly capable of completing maintenance. To facilitate this, NSP should provide more detailed community selection guidelines for FPs, including a “Community Assessment and Suitability” form that must be completed before any grants are distributed. During the subproject implementation phase, the organisational structures that were deemed to have a positive impact on subproject condition and maintenance should be instituted, such as O&M Committees, O&M Officers, and the development of a detailed and thorough O&M plan, with a clear breakdown of which community members will contribute to O&M activities and with what resources. Subproject construction should be closely monitored by the O&M Committee and Officer, working alongside FP engineers in order to gain a more complete understanding of the subproject, particularly for more technically complex subprojects. Alongside these processes, comprehensive and ongoing training should be provided to the O&M Committee and CDC members. NSP should penalise FPs that are unable to follow through with these necessary steps in subproject implementation. 3 Currently, little to no contact between the FP and communities occurs after subproject completion. Thus, in the short term (3-6 months after completion), the FP should follow up with communities, filling any gaps in training or expertise required by the CDC and community members. The FP should revisit communities on a monthly basis to address any developing issues or concerns, depending on the capacity of the community in question. During this phase, the FP can also help communities to link up with relevant line ministries, who could then take over training and maintenance supervision for more technically complex and large subprojects. To achieve these goals, NSP should plan for FPs to follow up with communities, and in some cases a percentage of the initial grant should be set aside for short term O&M costs. 4 INTRODUCTION AND METHODOLOGY INTRODUCTION In 2003, the Ministry of Rural Rehabilitation and Development (MRRD), with the support of the World Bank and a broad range of other international donor agencies, initiated the National Solidarity Program (NSP). NSP was a Community Driven Development (CDD) programme designed (i) to provide public services and development opportunities to rural communities and (ii) to form local representative institutions called Community Development Councils (CDCs). Between 2003 and 2010, NSP implemented activities in approximately 23,000 communities throughout Afghanistan. In 2010, the programme entered in its third phase, which aims at covering approximately 17,200 newly included communities for a first round of block grants and approximately 10,500 already included communities for a second round of block grants. Activities are implemented on the ground through Facilitating Partners (FPs), NGOs or International Organisations NSP is the largest development programme in the country, with donor contribution totalling more than US$1.5 billion as of 2010. The proposed budget for NSP-III from September 2010 to September 2015 is nearly US$1.5 billion, or roughly US$300 million per year. This proposed annual budget represents about 20 per cent of the country’s total annual development budget. A largely unasked and unanswered question regarding NSP is the cost of maintaining subprojects built over the past 6 years. Although many of the subprojects constructed require little to low maintenance which can be performed and funded by communities, others, such as tertiary roads, water supply networks, and micro hydro plants require often costly or technically complex labour or replacement parts. The purpose of this subproject sustainability study is to assess the condition of assets built under NSP as well as maintenance activities performed to date, to understand the underlying factors that influence the condition of subprojects, and to estimate the annual O&M cost for each type of asset. Using this information, an O&M budget was estimated for all assets built under NSP, identifying any gaps in funding and highlighting constraints to effective O&M expressed by community and CDC members. Possible solutions and recommendations, based on interviews with community members and Facilitating Partners (FPs) were then provided at the end of the report. This study complements the ongoing ARTF third-party monitoring which assesses NSP subprojects from an engineering perspective. The subproject sustainability assessment is one part of a two-part study conducted by Altai Consulting for the World Bank and MRRD-NSP. The second study focuses on governance and sustainability of the CDCs. APPROACH The objective of this report is to present an overview of the condition of assessed subprojects, an overview of maintenance activities performed, and an analysis of variables affecting both subproject 5 condition and O&M (through a framework illustrated in Figure 1 below). The first section, “Subproject Status and O&M”, focuses on subproject condition and O&M activities, and presents an overview of the subprojects assessed including brief descriptions of subproject types alongside an overview of O&M activities performed to date on the assessed subprojects, including how O&M is organised and, where applicable, funded. The second section, “Variables Influencing Subproject Condition and O&M”, is an analysis of institutional and structural variables tested against subproject condition and O&M activities. Institutional variables are divided into three categories: Organisational Structure: includes the way in which a CDC organises maintenance, how involved they are in the subproject implementation and O&M, whether a specialised O&M committee has been formed, and whether an O&M plan has been established. Technical Capacity: includes previous experience in construction and repairs, previous experience with maintaining a similar subproject, and quality and type of training received from the FP. Community Investment: includes the subproject’s level of response to a community need, the transparency and accountability of the CDC, community awareness of the subproject’s importance, and the level of investment in O&M activities. Structural variables are divided into two categories: Community-Based: the geography and topography of the village, the levels of security and poverty, the size of the community, and the number of subprojects the community has constructed. Subproject-Based: The initial and O&M cost of the subproject, number of other NSP subprojects constructed, date of construction (NSP I or NSP II), the technical complexity of the subproject, the quality of initial construction, and fidelity to initial project design. 6 Figure 1 - Analysis Framework Organisational Structure Subproject Community- Condition based Technical Institutional Structural Capacity of CDC Variables Effective Variables Subproject- Performance of based Community O&M Investment Recommendations Policy Options for Effective Subproject Sustainability The main distinction between institutional and structural variables is that the former may be more easily manipulated in order to improve subproject condition and O&M activities. For example, NSP could encourage FPs to institute O&M committees in all communities, or to improve the quality of O&M training received, while improving the economic situation of a community or reducing the construction cost of subproject is more difficult. In this sense, institutional variables are actionable, and can be changed through a set of policy decisions, while structural variables effecting subproject condition and O&M activities are more static: the only way to control them is through the selection of communities and subproject types. The third section of the report presents a series of O&M cost scenarios based on assumptions such as frequency of maintenance, attrition rates, and replacement costs of key components. Scenarios are provided for the sample group and also for all relevant subprojects completed under NSP in order to estimate total O&M cost. METHODOLOGY PROJECT OBJECTIVES The subproject sustainability study aimed to carry out a systematic assessment on:  The functionality of the completed sub-projects;  Types and costs of O&M activities that CDCs are providing;  Total funding gap; 7  Underlying factors that influence local decisions to maintain the sub-projects;  Identification of key constraints that prevent CDCs from providing adequate O&M. GEOGRAPHICAL SCOPE The study was conducted in 100 villages in 10 districts (10 villages per district) in Northern (Badakhshan), Central (Bamiyan and Balkh), Eastern (Nangarhar), and Western (Herat) Afghanistan, as illustrated by Figure 2 below. Figure 2 - Geographical Scope of Study Provinces were selected not only to provide as much of a geographical cross-section of Afghanistan as possible, but also to highlight different types of terrain including valleys, mountains and rivers. Secure villages closer to the district centre were prioritised due to timeline constraints and team safety. Sampling methodology was contributed by an external consultant to the World Bank. OVERVIEW OF LOCATIONS Although surveyed districts were representative of the geographic (flatlands, plains, valleys, mountains, and varying degrees of access to water) and economic diversity of Afghanistan, they were by and large relatively secure and easily accessible, with the exception of Zindajan district in Herat where heightened insurgent activity was reported by field interviewers. BADAKHSHAN Badakhshan province in Northern Afghanistan is a mountainous region with multiple valleys and rivers and Table 1 offers a district profile of the province. The province’s difficult terrain increases the cost and the technical complexity of construction and infrastructure projects. The two surveyed districts in 8 Badakhshan, Baharak and Jorem, are moderately poor, living mostly on subsistence agriculture. Surveyed residents in Baharak reported contributing on average 11,300 AFN per year to O&M, while those in Jorem reported contributing 6,950 AFN. Both contributions are well below the surveyed average of 14,100 AFN per year. Although roads from Faizabad, the provincial capital, to district centres were in excellent condition, access roads from the DC to villages were in very poor condition. Villages visited in both districts were primarily Tajik, although four villages were ethnically heterogeneous. Two exceptions were Shash Gul in Jurm (Uzbek) and Do Abgi in Baharak (Pashtun). Table 1 - District Profiles: Badakhshan Badakhshan Profile Subproject Economic Security Number of District Village 1 2 3 Households Assessed Rating Rating Subprojects Chapchi Bala MHP 2 1 4 125 Chapchi- Payan Culvert 2 1 3 150 Deh Ta WSN 4 1 1 83 Do Abgi MHP 4 1 1 55 Doab MHP 4 1 1 212 Baharak Hatam Beki Culvert 2 1 1 92 Khoshdrew Tertiary Road 5 4 2 173 Madrasa Khair Abad MHP 3 1 1 118 Masjed Haje Sayd Abdul Ghaias Culvert 2 1 1 84 Obkhora MHP 4 1 1 75 Baba Qasim Culvert 4 1 4 96 Dah-e-Payen Naw MHP 3 1 1 198 Dara Pushkan Tertiary Road 3 1 1 174 Madrasa Awan MHP 3 2 1 119 Pata Ghozar Tertiary Road 3 1 1 142 Jurm Sari Hous Protection Wall 4 1 2 204 Shah Gul Culvert 2 1 2 119 Sina Canal 3 1 1 178 Walarib Canal 3 2 1 153 Yabab Canal 2 1 3 270 BALKH Balkh province is located in the north of Afghanistan and shares a border with Uzbekistan, including one of the country’s most active border crossing points at Hairatan. Table 2 offers district profiles for the province. Strong trade and relative stability have spurred economic growth in Balkh, and as such several districts are relatively wealthy compared to the national standard. Geographically, Balkh is composed mostly of flatlands and although there are some areas where the security situation is poor, most of the province is relatively stable. Of the two surveyed districts, Nahr-e Shari was by far one of the wealthier surveyed across the nation, with communities contributing on average 60,000 AFN per year to O&M, 1 Economic rating is on a scale of 1 to 5, with 1 being very wealthy and 5 being very poor. 2 Security rating is on a scale of 1 to 5, with 1 being very secure and 5 being very insecure. 3 Number of subprojects constructed by NSP, taken along with data on households from the NSP Completed Subproject Database. 9 while Khulm was one of the poorest, with villagers contributing only 4,050 AFN per year. Villages in Khulm district were mostly Tajik, whereas those visited in Nahr-e Shari had mixed populations composed of Tajiks, Hazaras, Pashtuns, and Uzbeks. Table 2 - District Profiles: Balkh Balkh Profile Subproject Economic Security Number of District Village Households Assessed Rating Rating Subprojects Alimardan Bik Shallow Well 3 1 3 202 Chardarya Protection Wall 3 1 2 145 Chatgary Culvert 3 1 2 105 Deh Hassan Shallow Well 4 1 3 188 Guzar Sert Canal 2 1 3 192 Khulm Khanaqa Mirza Khairuddin Protection Wall 4 1 2 113 Logariha Solar Panel 5 2 1 84 Mulla Sultan Tangi Shallow Well 2 1 2 140 Naw Abad Shallow Well 4 1 2 62 Shahi Khail Arabia Solar Panel 4 2 1 78 Islamabad WSN 2 1 1 1001 Langar Khona Khord Tertiary Road 1 1 5 1050 Naw Abad Torkman Ha WSN 1 1 4 309 Noor Khoda Dohum Tertiary Road 1 1 4 1327 Nahr-e Pashtoon Kot WSN 3 1 3 789 Shahi Qala Qol Mohammad WSN 3 1 3 308 Qezel Abad Tertiary Road 3 1 2 174 Sajadia Culvert 3 1 3 964 Takhta Pul WSN 2 1 3 222 Yakatot Tertiary Road 1 1 2 175 BAMIYAN Bamiyan province is located in central Afghanistan and is characterised by mountainous valleys, similar to Badakhshan. As such, the cost and technical complexity of certain infrastructure projects is significantly increased. Table 3 outlines the district profiles of the province. Bamiyan and Shibar districts were two of the poorer districts in the sample group, with average community O&M contributions of 6,625 and 6,875 AFN per year, respectively. Access to many villages was severely restricted due to the poor quality of roads. Subprojects assessed were in villages composed entirely of Hazaras. 10 Table 3 - District Profiles: Bamiyan Bamiyan Profile Subproject Economic Security Number of District Village Households Assessed Rating Rating Subprojects Chap Dara Solar Panel 4 1 1 110 Dahan-e-Soumara Shallow Well 4 1 3 66 Dow Ab zard sang MHP 2 1 1 51 Dowabe Chabdara Tertiary Road 4 1 5 230 Fatmasti Shallow Well 3 1 4 125 Bamyan Gorwan-i-Bala Shallow Well 3 1 3 124 Haider Abad MHP 3 1 3 477 Jar soghee Culvert 1 1 2 44 Pitap Laghman Culvert 5 1 2 26 Zakareya MHP 3 1 1 90 Bar Galij Payin Shah Asker Solar Panel 3 1 1 205 Birgilich bala Solar Panel 4 1 1 215 Dahan Khoushkak Solar Panel 3 1 1 56 Dashti Qala MHP 3 1 2 62 Dashti Sayd Culvert 3 1 3 152 Shibar Eraq Ulia Solar Panel 3 1 2 63 Ghorbandak Tertiary Road 5 1 2 36 Ghulam wa sheia Solar Panel 5 1 2 57 Gounbod Tertiary Road 3 1 7 165 Jundar Qalai Bala Tertiary Road 4 1 2 54 HERAT Herat province is located in western Afghanistan, shares a long border with Iran and is mainly composed of flatlands and plains. Table 4 outlines the district profiles of the province. Injil and Zindajan were two of the wealthier districts surveyed, with reported villages contributing on average 13,750 AFN per year in both locations. Field teams also noted that CDC members and villagers were more educated and literate than in other provinces. Access to Zindajan district from the provincial centre was complicated due to a heightened insurgent presence and the poor conditions of the roads. Villages visited in Herat were entirely composed of Tajiks. 11 Table 4 - District Profiles: Herat Herat Profile Subproject Economic Security Number of District Village Households Assessed Rating Rating Subprojects Chashmaha Karez 4 3 2 130 Ehsaq Suliman WSN 2 1 2 666 Farahyan WSN 3 1 1 124 Frizan Sofla WSN 2 1 4 227 Ghaizan Cheshma Canal 2 1 2 1083 Injil Ghaizan Sangar Canal 3 1 3 675 Haft Qulb Tertiary Road 3 1 2 65 Kabrzan Protection Wall 4 1 5 721 Rabat -e- Sarwistan Shallow Well 3 1 4 140 Rakhna Protection Wall 3 1 2 160 Ali Chap Karez 5 3 3 184 Golow Ga Karez 4 4 2 78 Khan Mullah Karez 4 3 2 81 Kushkak Karez 4 4 7 347 Qala-e-Nayeb Karez 3 1 2 106 Zindajan Sang Tarkida Shallow Well 4 3 3 152 Sanjab WSN 3 1 4 259 Sarboland Karez 5 1 1 58 Sari Naw Protection Wall 4 3 2 119 Taghi Naghi Karez 4 2 2 470 NANGARHAR Nangarhar province is located in eastern Afghanistan and is mostly composed of flatlands with a mountain chain in the west dividing it from Kabul. Table 5 provides district profiles for the province. Villages in Behsood and Sukhrood districts ranged from moderately wealthy to moderately poor and contributed on average 8,750 and 9,300 AFN per year to O&M, respectively. Villages surveyed in Sukhrood were atypical of Nangarhar in that they are primarily Tajik, while villages in Behsood were mostly Pashtun and Arab, with some Tajiks. Villages in both districts were easily accessible, with the furthest being one hour from the district center. 12 Table 5 - District Profiles: Nangarhar Nangarhar Profile Subproject Economic Security Number of District Village Households Assessed Rating Rating Subprojects Aklas Abad Khail Shallow Well 3 1 3 153 Bardaman Protection Wall 5 1 4 302 Ghondai Malek Mashal Solar Panel 2 1 2 166 Ghundi kariz kabir Canal 2 1 3 310 Hesi Awal Najmul jihad WSN 2 1 2 328 Behsood Loy Khosh Gunbad Culvert 2 1 4 458 Muqam Khan Tertiary Road 4 1 2 330 Nahar Masi Canal 3 1 3 300 Nahar Shahi Dohum Solar Panel 3 1 2 350 Saracha Ali Khan Canal 2 1 3 310 Bagh Baha Wali MHP 2 4 3 180 Bala Qala MHP 4 3 1 240 Bar Nazar Abad Canal 2 1 4 262 Bemaran Qalai Abdullah Solar Panel 3 1 2 235 Charbagh Malakan Shallow Well 3 1 2 240 Sukhrood Charbagh Qala Painda Solar Panel 2 1 2 102 Daronta Canal Tertiary Road 1 1 3 351 Daronta Chapa Darya Protection Wall 1 1 3 212 Deh Bawal Payen Culvert 3 1 3 114 Dehbawal Bala Karez 2 1 5 102 TOOLS AND SAMPLE GROUP Five tools were designed in order to respond to the project research objectives outlined above: FOCUS GROUP DISCUSSION O&M COMMITTEE/CDC QUESTIONNAIRE (FGD O&M) FGD O&M questionnaires were used to interview members of the CDC O&M committee or regular CDC members responsible for O&M of subprojects in communities where O&M committees had not been formed. The objective of the FGD O&M was to understand the contributions, constraints, and O&M activities carried out by the CDC, as well as to determine their level of satisfaction with the subproject in question and FP training and support. Sample Size: 400 respondents FOCUS GROUP DISCUSSION HOUSEHOLD QUESTIONNAIRE (FGD HH) FGD HH questionnaires were used to interview regular community members who were neither CDC members nor members of traditional power structures (Maliks, commanders, gharyadars, mullahs, etc.). The objective of the FGD HH questionnaire was to determine the level of community contribution to O&M (labour, resources and funding), key constraints to effective O&M and satisfaction with CDC performance Sample Size: 400 respondents 13 Both FGD questionnaires included mostly closed-ended, multiple choice questions. FP SOCIAL ORGANISER KEY INFORMANT INTERVIEW (FP SO KII) FP Social Organisers were interviewed using the FP KII questionnaire in order to determine FP post- completion community involvement/activity, level of support from third-party institutions (line ministries, other donors, PRTs, etc.), quality and type of O&M training provided by the FP, and constraints to effective O&M. Sample Size: 8 respondents FP ENGINEER KEY INFORMANT INTERVIEW (FP ENG KII) FP Engineers were interview using the FP Eng KII questionnaire in order to determine the level of experience of the engineer, quality of initial design/construction, maintenance training provided, and constraints to effective O&M. Sample Size: 8 respondents Table 6 outlines the sample sizes and locations of both FGDs and KIIs conducted in the study. Table 6 - Sample Size and Locations: FGDs and KIIs FGD (O&M and Household) FP KIIs (SO and Eng) Province Districts Communities Respondents FPs Respondents Baharak 10 80 1 2 Badakhshan 2 Jurm 10 80 1 Khulm 10 80 1 2 Balkh 2 Nahr-e Shari 10 80 1 Bamyan 10 80 1 2 Bamiyan 2 Shibar 10 80 1 Injil 10 80 1 2 Herat Zindajan 10 80 Behsood 10 80 1 2 Nangarhar Sukhrood 10 80 TOTAL 100 800 8 16 TECHNICAL ASSESSMENT TOOL (TA) The TA tool was used by the team of engineers to assess the condition and functionality of completed subprojects, the status of their key components, as well as maintenance activities performed. Subprojects included canals, micro hydro plants, solar panels, water supply networks, tertiary roads, shallow wells, culverts, protections walls, and karezes. 14 Three classifications were used to distinguish subprojects: functional and in good condition (all key components properly maintained), functional and in poor condition (at least 1 key component not properly maintained), and non-functional. A subproject is deemed functional when it is capable of providing its intended benefit to the community in question (e.g. a canal that irrigates farm land or a micro-hydro plant that provides electricity). Key components are defined as parts without which a subproject cannot function, or whose absence would cause the condition of the subproject to rapidly deteriorate. Three categories were also used to distinguish subproject maintenance: adequate (all required maintenance activities performed), not adequate (at least one required maintenance activity not performed) and not necessary to date (not maintenance activities require to date). Sample Size: 100 subprojects Table 7 outlines the sample size and locations of technical assessments made in the study. Table 7 - Sample Size and Locations: Technical Assessments Technical Assessments Province Districts Micro Water Solar Tertiary Shallow Protection Canal Hydro Supply Culvert Karez Panel Road Well Wall Plant Network Jorem 3 2 0 0 2 0 2 1 0 Badakhshan Baharak 0 4 0 1 1 0 3 1 0 Khulm 1 0 2 0 0 4 1 2 0 Balkh Nahr-e Shari 0 0 0 5 4 0 1 0 0 Shibar 0 1 5 0 3 0 1 0 0 Bamyan Bamyan 0 3 1 0 1 3 2 0 0 Zindajan 0 0 0 1 0 1 0 1 7 Herat Injil 2 0 0 3 1 1 0 2 1 Behsood 3 0 2 1 1 1 1 1 0 Nangarhar Sukhrood 1 2 2 0 1 1 1 1 1 TOTAL 10 12 12 11 14 11 12 9 9 O&M COSTING MODEL An O&M Costing Model was developed to estimate the O&M costs of all NSP subprojects. Community- based monetary contributions (which varied greatly depending on the district surveyed) will be deducted from this cost, showing in many cases a gap in available O&M funds. Several assumptions were made when building the O&M costing model. First, materials used, subproject design, and construction are all assumed to be of the highest quality. Second, attrition rates assume that asset key components are properly maintained. 15 Information on attrition rates was supplied by the engineering team and key component replacement costs were determined by a survey of suppliers in provincial capitals. Where possible, both high and low estimates are given. Assumptions made during O&M cost estimations, as well as the overall approach, will be further developed in the section “Subproject O&M Cost and Funding Gap.” CONSTRAINTS AND LIMITATIONS SECURITY AND ACCESSIBILITY Sample locations were chosen based on ease of access and good security, due to time constraints and team safety, respectively. As such, while sample locations are relatively well distributed, the least accessible and most-security constrained areas are not well represented in the study. As such, the estimations for O&M costs should be viewed as a lower range estimate, given that security and accessibility will only increase the logistical difficulties and ultimate costs of subproject maintenance. Furthermore, Afghan FPs could not be included in the sample group as they were operating in insecure and difficult to access districts. Security threats occurred at the border between Jurm and Warduj districts and throughout Zindajan but they did not disrupt data collection. INCLUSION OF WOMEN As the teams were composed of male members, access to women remained a strong limitation despite best efforts to engage with female community members. No women were interviewed in Nangarhar, while only 8 female CDC members could be interviewed across the remaining four provinces. No women were interviewed for the household-level focus group discussions. SAMPLE SIZE AND LOCATIONS Subproject types constructed and thus eligible for assessment in given districts/provinces were limited. As a result, certain provinces have more of a certain type of subproject. As subproject condition, maintenance activities, and especially cost of maintenance depend heavily on the type of subproject in question, location was not taken into account during quantitative data cross-analysis. Nevertheless, variables such as geography, poverty and accessibility will be discussed in a more qualitative fashion in the section “Factors Influencing Subproject Condition.” 16 SUBPROJECT STATUS AND O&M The following section begins with a brief introduction to subprojects, including construction process, descriptions of intended benefits, and an overview of maintenance activities. Following this, a breakdown of subproject condition and maintenance for the assessed sample, including key component status, most frequent causes of poor condition, and maintenance activities performed, is provided. SUBPROJECT CONSTRUCTION PROCESS The subproject construction process follows the following three phases outlined in the NSP Operations Manual.4 Subproject Proposal Preparation: The FP works with elected CDC members in order to prepare a proposal for subprojects that best address community needs. While FP engineers are responsible for preparing the technical subproject proposal and design, CDC members are meant to be trained and involved in the process in order to increase their capacity for project management. The FP ensures that the proposed subproject meets NSP technical requirements, and helps the community to prepare a realistic O&M and cost recovery plan. They also facilitate the formation of an O&M Committee within the community, responsible for maintaining the subproject. The mandatory minimum community contribution required for an NSP-financed subproject is 10% of the subproject total costs. Subproject Proposal Approval and Initial Grant Disbursements: The proposal is reviewed by the NSP PMU in order to ensure that:  The subproject is actually endorsed by the CDC  The subproject provides equitable access to benefits  All reasonable administrative expenses are included in the subproject budget  The subproject is technical and financially sound  The proposal includes an O&M plan where required which also includes cost recovery components where necessary  The proposal includes all the required approvals from the line ministries’ provincial departments Once the proposal is approved, the initial grant is distributed to the CDC. Subproject Implementation: The CDC is responsible for the subproject implementation with technical assistance, monitoring and oversight provided by the FP engineers and social organisers. The CDC must 4 The phases listed below include many requirements and checks designed to ensure subproject quality and appropriateness, as well as CDC training and preparation. Nevertheless, many tasks that were the responsibility of the FP were not adequately performed. For example, in some sample communities subprojects were built using sub-par materials and did not adhere to the initial design, while only 34 out of 100 CDCs reported having O&M Committees. 17 provide updates to the NSP PMU at various stages of project completion, and is required to report progress and budgets on a monthly basis to the community. The timely and satisfactory completion of the subprojects is the joint responsibility of the FP and the CDC. Once the subproject is completed, the FP hands over full control of the project to the CDC, who is then responsible for all O&M activities. CONTRIBUTIONS TO O&M COMMUNITY-BASED CONTRIBUTIONS During FGD HH surveys, respondents were asked how many community members could afford to commit funds, time/labour and materials/resources to O&M activities. They were also asked to explain how these activities were organised, either by CDCs, O&M Committees, village elders, the FP, or other authority figures. In all cases except funding, the majority of respondents reported that at least most of their fellow community members could contribute to O&M activities, which were in the majority of cases organised by the CDC or O&M Committee. FUNDING Funding contributions to O&M involve community members providing money for O&M activities. Funding is collected in three ways5: User Fees: Community members pay in exchange for a service, such as the provision of electricity through solar panels or MHPs or the provision of water through water supply networks. Fees are not set at any specific point in subproject implementation and are only collected from community members who can afford to pay. Poor community members are not obliged to pay a user fee to benefit from a subproject. Regular Collection: Community members contribute money regularly to a collections box. Funds are set aside and used for routine or periodic maintenance. Funds are also used in the event of an unforseen subproject breakdown caused by the failure of a key component or a natural disaster such as flooding. Spontaneous: Community members contribute additional funds when a project breaks down or loses functionality. This kind of contribution occurs if there is no money set aside in the collections box or if additional money must be raised following project breakdown or a natural disaster. When funding is collected, 45% of villagers surveyed reported that most of the community contributes, while 39% reported that few commit funds and 16% reported that no one contributes funds to O&M activities (illustrated in Figure 3). Funding contributions, as illustrated in Figure 4 are usually organised by the CDC or the O&M Committee; however in some communities contributions are organised by village officials or other authority figures (usually FP representatives). 5 Any one or a combination of these three methods may be in use in a given village. 18 Figure 3 - How Many Community Members Contribute Funding to Figure 4 - Who Organised Monetary Contributions? 6 Maintenance? The majority (63%) of villages surveyed made no monetary contributions to subproject O&M. Exceptions included certain villages in Baharak, Khulm, Shibar, Behsood, where MHPs and water supply networks received the most amount of funding, likely due to user fees collected for service provision and the overall cost of maintaining these subprojects. Figures 5 and 6 provide a breakdown of O&M contributions according to subproject type and district, respectively. Figure 5 - Yearly O&M Contributions per Village (AFN) by Subproject Type Yearly O&M Contribution by Subproject (AFN) Subproject Type None Up to 20,000 Up to 50,000 Up to 75,000 More than 75,000 WSN 36% 27% 9% 2% 25% MHP 23% 38% 23% 15% 0% Canal 90% 0% 10% 0% 0% Tertiary Road 83% 8% 0% 0% 8% Protection Wall 72% 28% 0% 0% 0% Karez 89% 0% 0% 0% 11% Shallow Well 40% 55% 5% 0% 0% Culvert 83% 17% 0% 0% 0% Solar Panel 57% 43% 0% 0% 0% AVERAGE 63% 24% 6% 2% 5% 6 For the charts in this section, only completed subprojects requiring maintenance are included in the analysis. The sample group is composed of 272 respondents (for both the FGD HH survey and the FGD CDC survey) from 68 communities. 19 Figure 6 - Yearly O&M Contributions per Village (AFN) by District Yearly O&M Contribution per Village (AFN) District None Up to 20,000 Up to 50,000 Up to 75,000 More than 75,000 Baharak 50% 40% 5% 10% 0% Jorem 70% 20% 0% 10% 0% Khulm 40% 60% 0% 0% 0% Nahr-e Shari 50% 20% 0% 0% 30% Shibar 53% 38% 10% 0% 0% Bamyan 70% 20% 10% 0% 0% Zindajan 80% 10% 0% 0% 10% Injil 80% 0% 10% 3% 8% Behsood 63% 25% 25% 0% 0% Sukhrood 72% 9% 22% 0% 0% AVERAGE 63% 24% 6% 2% 5% LABOUR Labour contributions occur when basic cleaning or maintenance work is required for subprojects such as canals, karezes, protection walls, tertiary roads, and culverts. This usually involves cleaning the trench of a canal or a karez, removing debris at the base of protection walls and culverts, or removing large rocks and stones from tertiary roads. 61% of FGD HH respondents reported that at least most of the community could contribute labour to maintenance activities, while 24% reported that only a few members of their communities could help with labour. 15% reported that no one in their communities could commit labour to O&M (illustrated in Figure 7). As with labour activities, the collection of funds is usually organised by the CDC or the O&M Committee (shown in Figure 8). Figure 7 - How Many Community Members Commit Time/Labour to Figure 8 - Who Organises Time/Labour Contributions? Maintenance? MATERIALS OR RESOURCES Material or resource contributions involve the donation of materials in order to repair or maintain a subproject. Examples include the donation of gravel to repair a tertiary road, cement to repair cracks in 20 a culvert or protection wall, small parts for MHPs and water supply networks, or other key components such as batteries for solar panels. 61% of FGD HH respondents reported that at least most community members could commit materials or resources to O&M when required, compared to 29% who reported that few could commit resources. Only 11% of respondents reported that no community members in their village could commit resources to subproject maintenance (shown in Figure 9). The CDC or O&M Committee was the organisation responsible for organising maintenance activities involving the donation of materials or resources (Figure 10). Figure 9 - How Many Community Members Commit Figure 10 - Who Organises Material/Resource Contributions? Materials/Resources to Maintenance? HOW DO CDC MEMBERS PREFER TO CONTRIBUTE TO O&M ACTIVITIES? The majority (66%) of CDC members reported that they prefer to organise material or resource contributions to O&M. 30% reported that they preferred to organise labour contributions (as shown in Table 8). The high percentage of respondents preferring to contribute materials or resources may be due to the ease of organising such contributions, compared to organising labour. Collecting and convincing a group of community members with varying time-schedules to commit labour over a specific period of time requires more organisational effort than asking a few people to donate materials for maintenance activities. In addition, in some cases labour for maintenance is not free and requires a monetary investment on the part of the CDC. 21 Table 8- Preferred O&M Contributions (CDCs) Preferred O&M Contributions Subproject Type Technical Labour Funding Materials Expertise WSN 17% 0% 83% 0% MHP 59% 0% 36% 5% Canal 20% 0% 80% 0% Tertiary Road 9% 2% 88% 0% Protection Wall 25% 0% 71% 4% Karez 17% 4% 79% 0% Shallow Well 50% 3% 42% 5% Culvert 14% 2% 81% 2% Solar Panel 50% 0% 40% 10% AVERAGE 30% 1% 66% 3% OTHER SOURCES OF FUNDING OR SUPPORT LINE MINISTRIES AND OTHER DONORS Only one CDC, from Nayeb village in Zindajan district, reported having received support on O&M from line departments, in the form of technical assistance. FP engineers explained that although line departments must often sign off on subprojects, promising their support, due to budgetary and organisational constraints they rarely ever contribute funding, materials, or labour to maintenance activities. FP representatives explained that MAIL and MRRD were the most active line departments, providing mostly training and technical expertise to CDCs and their communities. Indeed, although line ministries must sign off on each subproject, only in one community out of one hundred did they actually contribute to O&M. As more costly and technically complex subprojects begin to deteriorate, the lack of O&M support for subprojects becomes a serious issue. Especially for roads, but also for MHPs and water supply networks, line departments should become involved in monitoring and maintaining subprojects built through NSP. Linking CDCs with district government authorities and laying out an appropriate O&M budget per subproject could be one solution to subproject sustainability. OVERVIEW OF CONDITION AND MAINTENANCE TECHNICAL ASSESSMENTS The engineering team was responsible for determining the condition of subprojects, as well as maintenance activities performed. Subprojects were categorised in three ways through a series of physical inspections: functional and in good condition (functional and all key components in good condition), functional and in poor condition 22 (functional and at least one key component in poor condition), and not functional.7 If the subproject was non-functional, the engineers would then determine the cause through physical evidence and discussions with CDC members. Informal interviews were also conducted with community members to confirm maintenance activities and garner further information about the subproject including how and when it was built, what materials were used, and the initial quality of construction. Engineers also recovered initial subproject designs from the FP and compared these to the constructed subproject. SUBPROJECT CONDITION The majority of subprojects were functional but not in good condition. In total, as Table 9 shows, 21% of subprojects were functional and in good condition, either because of good construction quality, adequate maintenance, or a combination of these two factors. 20% of subprojects were not functional due to natural disasters such as floods having destroyed the subproject, the breakdown of a key component that had yet to be replaced, or insufficient operational funding (the case of several water supply networks). Table 9 - Subproject Status by Subproject Type % functional % functional Quality of Type and in good and in poor % not functional Construction condition condition Good Poor Canal 71% 14% 14% 50% 50% Culvert 36% 36% 27% 55% 45% Karez 72% 28% 0% 78% 22% Micro Hydro Power Plant 0% 92% 8% 62% 38% Protection Wall 38% 62% 0% 75% 25% Shallow Well 9% 64% 27% 73% 27% Solar Panel 0% 100% 0% 83% 17% Tertiary Road 0% 93% 7% 21% 79% Water Supply Network 36% 18% 45% 91% 9% AVERAGE 29% 56% 14% 64% 36% In addition to functionality, the TA team also assessed the status of subproject key components. An initial list of key components was provided by the WB and was complemented with input from the engineers hired by Altai to conduct the technical assessments. The most common cause of poor key component condition, as Figure 11 shows, was lack of O&M, for example when the trench of canal was not properly cleaned by the community or when hand pumps were not replaced in water supply networks and shallow wells. Poor construction and poor materials were also reported as drivers of poor subproject condition. Examples include a protection wall whose foundation was not properly built, 7 Only 1 subproject out of 100 was reported as being functional but not in use. 23 another protection wall that was not built with proper grade cement, and water supply network with pipes that were too small and not made of good quality polyethylene. Figure 11 - Most Frequent Cause of Poor Key Component Condition O&M PERFORMED Maintenance on subprojects was classified under two categories: routine and periodic. Routine maintenance is performed frequently on a regular schedule set in the O&M plan. Periodic maintenance may be performed on a regular basis with larger gaps in time between maintenance (re-gravelling of tertiary roads or replacement of a battery, for example), or as a response to an unexpected key component breakdown. Subproject key components that were not completed during the construction phase were not included in this analysis. To determine maintenance activities, the engineering team began by physically inspecting the subprojects, identifying any key components that were in poor condition due to lack of O&M. This initial assessment was complemented and qualified through discussions with CDC members and villagers, who provided details of maintenance activities performed to date. Of 71 subprojects requiring maintenance, only 24 were adequately maintained, although 17 were somewhat maintained. While they are included in Figure 12 below, the remaining subprojects requiring no maintenance to date are not included in the analysis presented below, as maintenance activities do not apply. 24 Figure 12 - Maintenance Performed: Overall SUBPROJECT SPECIFIC CONDITION AND MAINTENANCE ACTIVITIES CANALS8 OVERVIEW Irrigation canals are channels for directing the flow of water. The first step in building a canal is to assess the water requirement of the area in need of irrigation. Once this is done, canal design must take into topography measurements to determine the correct ratio of inertia to gravity forces used to define erosion and silting levels, essentially regulating the flow of water to prevent excessive erosion or siltation, the two most common problems with canals, generally caused by poor construction and design. In addition, canals may be lined in order to prevent seepage and erosion or to reinforce weak canal banks. Of the ten canals the engineering team assessed across locations shown in Figure 13, one canal was not linked up to a water source and was non-functional, three were not completed due to lack of budget, five were functional and in good condition, and one was functional but in poor condition due to poor design and construction quality. 8 th Technical information on subprojects was mostly taken from the 5 edition of the NSP Engineering Manual. 25 Figure 13 - Assessed Canal Locations KEY COMPONENT CONDITION Key components of canals such as the intake and the trench, as Table 10 shows, were usually adequately maintained and cleaned, although generally only community members who benefitted most from irrigation took part in maintenance activities. The most common cause of poor key component condition was poor construction. For example, in Sina, a community in Jorem district, the canal was not built properly due to lack of budget and had no stone masonry or lining. Table 10 - Canal Key Component Condition (%) Canals Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Lining of Canal 60% 40% Poor Construction Intake (Weir) 50% 50% Poor Construction Trench 70% 30% Poor Construction O&M ACTIVITIES Average Yearly Maintenance Costs: 400 USD (High), 40 USD (Medium), 0 USD (Low) Although canals were generally well maintained, as Table 11 illustrates, in some cases labour-based maintenance activities were not performed. For example, trenches in 2 out of 6 canals assessed and an intake in one canal were not cleaned as is required. As canal cleaning depends entirely on available labour and not funding, these activities can and should be regularly completed by the communities. Field teams noted that for canals and other irrigation subprojects, only those community members who directly benefitted from the subproject would contribute time and labour to maintenance activities. Convincing other community members to assist in maintenance often required payment. 26 Table 11 - Maintenance Activities: Canals Canals Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Cleaning of Trench 4 2 4 Routine Cleaning of Intake 0 1 5 Relining of Canal 0 1 8 Periodic Repair of Intake 0 2 4 CULVERTS OVERVIEW The function of a culvert is to convey surface water across or away from a road. In addition to this it must be able to carry traffic and the weight of earth load. Culverts with circular passages are the most common, although rectangular culverts are more adaptable to various degrees of headwater, and both plain cement concrete and RCC were used in culvert construction under NSP. Supporting structures for culverts include head and wing walls, which serve to retain the fill materials and reduce erosion, improve hydraulic efficiency, provide structural stability, and provide security of traffic. Culvert construction follows a series of steps involving compaction of the abutment bed, construction of support structures such as head and wing walls, cement or RCC work, stone masonry work, and back- filling and final compaction. KEY COMPONENT CONDITION Of the twelve culverts assessed the design team, five were functional and in good condition, three were functional but in poor condition, four were not functional. In addition, one culvert was not completed due to lack of budget, despite being reported as complete. Good functionality and condition generally depended on the initial design and construction quality: in one village the culvert base was below the drainage level and water had begun to collect at the base of the culvert; in another, the wing walls needed to be heightened; several others were not built with proper grade concrete; others did not have parapet walls leading to water collecting at the base of the culvert. One culvert in Hatam Biki village in Baharak was built using only one large, precast slab and could not be cleaned. 27 Figure 14 - Assessed Culvert Locations Khulm : 1 Baharak: 3 Nahr-e Shari: 1 Jorem: 2 Behsood: 1 Sukhrood: 1 Bamiyan: 2 The functional culverts in good condition were built properly, using appropriate designs and materials, and most key components were in satisfactory condition, as Table 12 shows. Table 12 - Culvert Key Component Condition (%) Culverts Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Foundation 82% 18% Poor Materials Side Railing 71% 29% Poor Construction Width 82% 18% Poor Construction Wing Walls 82% 18% Design Flaw Top Surface 80% 20% Poor Construction and Materials O&M ACTIVITIES Average Yearly Maintenance Costs: 50 USD (High), 0 USD (Low) Very little maintenance was necessary for culverts assessed from the sample group, Table 13 illustrates, most likely due to high quality of construction and materials used. Only in a small minority of cases was necessary maintenance not performed, such as concrete resurfacing in 2 out of 10 culverts and culvert passage cleaning in 1 out of 10 culverts. All three of these incidences were due to poor materials used during construction or poor quality of construction. 28 Table 13 - Maintenance Activities: Culverts Culverts Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Cleaning of the Culvert Passage 1 1 9 Routine Maintenance of Foundation 0 0 10 Repair of Top Surface/Packing of 0 1 8 Temporary Patches Resurfacing of Concrete 0 2 8 Periodic Reconstruction of Culvert Footings 0 1 10 KAREZS OVERVIEW Karez networks use gravity to channel water through a series of pipelines or canals for irrigation and consumption.9 A karez network is made up of a horizontal series of vertically dug wells that are then linked by underground water canals to collect water from watershed surface runoff in mountains. The canals channel the water to the surface, taking advantage of the current provided by the gravity of the downward slope of a mountain or hill. The canals are mostly underground to reduce water evaporation. A karez system has wells, dams and underground canals built to store the water and control the amount of water flow. Vertical wells are dug at various points to tap into the groundwater flowing down sloping land from the source, the mountain runoff. The water is then channelled through underground canals dug from the bottom of one well to the next well and then to the desired destination, typically an irrigation system. Common problems associated with karez irrigation systems are generally related to design of the system and construction of the tunnels. In addition, tunnels and wells must be properly maintained much in the same way as canals and shallow wells. KEY COMPONENT CONDITION As with the two irrigation assets listed above, karezes assessed by the engineering team were in relatively better condition than other asset types. Of the nine karezes assessed across locations shown in Figure 15, five were functional and in good condition, two were functional and in poor condition, and two were not completed due to lack of budget. 9 Karez systems assessed during this study were used for irrigation. 29 Figure 15 - Assessed Karez Locations All of the karez systems assessed were built according to appropriate designs and with good quality materials. The two karezes which were not in good condition needed cleaning. As with canals, FGD and TA participants reported that only those villagers who benefitted from the irrigation provided by the karez could be motivated to maintain the assets. Karez key components were also in satisfactory condition, as Table 14 shows, with the exception of one tunnel in Golgaw in Zindajan which needed cleaning. Table 14 - Karez Key Component Condition (%) Karezs Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Tunnel 75% 25% O&M Not Performed Intake 100% 0% N/A Vertical Shaft 100% 0% N/A Trench 100% 0% N/A Excavation and Maintenance 86% 14% O&M Not Performed Wells MAINTENANCE ACTIVITIES Average Yearly Maintenance Costs: 400 USD (High), 40 USD (Medium) 0 USD (Low) Required maintenance was performed on most karezes assessed during the survey, as Table 15 shows. However, like canals, there were several instances of cleaning not being performed where required, possibly due to the reasons listed above. 30 Table 15 - Maintenance Activities: Karezs Karezs Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Cleaning of Tunnel 1 2 4 Cleaning of Intake 0 0 1 Routine Cleaning of Trench 2 0 5 Cleaning of Maintenance Wells 1 1 5 Relining of wells 0 0 7 Periodic Relining of Tunnel 0 0 7 Extension of Retaining Wall 0 0 2 MICRO-HYDRO PLANTS OVERVIEW MHPs use the natural flow of water to produce hydroelectric power, and are by far the most complex subprojects assessed. First, a weir is used to raise water levels and divert water to a canal through an intake. Large particles such as gravel and stones are filtered through a gravel bay as the water reaches the forebay. The forebay allows for the transition of water from open channel to pressure flow conditions and regulates the flow into the penstock, which is accompanied by a spillbay that helps to divert water in the event of the turbine valve being closed during emergencies. Water flows through the penstock pipe (anchored to prevent pipe movement) to the turbines, which turn a generator in a power house connected to electrical loads. Finally, electrical power is then transmitted from the generator using transmission lines attached to poles. Due to the complexity of the MHP, many problems may arise as a result of poor design of canals, improper regulation of water flow, quality of construction materials used, or lack of maintenance performed on the penstock, forebay or turbines. KEY COMPONENT CONDITION Eleven of the twelve micro-hydro plants assessed across locations shown in Figure 16 were functional but in poor condition and one was non-functional. As they provide electricity, micro-hydro plants are important assets for communities, but they are also technically complex and expensive to maintain. 31 Figure 16 - Assessed MHP Locations Baharak: 5 Jorem: 2 Sukhrood: 2 Bamiyan: 3 Shibar: 1 The construction quality of some MHPs was questionable, and repairs done by the communities often did not meet acceptable safety standards. The photos in Figure 17 below show a pen stock pipe without a concrete block fixing it in place, which will lead to damage from rain and flooding, and poor quality transmission lines that are frayed and in need of replacement. None of the MHPs assessed had fuse boxes or invertors and had no option for regulating flow of electricity (see Table 16). Further complicating these technical issues, many CDC members reported that villagers would often plug in devices not designed to be run on the low and unregulated amount of electrical current provided by most MHPs, risking additional damage to the subproject. Figure 17 - MHP Components in Deh Payan Naw and Chapchi Bala 32 Table 16 - MHP Key Component Condition (%) MHPs Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Canal 83% 17% Poor Construction Weir 75% 25% O&M Not Performed De-silting basin 75% 25% O&M Not Performed Fore bay 92% 8% Poor Construction Spill way 92% 8% Not Completed Pen stock 92% 8% Not Completed Turbine 100% 0% N/A Fuse box 0% 100% Not Completed Power house 75% 25% Poor Construction Transmission lines 83% 17% O&M Not Performed Poles 83% 17% O&M Not Performed O&M ACTIVITIES Average Yearly Maintenance Cost: 480 USD In the majority of instances, maintenance was adequately performed on assessed micro-hydro plants, as is shown in Table 17. Notable exceptions were the maintenance and upkeep of the power house and the replacement of transmission poles. In addition, 2 MHPs required maintenance on turbines and de-silting basins. Table 17 - Maintenance Activities: Micro-Hydro Plants Micro-Hydro Plants Maintenance Performed? Not Maintenance Activity Yes No Necessary to Date Cleaning of Canal / Forebay 11 1 0 Maintenance of Electrical Panels 9 2 1 Maintenance of Invertors 0 0 0 Routine Maintenance of Turbine 6 2 4 Maintenance of Power House 2 4 6 Maintenance of Poles 4 2 6 Cleaning of De-silting Basins 7 2 3 Replacement of Electric Panels 8 0 3 Periodic Replacement of Invertors 0 0 0 Replacement of Poles 3 2 7 In addition, as Figure 18 shows, electrical panels and turbines in some instances were not properly maintained. The photos below show a leak in a turbine pipe that has been poorly sealed with cloth and another with cables not properly fixed and without using PVC tape, increasing the operational risk of the MHP. 33 Figure 18 - MHP Components in Deh Payan Naw and Khairabad PROTECTION WALLS OVERVIEW A protection wall, exemplified in Figure 19, is a masonry or concrete structure, which resists earth on one of its sides, serving the same purpose as a dam preventing flooding, except with earth. Design of a protection wall involves surveying the soil conditions and calculating the amount of stress that the wall will have to endure. Figure 19 - Examples of Protection Walls from NSP Engineering Manual Common causes of protection wall failure include uneven settlement of masonry walls during the rainy season, poor bonding and wrong bedding of stones in masonry walls, roots of big trees pushing the wall from behind, unstable boulders not cleared from the foundation, designed cross-section not fully constructed, poor backfilling resulting in high seepage water pressure, blocked weep holes in cement masonry walls, and lack of toe protection for walls. These failures are generally due to inadequate design and supervision or poor construction. 34 KEY COMPONENT CONDITION Out of the nine protection walls assessed across locations shown in Figure 20, 4 were functional and in good condition, while 5 were functional but in poor condition. Figure 20 - Assessed Protection Wall Locations Khulm : 2 Jorem: 1 Injil: 2 Zindajan: 1 Behsood: 1 Sukhrood: 1 Bamiyan: 3 Poor condition, as recorded in Table 18, was a result of missing components such as side joints, back fill, or adequate foundation, or poor design such as inadequate width or length of the protection wall. As properly constructed and designed protection walls do not require much effort in terms of O&M, maintenance was not an important factor in subproject condition. Table 18 - Protection Wall Key Component Condition (%) Protection Walls Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Pointing 100% 0% N/A Concrete Capping 88% 13% Poor Construction Back Filling 75% 25% Poor Construction Weep Holes 50% 50% Poor Construction Construction Joint 43% 57% Poor Construction O&M ACTIVITIES Average Yearly Maintenance Costs: 50 USD (High), 0 USD (Low) The majority of protection walls assessed, as Table 19 shows, required no maintenance to date. Exceptions include one instance of concrete topping repair and one instance of debris cleaning not being performed, as well as one instance where a wall did not receive the necessary extension. 35 Table 19 - Maintenance Activities: Protection Walls Protection Walls Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Repair of Loose Stones/Bricks/Gravel 0 0 8 Maintenance of Gabions 0 0 1 Repair of Concrete Topping 0 1 7 Routine Replacement of Settled Segments 0 0 8 Cleaning and Removal of Weeds and Other 1 1 6 Debris Reconstruction of Settled Segments of 0 0 8 Protection Wall Periodic Re-pouring of Concrete 0 0 8 Extension of Retaining Wall 0 1 7 SHALLOW WELLS OVERVIEW A well is a vertical cylindrical hole from which drinking water is drawn by a pump or by hand. The depth and diameter of a shallow well are determined by a hydrologist according to a set of geological conditions. Once these conditions have been determined, communities can perform most of the construction work themselves. Shallow wells are hand dug and consist essentially of three parts:  A well head structure with a head wall to prevent rubbish and surface waste from running into the well and a drain apron on which users stand as they lift water, which also provides the base for a pumping mechanism  A thin water-tight shell of reinforced concrete, lining the well shelf from the surface to the water bearing strata  A porous intake section plug telescoped inside the main lining Common problems associated with shallow wells usually involve poor design or construction and include inappropriate site selection or poor materials used to line the well, leading to collapse. Another common problem is breakdown of the hand pump used to draw water. KEY COMPONENT CONDITION Out of ten shallow well subprojects assessed by the engineering team across locations shown in Figure 21, seven were functional but in poor condition, three were not functional, and one was not completed, but reported as completed in the NSP database. Eight wells lacked top rings and four did not have any protection. The three non-functional well subprojects all had broken hand-pumps, while two had water that was not fit for drinking, described as too salty. Rarely was an entire subproject group of shallow wells in good condition in a given village. 36 Figure 21 - Assessed Shallow Well Locations Khulm : 4 Injil: 1 Zindajan: 1 Behsood: 1 Sukhrood: 1 Bamiyan: 3 Table 20 - Shallow Well Key Component Condition (%) Shallow Wells Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Bore Hole 100% 0% N/A Bore Depth 100% 0% N/A Apron 100% 0% N/A Top Ring 20% 80% Poor Construction Bore Lining 82% 18% Poor Materials Protection (Wall/Railing) 33% 67% O&M Not Performed Hand Pump/Buckets 70% 30% O&M Not Performed O&M ACTIVITIES Average Yearly Maintenance Cost: 30 USD The majority of shallow wells assessed did not require maintenance to date, as Table 21 shows. Nevertheless, maintenance was not performed in several instances where it was required, notably for well lining and well protection repairs and for hand-pump replacement. In these cases, the increased cost of replacing a pump or the technical complexity of repairing the well lining may have inhibited the necessary maintenance. 37 Table 21 - Maintenance Activities: Shallow Wells Shallow Wells Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Repair of Well Linings 0 2 9 Repair of Well Protections 0 3 8 Maintaining Proper Drainage 0 0 11 Routine Replacement of Pulley Bearings 0 0 8 Maintaining of Animal Troughs 0 1 4 Cleaning of Tanks and Platforms 1 1 6 Replacement of Rope and Pulley 1 0 7 Replacement of Bucket or Hand Pump 0 3 6 Periodic Reconstruction of Platform and Animal 0 0 7 Trough SOLAR PANELS OVERVIEW Solar panels are used as part of a photovoltaic system to generate and supply electricity. Setting up a solar panel system is not complicated and mostly requires that individual groups of panels are properly arranged. The solar panels transfer collected energy to a series of batteries, where it is stored for community use. As solar panels are relatively simple to maintain, the main O&M activity involves replacing the battery. Generally the solar panel system should be able to collect energy for 10 to 15 years, depending on the quality of the system. In addition to providing electricity for home use, solar panels are used by communities to power water supply networks. KEY COMPONENT CONDITION Eleven out of twelve solar panels assessed across locations shown in Figure 22 were functional but in poor condition, while one was not functional (see Table 22). No solar panels had invertors to regulate flow of electricity. As with MHPs, lack of regulation coupled with improper use of electrical devices leads to increased deterioration of the solar panel system and the battery. In addition, all assessed solar panels lacked grounding increasing the safety risk of operation. 38 Figure 22 - Assessed Solar Panel Locations Khulm : 2 Behsood: 2 Sukhrood: 2 Bamiyan: 1 Shibar: 5 Table 22 - Solar Panel Key Component Condition (%) Solar Panels Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Solar System Panel 83% 17% O&M Not Performed Battery 33% 67% O&M Not Performed Invertor 0% 100% Not Completed Joint Box 8% 92% Not Completed; Poor Materials Grounding 0% 100% Not Completed Bulbs 17% 83% Poor Materials O&M ACTIVITIES Average Yearly Maintenance Cost: 193 USD In the majority of cases, as Table 23 shows, all required maintenance activities were performed on solar panels, with the notable exception of battery maintenance. As batteries are generally the most expensive component of a solar panel to replace, proper maintenance is essential to ensuring low attrition rates and reduced O&M costs. 39 Table 23 - Maintenance Activities: Solar Panels Solar Panels Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Cleaning of Solar Panels 8 2 2 Routine Maintenance of Batteries 4 7 1 Cleaning of Electrical Connections 0 0 12 Replacement of Batteries 12 0 0 Periodic Replacement of Damaged Solar Panels 0 1 11 TERTIARY ROADS OVERVIEW In a national road network, roads are often categorised as Primary, Secondary, Tertiary, and Basic Access. In Afghanistan, the following distinctions generally apply: Primary Roads: primary roads are often paved with asphalt and are designed to support heavy amounts of traffic. Regional highways that link Afghanistan to its neighbours, the ring road that circles the country, and provincial highways are all examples of primary roads. Secondary Roads: link district centres to provincial centres. Secondary roads may be asphalted or gravelled, but may also be cleared and lined with stones. Tertiary Roads: rural roads that link villages with district centres and in some cases other villages. Tertiary roads may be gravelled, but are sometimes flattened and compacted or lined with stones to provide access to villages. Under NSP, gravel surfacing is only used when traffic is expected to be about 100 vehicles a day. Basic Access Roads: rural roads that provide minimum access between villages and district centres, basic access roads are not gravelled. Road construction is a technically complex process involving many variables that must be taken into consideration, such as camber, stopping sight distance, overtaking sight distance, design speed, super elevation, horizontal curve, passing places, gradients, surfacing, and other detailed surveys. Supporting structures for roads include side ditches, drainage structures, culverts, causeways, and bridges. Draining structures are particularly important in areas with frequent flooding in order to prevent attrition. The Ministry of Public Works (MPW) is responsible for maintenance of all national highways, regional highways, and provincial roads, while MRRD is responsible for rural tertiary and basic access roads. Under NSP, communities may apply for tertiary and basic access roads to connect their villages with the district and other villages. 40 KEY COMPONENT CONDITION The engineering team assessed fourteen tertiary roads across locations shown in Figure 23, of which thirteen were in poor condition and one was not functional. Roads in poor condition were missing key component structures such as side drains, culverts, shoulders, or side protection, or were built using poor materials such as over-sized gravel or poor grade cement in the case of structures such as culverts (see Table 24). Figure 23 - Assessed Tertiary Road Locations Drainage was a major issue for the majority of tertiary roads assessed: only two of fourteen roads had proper drainage while only four had a properly angled slope, allowing water and other debris to collect on the roads leading to quick deterioration of road condition. CDC members reported that road maintenance was too expensive and time consuming for villagers to carry out without additional financial and technical assistance. Table 24 - Tertiary Road Key Component Condition (%) Tertiary Roads Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Base course 33% 67% O&M Not Performed Sub Base 57% 43% Poor Materials Road Surface 36% 64% O&M Not Performed Retaining Wall 33% 67% Poor Construction Side Protection 44% 56% Poor Construction Culverts 33% 67% Not Completed Side Drains 14% 86% O&M Not Performed Road Slope 36% 64% O&M Not Performed Road Drainage 14% 86% O&M Not Performed 41 O&M ACTIVITIES Average Yearly Maintenance Costs: 2,667 USD (High), 0 USD (Low) No maintenance, routine or periodic, was performed on tertiary roads in the sample subproject group (see Table 25). Some of the necessary maintenance includes very high-cost activities such as re- gravelling and reconstruction of settled segments, in which case lack of maintenance is understandable given the poor economic situation of many communities throughout Afghanistan. On the other hand, many low-cost, labour-intensive activities such side drain cleaning and repacking road- protection/shoulders with earth were also not completed. Table 25 - Maintenance Activities: Tertiary Roads Tertiary Roads Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Packing of stones/sand and repairing of 0 3 10 small patches on road surface Maintaining road protection/shoulders with 0 2 9 earth Routine Cleaning of side drains 0 4 1 Maintenance of side drain 0 4 1 Maintenance of protection/retaining wall 0 1 8 Reconstruction of settled segments of roads 0 6 8 Periodic Re-gravelling of gravel packed roads 0 8 6 Reconstruction of retaining wall/culverts, 0 0 8 etc. WATER SUPPLY NETWORKS OVERVIEW A water supply network is a system of engineered hydrologic and hydraulic components which provide water supply through a series of pipes. Two types of water supply networks with different source water collection methods were built under NSP: gravity and motorised. Gravity WSNs draw water from springs, karezes or surface water. The water flows by gravitational force from the source through the transmission pipeline to the water reservoir, and then the stored water is distributed through a network of pipes to tap stands for public use. Motorised WSNs draw water from the surface or underground using pumps, and then distributed in much the same way as a gravity-based WSN. For gravity WSNs, the construction process begins by locating and testing the source of drinking water, and then protecting it from outside contamination, usually by building a protection wall. The transmission pipeline is then laid out and connected to a water reservoir, which must meet specific capacity requirements, determined using the NSP Engineering Manual. Valve boxes are installed to 42 regulate the flow of water from the reservoir through the pipe scheme connected to houses, where stand taps are installed. The main difference between gravity WSNs and motorised WSNs is the way water is drawn from a source: motorised WSNs need powered pumps to pull water up into a reservoir. The advantage of motorised pumps is that water supply is not dependent on seasonal variables or natural disasters such as drought or flooding. As WSNs are some of the most complex subprojects in terms of design and parts, common problems are usually associated with poor design, construction and materials, or insufficient O&M. KEY COMPONENT STATUS Of the eleven water supply networks assessed, five were functional and in good condition, one was functional in poor condition, and five were not functional. Most WSN key components were in good condition, although in four subprojects the pump houses, electrical panels, motors and pumps were not satisfactory generally due to poor maintenance or poor construction. In addition, the pipes used in several WSNs were not polyethylene as indicated in the design of several subprojects. Figure 24 - Assessed WSN Locations Nahr-e Shari : 5 Baharak: 1 Injil: 1 Behsood: 1 Bamiyan: 1 Shibar: 3 Lack of functionality was generally due to inability or unwillingness to pay for the electricity needed to run WSNs: of the five inactive projects, four were non-functional due to lack of budget for electricity, while one had a broken water pump. Reasons for lack of budget varied between the four communities. CDC members from two villages in Nahr-e Shahi (Balkh) reported not being able to afford city electricity10, while one village had a commander who stole the funds available for electricity (Qala-e Qul Mohamad). The fourth community in Behsood did not have access to electricity and could not afford to pay for generator fuel. 10 NCs reported that the three villages in Nahr-e Shari were relatively wealthy and could most likely afford city electricity, but did not pay because they believed that it was the government’s responsibility to provide free power. 43 Table 26 - Water Supply Network Key Component Condition (%) Water Supply Networks Condition Most Frequent Cause of Key component Satisfactory Unsatisfactory Unsatisfactory Condition Bore Hole 100% 0% N/A Pump House 70% 30% O&M Not Performed Electrical Panel 60% 40% O&M Not Performed Motor 60% 40% O&M Not Performed Pump 60% 40% O&M Not Performed Valves 90% 10% O&M Not Performed Over Head Tank 89% 11% Not Completed Ground Water Tank 100% 0% N/A Generator 67% 33% O&M Not Performed Delivery Pipe 80% 20% Poor Construction Distribution Network (Pipes) 82% 18% Poor Construction House Connections 91% 9% O&M Not Performed O&M ACTIVITIES Average Yearly Maintenance Cost: 722 USD Water supply networks were generally maintained in better condition than roads (see Table 27). Routine maintenance activities were conducted on pipes (3 of 6 WSNs where maintenance was required), motors or valves (2 out of 5 WSNs) and the generator or solar panel (1 out of 2 WSNs). In addition, periodic maintenance was conducted in the replacements of pumps, motors and valves (2 out 4, 3 out of 5, and 4 out of 5 WSNs respectively) as well as the replacement of batteries (2 out of 2 WSNs). Table 27 - Maintenance Activities: Water Supply Networks Water Supply Networks Maintenance Performed? Not Necessary Maintenance Activity Yes No to Date Maintenance of Pipes 3 3 4 Maintenance of Motors/Valves 2 3 5 Repair and Maintenance of Leaks and 0 1 10 Routine Broken Pipes Maintenance of Water Storage Tanks 1 1 9 Maintenance of Stand Posts 1 0 8 Maintenance of Generators/Solar Panels 1 1 7 Replacement of Pump 2 2 6 Replacement of Motor 3 2 5 Replacement of Valves 4 1 6 Periodic Replacement of Undersized Pipes 0 1 9 Construction of New Water Tanks 0 0 9 Replacement of Batteries 2 0 5 44 VARIABLES INFLUENCING SUBPROJECT CONDITION AND O&M One of the objectives of the subproject sustainability study is to examine the effects of a number of variables on subproject condition and maintenance activities. According to the technical assessments discussed above, the main cause of non-functionality in subprojects was lack of adequate O&M, followed by major breakdown (usually caused by floods or other natural disasters) and construction faults (see Figure 25). Effective operations and maintenance is thus essential to subproject condition. 11 Figure 25 - Cause of Non-Functionality in Subprojects For the following analysis two independent sets of variables, institutional and structural, are tested against to determine what effect, if any, they have on subproject status and maintenance activities. Institutional variables focus on the organisational structure and the technical capacity of CDCs, as well as the level of community investment, such as the creation of O&M Committees, CDC involvement in project construction and design, capacity for organising O&M activities, technical experience of CDC members, whether the project responds to an actual community need, and community contributions to the project. Structural variables focus on village and subproject level dynamics, such as security, poverty and access, as well as initial subproject construction and O&M cost, technical complexity of the subproject and quality of design and construction. Figure 26 provides an illustration of the structural and institutional variables at work in subproject status and maintenance. 11 Lack of O&M includes lack of maintenance activities performed and lack of funds for operational costs. 45 12 Figure 26 - Institutional and Structural Variables Institutional Structural Variables Variables Organisational Technical Community Community- Subproject- Structure Capacity of CDC Investment based based Access, O&M Experience in Community Technical Geography, Committees Construction Need Complexity Topography Transparency and Experience with Security and Construction Fee Collection Accountability of Similar Projects CDC Poverty and O&M Cost Organisation of Quality of Community Number of Fidelity to Maintenance Training Awareness Households Project Design O&M Number of O&M Plan Contributions Subprojects VARIABLES INFLUENCING SUBPROJECT CONDITION INSTITUTIONAL VARIABLES ORGANISATIONAL STRUCTURE CDCs are organised and trained to supervise subproject condition and to organise and conduct maintenance activities when necessary. The following section examines whether certain organisational structures, such as O&M Committees or appointed or elected O&M Officers have any effect on subproject condition. O&M Committees The formation of an O&M Committee has a significant impact on subproject condition and functionality, as Figure 27 shows. 59% of subprojects that were functional and in good condition had an O&M Committee, compared to 35% of subprojects that were functional and in poor condition. Even more telling, at no point did CDCs associated with non-functional subprojects have an O&M Committee. 12 Fee collections, otherwise known as user fees, are collected routinely and are normally related to a pay-for-use scheme; O&M contributions may include any contribution to O&M such as money, labour, or materials. 46 Figure 27 - Project Status: Was an O&M Committee Formed? Whether an O&M Committee is formed before, during or after project construction also has a significant impact on subproject condition. 41% of subprojects in good condition had associated O&M Committees formed before construction during the design phase, compared to 16% of subprojects in poor condition. Committees that are formed during the design phase gain a deeper understanding and more hands-on experience with the subproject, thus improving their chances of properly organising and carrying out maintenance activities. In addition, O&M Committees that are formed before construction have a better chance of guaranteeing that construction quality is higher, as they are at hand to give input and monitor construction activities such as the procurement of good quality materials and adherence to design. These relationships are shown below in Figure 28. Figure 28 - Project Status: When was the O&M Committee Formed by Construction Quality and by Project Status? Organising Community Based Contributions The number of community members (all, most, few, none) who contribute labour, materials or funding to O&M appears to have no impact on subproject condition. However, the body responsible for organising O&M contributions and activities has a noticeable effect on subproject condition. When activities are organised by the O&M Committee or the CDC, subprojects are more likely to be functional and in good condition. Through the O&M Committee, CDC members receive training in O&M and are thus better equipped to mobilise efficient and effective maintenance activities. The CDC members have 47 a stronger knowledge of the subproject and its maintenance needs and have normally received training in motivating and organising community members for maintenance activities. The involvement of the CDC in O&M activities, as well as its involvement in subproject construction and its expertise (variables that will be further developed below), appears to be an important contributing factor to subproject condition. As illustrated in Figures 29 through 31 below, the effect of maintenance contribution by the CDC or O&M committee is nearly identical regardless of whether it is in the form of labour, material or monetary contributions. Figure 29 - Project Status: Who Organised Labour Contributions? Figure 30 - Who Organised Material Contributions? Figure 31 - Who Organised Monetary Contributions? Involvement of CDC Members All subprojects in good condition had CDC members who provided input during the design/inception stage, compared to 90% of subprojects in poor condition and 87% in non-functional projects, as shown in figure 32 below. Being involved in the project development process earlier has a small positive effect on subproject condition. CDC members have the chance to learn more throughout the process and to better familiarise themselves with the subproject and its maintenance needs. Early involvement also means that the CDC members can supervise and monitor subproject construction, helping to ensure that good quality materials are used and that construction is done properly. In addition, as indicated in Figure 33 below, continued involvement through construction monitoring and auditing of the use of 48 funds also showed similar trends, with functional and in good condition projects being more likely to be monitored and audited. Figure 32 - Project Status: Did CDC Members Provide Input During Figure 33 - Project Status: Were CDC Members Involved in Design Phase? Construction Monitoring and Audit of Funds? TECHNICAL CAPACITY OF CDC MEMBERS Training CDCs who received some form of O&M training were 21% more likely to maintain functional projects in satisfactory condition (Figure 34). Teaching CDC members how and when to maintain a project, as well as how to mobilise maintenance efforts, has a significant effect on subproject condition. In addition, CDC members from villages with functional subprojects in good condition ranked their satisfaction with the training received 2 points higher (on a 10 point scale) than CDCs with subprojects in poor condition (see Figure 35). Thus, quality and relevance of training is equally important for subproject condition. Figure 34 - Project Status: Did the CDC Receive O&M Training? Figure 35 - On a Scale of 1-10 (1 very dissatisfied 10 very satisfied) How Satisfied are CDC Members with O&M Training? 49 Finally, developing an O&M plan, as part of the training received by the FP, is also an important step to ensure good subproject condition. 40% of CDCs with subprojects in good condition reported that all members were familiar with the O&M plan, compared to only 21% of CDCs with subprojects in poor condition and 18% of CDCs with non-functional subprojects. In addition, none of the CDCs with good condition projects reported that no one was familiar with the O&M plan, while 20% of respondents from CDCs with poor condition projects reported this absence. This familiarity highlights the necessity for CDC engagement with and awareness of the O&M plan. Figure 36 - Project Status: Are CDC Members Familiar with the O&M Plan? Prior Experience with Construction or with Similar Projects Prior experience in construction and development has a strong positive impact on subproject condition (as shown in Figures 37 and 38). CDCs with members experienced in construction or development projects were 20% more likely to own a subproject in good condition and 23% more likely to own functional subprojects. Members with construction experience have a greater understanding of maintenance needs and can thus help to ensure subprojects are properly maintained. Furthermore, experience in construction helps CDC members to better monitor subproject construction and ensure initial quality, both in terms of materials used and the way the subproject is built. 50 Figure 37 - Project Status: Do CDC Members have Experience in Figure 38 - Quality of Construction: Do CDC Members have Experience Construction/Development Projects? in Construction/Development Projects? Similarly, CDC members who are already familiar with their subproject have a better understanding of how it works and how it should be maintained, and can help other CDC and community members with maintenance activities. 83% of subprojects in good condition had CDCs with members who had at least some experience maintaining a similar project, compared to 51% of subprojects in poor condition and 38% of non-functional subprojects (see Figure 39). Figure 39 - Project Status: Do CDC Members have Experience Maintaining Similar Projects? Perceived General Technical Capacity of CDC Members 86% of villagers with functional projects in good condition reported that they were confident that members of their community had the capacity and technical knowledge to maintain their subproject, compared to only 46% of villages where the subproject was in poor condition and 38% of villages where the subproject was not functional (Figure 40). Understandably, communities with subprojects that are in good condition have more faith in the technical capacity of CDC members. Nevertheless, CDC members with greater technical capacity who are involved in the construction of the subproject can better supervise activities, helping to improve subproject condition. 51 Figure 40 - Project Status: Do you Feel Community Members have the Capacity and Technical Knowledge to Maintain the Subproject? COMMUNITY INVESTMENT Community Need Despite NSP being intended as a bottom-up process, the percentages in Figure 41 indicate that delivering projects that respond to demonstrated need has had mixes success. Several factors help to explain the poor percentage of overall projects that actually respond to a community need. Most importantly, a project was not assessed as responding to a community need unless the entire community’s needs were met. For example, a water supply network that only provided half the community with drinking water did not address the community’s demonstrated need, because it would still exclude some of the community members with a demonstrated need. Nevertheless, projects that respond to a community need are more likely to be maintained and kept in functional condition. 58% of subprojects in good condition responded to the need of the entire community, compared to 54% of functional projects in poor condition and 29% of non-functional13 subprojects. In addition, as indicated in Figure 42 below, the likelihood that the subproject responded to an actual need was highly dependent on the type of subproject. Micro-hydro power plants were viewed very favourably by respondents, with over 70% responding to a community need. Alternatively, shallow wells were only viewed as addressing the community’s needs about 30% of the time. 13 Community members were asked to assess need at the time of subproject construction. 52 Figure 41 - Project Status: Does the Project Respond to an Actual Figure 42 - Percentage of Subprojects that Respond to an Actual Community Need? Community Need by Subproject Type STRUCTURAL VARIABLES AND SUBPROJECT CONDITION SUBPROJECTS Initial Project Cost Initial subproject cost appears to have a significant effect on project condition: the more money invested in the subproject the more likely it is to be in poor condition or not functional (see Figure 43). As more expensive subprojects generally require more community financial investment for initial construction, this finding seems counter-intuitive. However, more expensive subprojects such as water supply networks, MHPs and tertiary roads are also more costly to maintain. Other possible explanations lie in the subproject sampling as well as in the accuracy of the NSP database. For example, the majority of tertiary roads were in poor condition. As roads are generally the most costly subprojects to build, this may skew the test against subproject condition. Furthermore, construction cost figures in the NSP database are most likely inaccurate. For example, the cost to gravel 100m of road according to the NSP database (30-150 USD per 100m) is well below prices quoted from local construction companies (800- 1000 USD per 100m). Figure 43 - Subproject Status by Average Cost of Construction (AFN) 53 Cost of O&M As a corollary to the cost of construction, cost of O&M had a significant impact on subproject condition as well. Subprojects that are less expensive14 to maintain are more likely to be functional and in good condition (as shown below in Figure 44). As most communities are not capable of collecting the required funds to maintain high-cost subprojects, inexpensive subprojects whose maintenance requirements involve mostly labour contributions (canals, culverts, karezes, protection walls, and shallow wells) are more likely to be kept in better condition. Cost ratings (very high, high, average, low) are based on attrition rates identified by engineers and cost of replacement parts reported at provincial centres. Figure 44 - Subproject Status by Cost of O&M Type and Technical Complexity of Subproject The type and technical complexity15 of the subproject has a significant effect on the condition and functionality (see Table 28). Subprojects that were classified as technically complex (MHPs, WSNs and solar panels) were generally in poorer condition than less complex assets. Notably, 82% of subprojects in good condition are not technically complex, while 18% of the projects in good condition were highly complex, compared to only 33% of subprojects in poor condition, as illustrated in Figure 45 below. 14 Cost rating classification: tertiary roads (very high); micro-hydro plants, water supply networks (high); solar panels (average); canals, culverts, karezes, protection walls, shallow wells (low). 15 Technical complexity classification: MHPs and WSNs (very high); solar panels (high); tertiary roads (average); canals, culverts, karezes, protection walls, shallow wells (low). 54 Table 28 - Project Status by Subproject Type % Functional and in Good % Functional and in Poor Type % Not Functional Condition Condition Canal 71% 14% 14% Culvert 36% 36% 27% Karez 72% 28% 0% Micro Hydro Power Plant 0% 92% 8% Protection Wall 38% 62% 0% Shallow Well 9% 64% 27% Solar Panel 0% 100% 0% Tertiary Road 0% 93% 7% Water Supply Network 36% 18% 45% AVERAGE 29% 56% 14% Figure 45 - Subproject Status by Technical Complexity Quality of Initial Construction The quality of materials and the quality of construction both have a significant impact on subproject condition. All subprojects assessed to be functional and in good condition were built properly using high-grade materials (properly mixed concrete, necessary key components and accompanying structures), compared to 54% in poor condition that were not built properly. Examples of poorly constructed subprojects include:  tertiary roads that do not have the necessary accompanying key components such as proper drainage, leading to quick deterioration of the road surface;  other tertiary roads that were not built at the proper slope due to lack of budget and were thus inoperable during the winter;  a canal with no water;  a protection wall destroyed because it was not built high enough to contain a flood;  a culvert that was not built according to design and was impossible to maintain/clean; 55  another culvert constructed using poor grade cement which was quickly deteriorating;  solar panels and MHPs that lacked invertors or fuse boxes to regulate electrical flow. Indeed, in many scenarios, low-quality materials and sub-par construction efforts seem to be the main causes of poor subproject condition (see Figure 46). Figure 46 - Project Status: Was the Subproject Constructed Properly and with Good Materials? Project Design Although fidelity to the initial subproject design does not seem to greatly affect condition, it does have a noticeable impact on functionality. 80% of functional subprojects were built according to the design agreed upon by community members and FP engineers, compared to 67% of non-functional subprojects (as shown in Figure 47). Design modifications include reduced length or width of culverts, protection walls and tertiary roads and other design flaws causing some of these subprojects to deteriorate quickly, in the case of protection walls built with low-grade concrete (a wall in Noor Khoda Dohum in Nahr-e Shari), this means to not provide sufficient protection against flooding (a protection wall in Shar-e Naw in Zindajan), or to be difficult to clean or maintain (a culvert in Hatam Biki in Baharak). Figure 47 - Project Status: Does the Project Design Match the Completed Subproject? 56 Age of Subproject The age of the subproject did not have a noticeable effect on project condition. Functional projects in good condition had an average age of 5.36 years, while those in poor condition had an average age of 5.12 years and non-functional projects had an average age of 4.92 years. The initial quality of construction and materials used, as well as CDC capacity and, cost of maintenance, and technical complexity of the project, seem to have a much higher impact on project functionality. Indeed, for many of these projects (culverts, protection walls, and shallow wells) age should not be a factor if they are built properly with quality materials. COMMUNITIES Community-level variables tested against subproject condition include poverty, access (distance from DC in minutes), security, and geography/topography, population, and the number of NSP subprojects previously constructed. Poverty Based on profiles of the village sample group, the level of wealth or poverty does not appear to have any impact on subproject condition in sample communities. On average, Khulm and Zindajan districts had wealthier communities than the other districts, although economic disparity between villages existed within all districts. The economic situation on communities was rated on a scale of 1 to 5, with 5 being very poor and 1 being very wealthy. The chart below, Figure 48, illustrates the indeterminate effect of village economics on subproject condition. Figure 48 - Subproject Status: Community Economic Rating The level of economic development is strongly associated with the perceived performance of CDCs with regards to specific roles. In order to assess the impact of wealth on governance patterns, the villages have been categorised into wealth groups. The wealth of a community has been estimated based on various indicators such as the access to infrastructure and facilities, the sources of income available in the village, and the type of goods that people had access to. The wealthiest communities were usually closer to the district centres and benefited from advanced levels of urbanisation, while the poorest communities were facing major difficulties to ensure their livelihood from the little land or livestock they 57 possessed. In between, intermediary communities would typically have access to significant cropping areas and could sell agricultural produce on the local market. In wealthier communities, respondents expressed high levels of satisfaction with the CDC for its ability to manage and supervise projects (more than 70% considering the performance of the CDC in this regard to be very positive across respondents belonging to the first wealth group) but tend to be less appreciative of CDC’s capacity to resolve disputes or maintain a link with the government (respectively 55% and 48% of very positive appreciation within the same group). Yet, this perception changes when the levels of economic development start to decrease. Respondents from the lowest wealth group valuing the performance of the CDC in dispute resolution and in maintaining a link with the government (respectively 68% and 65% of very positive opinion) more than its ability to effectively manage and supervise projects (43% of very positive opinions).16 However, these trends notwithstanding, neither correlation nor causation of the effects of community economics on subproject condition could be established based on the sample group. Population As subproject grants are awarded based on community size, villages with a large number of households receive more money for subprojects than smaller villages. Nevertheless, number of households has no visible effect on subproject condition, most likely due to the increased size and complexity of subprojects that are built for larger communities (see Figure 49). Although a community of 500 households may receive more money for a water supply network than a community with only 50 households, the network in the larger community will necessarily be larger and more complex. Construction and O&M activities are thus more complicated, balancing out the added benefit of receiving a larger grant. Figure 49 - Project Status: Number of Households Number of Subprojects Built Similarly, the average number of subprojects built also has little effect on subproject condition (Figure 50). Even when organising the successive construction of subprojects that are similar in design and benefit, such as a canal and a karez, rarely do communities request to build the same kind of subproject 16 For more information please see the accompanying report on CDC Sustainability by Altai Consulting. 58 twice in a row. In addition, CDCs may find it more difficult to manage more than one project, counterbalancing any benefit construction and O&M experience that they may have gained from previous subprojects. Figure 50 - Project Status: Average Number of Projects Access Distance from the District Centre (DC) does not appear to have any effect on subproject condition, although incomplete projects are on average around 20 minutes further from the DC than other villages. Year-round access roads also do not appear to have any impact on subproject condition. Although these findings appear to demonstrate that ease of access is not important for maintenance activities, subprojects that require heavy or large materials/parts for maintenance are generally in poor condition (tertiary roads, for example). As no attempt has been made to re-gravel the tertiary roads in the sample villages, it is difficult at this point to estimate how much access will affect O&M cost and consequentially subproject condition. However, while accessibility appears in this sample set to have limited impact on subproject condition, it should be noted that due to time and resource constraints, the study could not cover the most remote communities where the impact may have been exacerbated. Security A security rating from one (very good security) to five (very poor security) was assigned to each village, based on kinetic activity (military operations, etc.) and insurgent as well as criminal presence. As most villages were relatively secure, this index did not provide any insight into how security effects subproject condition. Indeed, although Zindajan was the least secure district visited by the field team, subprojects were better maintained in Zindajan than in any other district except neighbouring Injil. Khoshdrew, a village described as highly insecure by the field team, had a tertiary road that was in use and functional if not perfectly maintained. Although testing levels of insecurity against subproject condition did not prove fruitful in this study, including more kinetic districts, if possible, in future assessments could yield different results. 59 Geography/Topography Geographical and topographical variables did not appear to have any significant impact on subproject condition. As was the case with access, this could be due to tertiary roads not being maintained at all by communities. 60 VARIABLES INFLUENCING O&M Variables tested against maintenance performance were similar to the ones used to test influences on subproject condition. Notably, the following variables were tested:  the formation of an O&M Committee, the capacity of its members and its transparency and legitimacy  the construction and O&M cost (as estimated by the CDC) of a subproject, as well as the amount of money invested by the community in O&M  who organises maintenance activities (CDCs, village elders, or the FP) and who participates in maintenance  how the community views the subproject, whether it is based on a community need, how the community views the CDC and how invested the community is in the subproject Figure 51 - Main Causes of Poor Maintenance Maintenance activities are classified in two ways: adequate and inadequate. Thus, subprojects that have only had some maintenance performed are considered to be inadequately maintained (marked as “Maintenance Not Performed” in the charts below). INSTITUTIONAL VARIABLES AND O&M Institutional variables are similar to those used when testing project condition, notably the formation of an O&M Committee and its effect on community contributions, CDC experience and capacity and community O&M investment ORGANISATIONAL STRUCTURE O&M Committees The formation of an O&M Committee has a strong positive effect on maintenance activities. Subprojects were more than twice as likely to be properly maintained in communities with O&M Committees than in communities lacking an institution responsible for O&M (see Figure 52). However, whether O&M 61 Committees are formed before, during or after construction has no demonstrable effect on subproject maintenance (Figure 53). Figure 52 - O&M: Has an O&M Committee Been Formed? Figure 53 - O&M: When was the O&M Committee Formed? The positive impact of the O&M Committee is likely the result of its specialisation and consequent expertise in maintenance activities, expertise that is slightly improved due to increased involvement in the design and construction process. The inclusion of an even more specialised O&M officer within the O&M Committee also significantly increased the likelihood of effective maintenance being performed (see Figure 54). Both the O&M Committee and presence of an O&M officer indicate that the CDC has prioritised O&M, resulting in a positive feedback loop. Figure 54 - O&M: Has an O&M Officer Been Appointed? In addition, the presence of an O&M Committee greatly increases the average community investment in O&M. Communities with O&M Committees invest on average 23,008 AFN per year in O&M, compared to 9,969 AFN per year in communities without O&M Committees, confirming that having CDC members who specialise in O&M increases the chance of a subproject being in good condition and properly maintained. 62 User Fees 42% of effectively maintained subprojects were managed by communities who had instituted a system to collect user fees, compared to 80% of poorly maintained subprojects managed by communities who did not collect user fees (Figure 55). Having money set aside for maintenance thus appears to be an effective way of ensuring that subprojects are properly maintained. While the majority of user fees were collected for MHPs and water supply networks, three communities also reported collecting money for the repair of a road, a karez and a shallow well. Of the three, the karez and the shallow well were properly maintained. Collecting user fees or regular O&M fees for the maintenance of other subprojects could improve subproject condition, although several challenges such as the perceived transparency and accountability of the CDC as well as its capacity to manage funds must also be considered. Furthermore, convincing community members to put money aside regularly for O&M would require communicating to them the importance of the subproject and increasing their level of investment, which could prove to be a difficult task especially if they are not direct beneficiaries of the subproject. Figure 55 - O&M: Is there a Regular System of Collecting User Fees? Organising Community Based Contributions The CDC and O&M Committees appear to be slightly more effective17 at organising and carrying out adequate maintenance. In all three types of contributions, when maintenance is organised by the CDC or the O&M Committee more subprojects are properly maintained. CDC members receive O&M training and generally have a better understanding of the subproject than other non-CDC village officials, which increases their capacity to organise and carry out O&M. Furthermore, CDC members understand the importance of the subproject for the community, and are thus more likely to be able to convince other community members of the importance of maintenance. More specifically, once contributions have 17 For the following 3 charts, Figure 56 to 58, CDCs and O&M Committees ranked higher than village officials. However, due to the smaller sample size of villages where maintenance was performed, the relatively small percentage difference between CDCs and other village officials makes it difficult to come to any definitive conclusion. Nevertheless, the trend points to CDC and O&M Committees being more effective at organising maintenance activities. 63 been organised, CDC can use their greater understanding of maintenance needs in order to most effectively make use of available labour, materials or funds. Figure 56 - O&M: Who Organises Monetary Contributions? Figure 57 - O&M: Who Organises Time/Labour Contributions? Figure 58: O&M: Who Organises Material/Resource Contributions? TECHNICAL CAPACITY OF CDC MEMBERS Capacity of CDC Members Whether CDC members received some form of O&M training and the amount of CDC members familiar with the O&M plan both have a small positive effect on effective subproject maintenance, emphasising the importance of training and CDC involvement and planning when organising O&M activities (see Figure 59 and 60). However, variables such as previous experience with construction and repairs and previous experience maintaining similar projects have no observable effect. 64 Figure 59 - O&M: Did CDC Members Receives some form of Training Figure 60: O&M: Are CDC Members Familiar with the O&M Plan? from NSP/FP Staff? COMMUNITY INVESTMENT Transparency and Accountability of CDCs When asked how their subproject could be improved, 15 villagers (around 4% of the entire sample) reported that the CDC should be more accountable and transparent. Examples include:  A village in Zindajan, where community members explained that they did not feel that the CDC would spend money collected for O&M on the subproject and were thus hesitant to contribute to maintenance  A village in Nahr-e Shari, where villagers explained that they could raise enough money to pay for city electricity to power their water supply network, but that the CDC head had kept the money for himself  A village in Baharak, where villagers expressed disappointment in the way the CDC was managing subproject maintenance and that the CDC was not listening to the people  A village in Jorem, where villagers reported that CDC members should be more honest and more responsive to the people Although the majority of non-CDC community members surveyed during the subproject sustainability study did not report that the CDC was corrupt or unaccountable to other villagers, transparency and accountability are key issues when motivating community members to commit to O&M activities. Communities who perceive their CDCs as corrupt or non-representative will hesitate to contribute funds, labour or materials to O&M. This perspective was confirmed by an AKDN social organiser in Bamiyan, who explained that “[CDC shuras] must have a very high level of transparency and should be able to prove to donors and villagers that they have operated in an ethical fashion… Sustainability is largely dependent on whether villagers believe that the CDC is behaving appropriately.” On the other hand, several CDC members responsible for subproject maintenance reported that as they had no way of proving to community members that funds contributed were actually spent on maintenance, they preferred to not perform any maintenance at all. They explained that if they had 65 spent the money, villagers would assume that they were corrupt and had stolen at least a portion of the funds. Community O&M Investment Communities with properly maintained subprojects invest almost four times more money into O&M than communities with unmaintained subprojects (Figure 61). A strong link exists between money invested and effective maintenance. Communities that invest more money can afford to pay labourers and also to pay for replacement parts. Figure 61 - O&M: Community O&M Monetary Investment (AFN) Awareness/Investment in Project During KIIs with FPs, social organisers noted that motivating community members to conduct maintenance was major challenge for O&M. They outlined two main causes for this lack of motivation and investment: that the subproject did not benefit the entire community and thus excluded members had no incentive to help with maintenance activities or to contribute user fees; and that often community members were not aware of the importance of the subproject and so did not conduct timely and effective maintenance. In the first scenario, two types of exclusion were apparent: exclusion due to lack of subproject capacity (for example when a water supply network did not reach all the households in the community) and exclusion due to poverty (for example when a household could not afford to pay user fees for a project). Both types of exclusion triggered corresponding problems in terms of adequate O&M being carried out. Villagers who were excluded due to lack of capacity were significantly less motivated to participate in O&M activities, often requiring pay to abandon their work for the day to help with maintenance, driving up the overall cost of O&M. In the case of canals, for example, CDC members estimated that when villagers could not be motivated out of self-interest to maintain the subproject, costs for labour could easily reach 20,000 AFN/year on average (400 USD) depending on the size of the canal and the intensity of the required labour. In villages where poor community members could not contribute, wealthier community members would often cover the cost of maintenance for them. In some instances, these unequal contributions led to 66 conflict between community members who paid user fees and those who could not afford to, hindering maintenance activities and community harmony. Donor Dependency Field teams as well as several FP social organisers reported that when asked why they had not performed maintenance on their subproject, CDC members from several communities explained that they would prefer to either wait for a donor to come and repair the project or to come and build a new project altogether. In the case of a new subproject being built, the CDC members would then benefit from the added prestige of bringing a new construction project to the community. Although this was not a widely reported phenomenon, instances of donor dependency could be combatted by increased efforts to convince community members of the subproject’s importance and by encouraging ownership. Community Need Subprojects that respond to a community need have a greater likelihood of being properly maintained (Figure 62). Effective maintenance was performed on 63% of subprojects that respond to a community, compared to 42% of subprojects that do not. Villagers and CDCs are more likely to be motivated to participate in maintenance activities when the subproject benefits the community. Figure 62 - O&M: Does the Project Respond to an Actual Community Need? User Fees As discussed above, when user fees are collected adequate maintenance activities are more likely to be performed (Figure 63). Having funds collected through regular user fees on hand when a subproject requires maintenance allows CDCs to avoid the often complicated task of organising community members to collect maintenance. In the context of community investment, subprojects for which user fees are necessary generally provide a direct and obvious benefit to community members (electricity or water), and thus encourage and facilitate contributions to maintenance. 67 Figure 63 - O&M: Is there a Regular System of Collecting User Fees? Impediments to Contribution Impediments to contribution were reported by FGD HH respondents, and included lack of funds (to pay for labour, materials or for technical assistance), lack of time to participate in maintenance, lack of materials where funds available, and lack of technical capacity. Other possible impediments identified by respondents included no one organising maintenance and not having any interest in maintenance. For monetary and material maintenance contributions (Figures 64 and 65), lack of funds was the main impediment identified by respondents. Figure 64 - What are the Main Impediments to Contributing Funds? Figure 65 - What are the Main Impediments to Committing Materials/Resources? For labour contributions (Figure 66), lack of time was identified as the main impediment, followed by lack of funds and lack of skills. While the impediment posed by lack of funds to monetary and material contributions is evident – as demonstrated above, many community members cannot afford to pay for maintenance or materials – lack of funds was also identified as a key impediment to labour contributions by 37% of FGD HH respondents. Funding is an issue for labour because when a significant amount of community members do not benefit from a given subproject, they have no incentive to contribute to maintenance activities and the CDC must pay them in exchange for their labour. Indeed, funding is likely as large of an impediment for 68 labour as it is for monetary and material contributions: community members may lack the time to participate in labour-based maintenance because they need to do other work involving monetary gain such as tending to their fields or herds. Thus, to compensate them for time taken away from their regular work, they must be paid, meaning that labour is essentially a funding issue. Figure 66 - What are the Main Impediments to Committing Time/Labour? STRUCTURAL VARIABLES AND O&M COMMUNITIES Population Communities where subprojects were poorly maintained had a slightly higher average population than communities where subprojects were properly maintained (Figure 67). Larger communities require larger subprojects, which may constrain effective maintenance. Organising the necessary labour to clean and maintain a canal or karez, for example, becomes more complicated as the size of the irrigation subproject increases. Furthermore, increased population may also hinder the basic management capacity of the CDC, as a greater potential for disputes over subproject benefits may exist. Figure 67 - O&M: Average Number of Households 69 Number of Subprojects Communities with properly maintained subprojects had a slightly lower average number of projects per community (2.1) than communities with poorly maintained subprojects (2.5), as shown in Figure 68. As was the case with subproject condition, the benefits of experience are most likely counterbalanced by the difficulties of managing multiple subprojects. Indeed, the trend we see here is actually reversed compared to subproject condition, where subprojects in good condition belonged to communities with more subprojects rather than communities with subprojects in poor condition or not-functional. The number of subprojects may have a greater effect on O&M than subproject condition because the more subprojects a community has, the more difficult it is for them to organise and conduct effective maintenance activities. Figure 68 - O&M: Average Number of Projects Poverty The economic situation in the village did not have any visible effect on whether or not adequate maintenance was performed in sample communities. The average economic community rating for both adequately maintained and poorly maintained subprojects was 3 out of 5 (1 being very wealthy and 5 being very poor). Thus, for the sample group, poverty is not a significant variable for O&M. Ethnicity Although ethnicity had no demonstrable effect on subproject condition in the sample group, homogeneous communities perform effective maintenance more often than heterogeneous communities. More homogeneous communities may be capable of better organising effective maintenance activities as the presence of ethnic conflicts or disputes is decreased, the likelihood of responsiveness to community needs as a whole increases and community members may work better with members of their own ethnicity. 70 Figure 69 - O&M: Ethnicity Access Similarly, the distance from the district centre did not have any noticeable effect on O&M activities. The average time to the DC in communities where maintenance was adequately performed was 50 minutes, compared to 57 minutes in communities where maintenance was not adequately performed. Access may become more of a factor if communities begin to maintain tertiary roads on their own, which is unlikely due to the high cost of maintenance. An additional explanation could be that most subproject key components, aside from pipes and large quantities of concrete that are not often required, are easily transportable. Furthermore, due to time and resource constraints, the study could not cover very remote communities. NSP I vs. NSP II Communities Subprojects built under NSP II had a greater chance of being adequately maintained than subprojects built under NSP I (see Figure 70). As age of subproject is not a significant factor for effective maintenance or subproject condition, this distinction is most likely due to structural issues such as improved maintenance procedures and planning under NSP II, as well as more effective sub-project selection. Figure 70 - O&M: NSP Phase 71 Security As with poverty and access, security does not appear to have any significant impact on effective subproject maintenance. The average security rating for communities where maintenance was adequately performed was 1.3, the exact same rating as in communities where maintenance was not performed (1 being very secure and 5 being very insecure). As most communities surveyed were ranked as 1 or very secure, drawing conclusions based on security rankings is difficult. Nevertheless, two communities that were ranked 3 and 4 on the security index, both in Zindajan, had conducted effective maintenance activities on their subprojects. Geography/Topography As both districts in Badakhshan, Baharak and Jorem, had different levels of subproject maintenance than Bamiyan and Shibar districts in Bamiyan even though they are both mountainous areas, topography does not seem to have any visible impact on maintenance activities. The same distinctions can be observed in four districts that are composed mostly of plains: Khulm and Nahr-e Shari in Balkh and Zindajan and Injil in Herat (see Table 29). Table 29 - Maintenance Performed by District Maintenance Maintenance District Performed Not Performed Baharak 22% 78% Jorem 0% 100% Khulm 43% 57% Nahr-e Shari 11% 89% Shibar 38% 63% Bamyan 25% 75% Zindajan 40% 60% Injil 100% 0% Behsood 29% 71% Sukhrood 83% 17% SUBPROJECTS Construction Cost of Subproject The average construction cost for subprojects where maintenance was adequately performed is 24,496 USD, compared to 21,305 USD for subprojects where maintenance was not performed. As communities must pay for a minimum of 10% initial subproject cost, the higher the investment the more likely they are to properly maintain the subproject. Nevertheless, as discussed above, construction cost as reported in the NSP database does not seem to match actual cost, and as such any conclusion drawn from these figures must be made with reservations. 72 Technical Complexity of Subproject More technically complex subprojects from the sample group were better maintained than less complex subprojects (Figure 71). More specifically, 48% of subprojects that were rated very high and 45% of subprojects that were rated high in terms of technical complexity were maintained, compared to 0% and 38% of average and low complexity subprojects. As the more technically complex subprojects from the sample group, MHPs and water supply networks, were also the most directly beneficial villagers, communities have more incentive to keep them maintained. Figure 71 - O&M: Technical Complexity of Subproject O&M Cost18 of Subproject The chart below, Figure 72, confirms the same findings as the chart on technical complexity above: MHPs and water supply networks (high O&M cost subprojects) are maintained because they respond to a direct community need. The average cost subprojects assessed in this study, solar panels, are also properly maintained, with the majority of communities replacing batteries on several occasions. Low cost subprojects were almost equally maintained and not maintained. Roads, the most expensive subprojects in terms of O&M cost, are not maintained because while communities understand the importance of paying for fresh water and electricity, paying the high cost of maintaining a road is not a priority. The lesson to be learned from these analyses is that need is more important than cost or technical complexity, and that some communities do have the capacity and funds available to maintain a subproject as long as that subproject responds to a basic community need. 18 O&M Cost is what CDC members believe they need in terms of O&M budget in order to maintain the subproject. Actual reported O&M expenditures are examined under Community O&M Investment. 73 Figure 72 - O&M: O&M Cost of Subproject Initial Construction Quality of Subproject The initial quality of construction also has an effect on maintenance activities. Good-quality subprojects are 28% more likely to be properly maintained than poor quality subprojects. Community members have more incentive to maintain a subproject when they believe that it is properly built. If the workmanship or materials are poor, maintenance costs decrease and communities are more willing to put money into maintenance: good quality subprojects receive an average yearly O&M contribution of 16554 AFN compared to poor quality subprojects, which receive 10885 AFN. Figure 73 - O&M: What is the Initial Construction Quality? 74 SUBPROJECT O&M COST AND FUNDING GAP INTRODUCTION O&M COSTING MODEL APPROACH The following section presents the findings of the O&M costing model exercise, including the O&M funding gap by subproject and by district, and suggests ways in which this gap could be mitigated. The O&M costing model estimates costs for the 100 communities surveyed during the subproject sustainability study. For the purpose of the model, one question from the focus group discussion questionnaire directed at CDC members was used: “How much does the community contribute to O&M costs per year?” The O&M expenditures reported per community by CDC members were then compared against O&M estimates prepared by Altai. These estimates were prepared with the help of a team of engineers, and are based on costs of key components and attrition rates. Cost estimates are presented as a number of scenarios influenced by the series of modifiable assumptions presented below. As an additional exercise, a funding gap estimate for the entire NSP database19 is calculated. A fixed community investment rate is used based on data from average community contributions per subproject collected from the 100 communities during the sustainability assessment. The total estimated cost per subproject type is then compared to the average community investment to determine any gaps in funding. ASSUMPTIONS In order to build the O&M costing model, a series of static and modifiable assumptions were made with regards to key component replacement cost, attrition rates, and maintenance activities. These assumptions are outlined below. GENERAL It is assumed that effective maintenance will be carried out on subprojects according to a standard schedule and that high quality materials will be used when key components must be replaced. In addition, it is assumed that funds will be saved for subproject key component replacement in the case of an eventual breakdown; in other words, the yearly O&M budget presented below is based on routine maintenance costs as well as the cost of key component replacement, based on key component attrition rates. Furthermore, it should be noted that while certain subprojects are relatively expensive to maintain (water supply networks, MHPs, and tertiary roads), others do not require a significant amount of maintenance and thus do not cost very much in terms of O&M. 19 Number of total subprojects was taken from the “NSP I&II Completed Subprojects” database from August 2011. 75 KEY COMPONENT REPLACEMENT COST Key component replacement costs were sourced from five provincial centres: Faizabad (Badakhshan), Mazar-e Sharif (Balkh), Bamiyan, Herat, and Jalalabad (Nangarhar). A minimum of two suppliers were contacted in each location, and the average cost was used in the O&M model. Where applicable, the replacement cost also includes the cost of specialised labour. ATTRITION RATES Attrition rates were determined with the assistance of the engineers for each subproject key component. The model assumes that high quality key components will be used and that subprojects will be properly constructed according to good project designs. INCOMPLETE SUBPROJECTS Incomplete subprojects are not included in the O&M cost assessment. In addition, while calculating O&M costs for the entire NSP database, the model provides the option of including the same percentage of subprojects types that were assessed as complete from the sample group. For example, 70 percent of canals from the sample group were completed, and thus the model includes the option of only calculating costs for 70 percent of total canals. MAINTENANCE ACTIVITIES Two types of maintenance activities are included in the model: high-frequency and low-frequency. High- frequency maintenance assumes that maintenance will be performed as required on subprojects and their key components, and as such the attrition rates will be lower than low-frequency maintenance. In addition to high and low frequency, the following two subproject specific assumptions are also included. For the scenarios presented in Tables 32 through 37 below, high frequency (adequate) maintenance is assumed to have been performed. Paid Labour One of the findings from the main subproject sustainability report suggested that often CDCs must pay community members to perform maintenance activities on canals and karezes. They explained that this was due to their difficulty in motivating community members who did not directly benefit from the subproject in question to perform maintenance activities. Thus, the costing model includes the options of paid or voluntary labour with regards to canals and karezes. Field teams reported that the estimate for labour ranged from 2,000 to 20,000 AFN per year: the option of these two costs is included in the costing model. In addition, communities also had to pay on occasion for labour contributions to culvert and protection wall maintenance and cleaning. The engineering team estimated, based on conversations with CDC members and from their own experience with O&M, that these activities would cost communities $5 a day per labourer, and require 10 man days per year, resulting in $50 USD in labour costs alone per year. 76 Tertiary Roads Tertiary roads are the most expensive subprojects to maintain, mainly due to the cost of re-gravelling. If a road has been gravelled, this gravel must be replaced on a regular basis, if not the road must be re- gravelled or it will cease to be a gravelled road and thus lose subproject functionality. All tertiary roads assessed during the subproject sustainability study were in poor condition and poorly maintained. Thus, the costing model provides the option to choose between maintenance performed on tertiary roads and maintenance not performed. O&M costs for roads are calculated per kilometre. Water Supply Networks Water supply network costs were calculated by combining unit costs per network (replacement of motor, turbines, and other key components) with the amount of pipe used to build the network. Thus, longer networks have a significantly higher maintenance cost than smaller networks. LIMITATIONS The main limitation for this costing exercise was lack of resources and the limited timeframe provided. While engineers were able to collect data on key component costs during visits to provincial centres, estimates for all the key components of a water supply network or an MHP (as based on NSP manuals) could not be completed. Instead, priority was given to the most expensive and crucial subproject key components. Thus, while the O&M estimates provided below are most likely low, an initial funding gap estimate may be calculated when compared to O&M costs reported at the community level. Finally, the funding gap estimate for the entire NSP database is based on data collected from the 100 communities assessed during the sustainability study. While certain communities did invest in protection walls and culverts, the O&M model is set to assume that these assets will not depreciate as good quality construction materials have been used. As such, positive budget sums for these subprojects are likely misleading. ESTIMATED FUNDING GAP BY SCENARIO The following section presents three possible O&M cost scenarios, illustrating how costs can change depending on the performance of maintenance activities, the motivation of community members, and the actor responsible for performing maintenance. In addition, for the following scenarios, the percentage of completed subprojects from the sample group is applied to the overall NSP database (for example, only 70 per cent of canals are assumed to be completed). The table below illustrates the funding gap per year for the best and worse-case scenarios calculated for all NSP subproject types assessed during the sustainability study. Even when costs are kept to a minimum due to proper maintenance and community investment in O&M activities, communities are still running an O&M cost deficit of 1.67 million USD. Table 30 - O&M Funding Gap by Scenario (all NSP subprojects, USD per year) Scenario Investment Costs Funding Gap 77 Worst Case 4,778,648 57,820,704 (53,042,055) Average Best Case 4,778,648 6,410,415 (1,631,766) AVERAGE COMMUNITY INVESTMENT AND ESTIMATED O&M COSTS Community investments per subproject varied significantly depending on the type of subproject constructed (Table 31). For example, MHPs and water supply networks received the highest average investments. For MHPs, investment per subproject was more or less equal independent of the MHP in question, whereas for water supply networks, the average investment depended to a large extent on the size of the network. For example, Injil and Nahr-e Shari, the two districts with the largest WSNs assessed, contributed the most to O&M for their respective networks. On the other side of the O&M investment spectrum, solar panels, culverts, canals, and protection walls received little to no O&M investment, mostly because they require low-cost maintenance to remain functional. Nevertheless, culverts, canals, and protection walls in Behsood, as well as karezes in Zindajan and protection walls in Khulm and Sukhrood did require some O&M investment, findings which confirm the field team’s assessment that in certain cases CDCs must pay for labour. Finally, although tertiary roads were assessed in all districts except Khulm and Zindajan, only communities in Nahr-e Shari were willing or capable of contributing to O&M costs. The average investment per km of tertiary road was 293 AFN or 5.65 USD, compared to an average estimated re-gravelling and maintenance cost of 3,000 USD per km. Indeed, maintaining tertiary roads is by far the most expensive O&M activity for communities to undertake. Table 31 - Average Community Investment (AFN) per Subproject and District Community O&M Investments (AFN) District Protection Shallow Solar DISTRICT Canal Culvert Karez MHP Tertiary Road WSN Wall Well Panel TOTALS Baharak - - - 21,600 - - - - 5,000 26,600 Jorem - 3,250 - 31,500 - - - - - 34,750 Khulm - - - - 3,000 7,938 1,375 - - 12,313 Nahr-e Shari - - - - - - - 39,188 88,003 127,190 Bamyan - - - 20,000 - 2,083 - - - 22,083 Shibar - - - 50,000 - - 3,750 - - 53,750 Injil - - - - - - - - 45,833 45,833 Zindajan - - 18,214 - - - - - 10,250 28,464 Behsood 10,000 20,000 - - 20,000 - - - - 50,000 Sukhrood - - - 25,000 2,500 21,750 - - - 49,250 20 Average 4,286 2,409 18,214 25,462 3,563 5,975 1,875 13,063 53,888 20 The average investment per subproject reported here is used to calculate the overall O&M funding gap for the entire NSP database. 78 SCENARIO 1: TERTIARY ROADS REQUIRE RE-GRAVELLING AND CANALS, CULVERTS, KAREZES, AND PROTECTIONS WALLS REQUIRE PAID MANUAL LABOUR In the following scenario, tertiary roads are not properly maintained and thus require re-gravelling paid for by communities, and canals, culverts, karezes, and protections walls require paid manual labour, most likely because some community members do not benefit and require payment for labour contributions (Table 32). Needless to say, the following scenario is not sustainable for communities as they will be unable to cover the large subproject maintenance costs. The table below illustrates the funding gap by subproject and by district. In certain scenarios (MHPs in Shibar, one shallow well in Sukhrood), communities could cover adequate O&M costs, but overall they could not raise enough money to pay for any subproject group. Table 32 - O&M Cost: Scenario 1 Funding Gap Subproject Investment minus Estimated Cost (AFN) Protection Shallow Solar DISTRICT Districts Canals Culverts Karezs MHPs Tertiary Roads WSNs Walls Wells Panels TOTALS Baharak - (70,200) - (107,143) - - - (1,145,733) (57,882) (1,380,959) Jorem - (43,550) - (19,997) (23,400) - - (2,551,467) - (2,638,414) Khulm (20,800) (23,400) - - (43,800) (10,783) (18,187) - - (116,970) Nahr-e Shari - (23,400) - - - - - (1,615,279) (1,063,685) (2,702,364) Bamyan - (46,800) - (58,728) - (15,077) (10,925) (312,000) - (443,530) Shibar - (23,400) - 23,757 - - (39,951) (1,500,720) - (1,540,313) Injil (41,600) - (20,800) - (46,800) (4,680) - (138,667) (445,223) (697,770) Zindajan - - (106,586) - (23,400) (4,680) - - (27,292) (161,958) Behsood (52,400) (3,400) - - (3,400) (17,160) (19,037) (55,467) (144,272) (295,136) Sukhrood (20,800) - - (20,621) (20,900) 1,470 (19,037) (83,200) - (163,089) SUBPROJECT (135,600) (234,150) (127,386) (182,733) (161,700) (50,909) (107,138) (7,402,533) (1,738,355) (10,140,503) TOTALS When these costing estimates are applied to the entire NSP subproject database, a similar trend occurs: overall, communities cannot afford to maintain tertiary roads, water supply networks, shallow wells, canals, and solar panels (Table 33). Re-gravelling of tertiary roads is the largest driver of the funding gap, accounting for more than 95% of all costs, followed by replacement of pipes in water supply networks. In this scenario, tertiary roads will deplete until they require full reconstruction, and once the majority of pipes break in water supply networks they will not be replaced, as communities are not putting aside enough money in the case of eventual breakdown. Finally, canals will require complete re-digging as communities will not be able to afford the labour required for maintenance. 79 Table 33 - O&M Costs: Scenario 1 (All NSP Subprojects) Funding Gap (Investment minus Investment Estimated Cost Subproject Estimated Cost) (AFN) (USD) (AFN) (USD) (AFN) (USD) Canals 1,926,000 37,038 9,347,520 179,760 (7,421,520) (142,722) Culverts 2,243,667 43,147 20,403,717 392,379 (18,160,050) (349,232) Karezs 7,522,500 144,663 8,590,400 165,200 (1,067,900) (20,537) MHPs 40,178,308 772,660 39,416,967 758,019 761,340 14,641 Protection Walls 7,050,188 135,581 5,145,400 98,950 1,904,788 36,631 Shallow Wells 20,046,125 385,502 46,755,072 899,136 (26,708,947) (513,634) Solar Panels 1,849,375 35,565 9,887,558 190,145 (8,038,183) (154,580) Tertiary Roads 73,424,313 1,412,006 2,629,848,000 50,574,000 (2,556,423,688) (49,161,994) WSNs 94,249,238 1,812,485 237,281,964 4,563,115 (143,032,726) (2,750,629) TOTAL 248,489,712 4,778,648 3,006,676,597 57,820,704 (2,758,186,886) (53,042,055) SCENARIO 2: TERTIARY ROADS REQUIRE RE-GRAVELLING, BUT CANALS, CULVERTS, KAREZES, AND PROTECTION WALLS DO NOT REQUIRE PAID MANUAL LABOUR In the following scenario, tertiary roads are still not properly maintained but the CDC can organise enough free labour contributions that canals, culverts, karezes, and protection walls do not require paid manual labour (Table 34). Thus, money collected for the maintenance of the above subprojects may be put aside in the case of an eventual breakdown due to a natural disaster such as flooding. While free labour facilitates the affordable maintenance of canals, culverts, karezes, and protection walls, the O&M cost for roads is still much too expensive for communities to cover and thus communities responsible for tertiary roads are unable to perform effective maintenance due to the high cost of re-gravelling. Table 34 - O&M Cost: Scenario 2 Funding Gap Subproject Investment minus Estimated Cost (AFN) Protection Shallow Solar DISTRICT Districts Canals Culverts Karezs MHPs Tertiary Roads WSNs Walls Wells Panels TOTALS Baharak - - - (107,143) - - - (1,145,733) (57,882) (1,310,759) Jorem - 3,250 - (19,997) - - - (2,551,467) - (2,568,214) Khulm - - - - 3,000 (10,783) (18,187) - - (25,970) Nahr-e Shari - - - - - - - (1,615,279) (1,063,685) (2,678,964) Bamyan - - - (58,728) - (15,077) (10,925) (312,000) - (396,730) Shibar - - - 23,757 - - (39,951) (1,500,720) - (1,516,913) Injil - - - - - (4,680) - (138,667) (445,223) (588,570) Zindajan - - 18,214 - - (4,680) - - (27,292) (13,758) Behsood 10,000 20,000 - - 20,000 (17,160) (19,037) (55,467) (144,272) (185,936) Sukhrood - - - (20,621) 2,500 1,470 (19,037) (83,200) - (118,889) SUBPROJECT 10,000 23,250 18,214 (182,733) 25,500 (50,909) (107,138) (7,402,533) (1,738,355) (9,404,703) TOTALS 80 Scenario 2, applied to the entire NSP subproject database, yields results similar to Scenario 1 with one major exception: canals, culverts, karezes, and protection walls are voluntarily maintained by communities, reducing O&M cost. In this scenario, tertiary roads and water supply networks are the biggest driver of overall O&M cost, and communities still cannot afford to maintain and shallow wells and solar panels in the long term (Table 35). Table 35 - O&M Cost: Scenario 2 (All NSP Subprojects) Funding Gap (Investment minus Investment Estimated Cost Subproject Estimated Cost) (AFN) (USD) (AFN) (USD) (AFN) (USD) Canals 1,926,000 37,038 - - 1,926,000 37,038 Culverts 2,243,667 43,147 - - 2,243,667 43,147 Karezs 7,522,500 144,663 - - 7,522,500 144,663 MHPs 40,178,308 772,660 39,416,967 758,019 761,340 14,641 Protection Walls 7,050,188 135,581 - - 7,050,188 135,581 Shallow Wells 20,046,125 385,502 46,755,072 899,136 (26,708,947) (513,634) Solar Panels 1,849,375 35,565 9,887,558 190,145 (8,038,183) (154,580) Tertiary Roads 73,424,313 1,412,006 2,629,848,000 50,574,000 (2,556,423,688) (49,161,994) WSNs 94,249,238 1,812,485 237,281,964 4,563,115 (143,032,726) (2,750,629) TOTAL 248,489,712 4,778,648 2,963,189,561 56,984,415 (2,714,699,849) (52,205,766) SCENARIO 3: TERTIARY ROADS DO NOT REQUIRE RE-GRAVELLING OR ARE MAINTAINED BY LINE DEPARTMENTS, AND COMMUNITY MEMBERS CAN ORGANISE FREE LABOUR-BASED MAINTENANCE CONTRIBUTIONS FOR CANALS, CULVERTS, KAREZES, AND PROTECTION WALLS In the following scenario, tertiary roads are properly maintained by community members or are re- gravelled by line departments, and canals, culverts, karezes, and protection walls require only free labour provided by community members (Table 36). The main driver of O&M cost in this scenario is money put aside for the replacement of solar panel systems, and especially funds saved for the replacement of pipes in water supply networks. 81 Table 36 - O&M Cost: Scenario 3 Funding Gap Subproject Investment minus Estimated Cost (AFN) Protection Shallow Solar DISTRICT Districts Canals Culverts Karezs MHPs Tertiary Roads WSNs Walls Wells Panels TOTALS Baharak - - - (107,143) - - - - (57,882) (165,025) Jorem - 3,250 - (19,997) - - - - - (16,747) Khulm - - - - 3,000 (10,783) (18,187) - - (25,970) Nahr-e Shari - - - - - - - 39,188 (1,063,685) (1,024,497) Bamyan - - - (58,728) - (15,077) (10,925) - - (84,730) Shibar - - - 23,757 - - (39,951) - - (16,193) Injil - - - - - (4,680) - - (445,223) (449,903) Zindajan - - 18,214 - - (4,680) - - (27,292) (13,758) Behsood 10,000 20,000 - - 20,000 (17,160) (19,037) - (144,272) (130,469) Sukhrood - - - (20,621) 2,500 1,470 (19,037) - - (35,689) SUBPROJECT 10,000 23,250 18,214 (182,733) 25,500 (50,909) (107,138) 39,188 (1,738,355) (1,962,983) TOTALS Scenario 3 is the most optimistic of all three scenarios. When applied to the entire NSP subproject database, canals, culverts, karezes, protection walls, and tertiary roads become affordable. However, as with Scenario 2, water supply networks are still the biggest driver of O&M cost, along with shallow wells (Table 37). Table 37 - O&M Costs: Scenario 3 (All NSP Subprojects) Funding Gap (Investment minus Investment Estimated Cost Subproject Estimated Cost) (AFN) (USD) (AFN) (USD) (AFN) (USD) Canals 1,926,000 37,038 - - 1,926,000 37,038 Culverts 2,243,667 43,147 - - 2,243,667 43,147 Karezs 7,522,500 144,663 - - 7,522,500 144,663 MHPs 40,178,308 772,660 39,416,967 758,019 761,340 14,641 Protection Walls 7,050,188 135,581 - - 7,050,188 135,581 Shallow Wells 20,046,125 385,502 46,755,072 899,136 (26,708,947) (513,634) Solar Panels 1,849,375 35,565 9,887,558 190,145 (8,038,183) (154,580) Tertiary Roads 73,424,313 1,412,006 - - 73,424,313 1,412,006 WSNs 94,249,238 1,812,485 237,281,964 4,563,115 (143,032,726) (2,750,629) TOTAL 248,489,712 4,778,648 333,341,561 6,410,415 (84,851,849) (1,631,766) 82 CONCLUSIONS AND RECOMMENDATIONS The objective of the subproject sustainability study was to isolate variables affecting O&M and subproject condition, to determine subproject O&M cost for the assessed communities, and to develop strategy recommendations for reducing cost and improving sustainability of subprojects. This section begins by presenting key findings from tests performed on the variables outlined in Figure 74 and from the three subproject O&M costing scenarios presented in greater detail in the body of this report. The section concludes with a discussion of general conclusions, followed by a series of strategy recommendations for NSP/MRRD. KEY FINDINGS VARIABLES INFLUENCING SUBPROJECT CONDITION AND MAINTENANCE ACTIVITIES The key findings presented below (Table 38) focus on five main categories: institutional bodies and planning, community technical capacity, level of community investment, community-based factors, and subproject-based factors. In the body of this report, variables falling under these five categories were tested to determine any significant impact on subproject condition and maintenance activities. The results below discuss key findings drawn from these tests. Figure 74 - Institutional and Structural Variables Institutional Structural Variables Variables Organisational Technical Community Community- Subproject- Structure Capacity of CDC Investment based based Access, O&M Experience in Community Technical Geography, Committees Construction Need Complexity Topography Transparency and Experience with Security and Construction Fee Collection Accountability of Similar Projects CDC Poverty and O&M Cost Organisation of Quality of Community Number of Fidelity to Maintenance Training Awareness Households Project Design O&M Number of O&M Plan Contributions Subprojects 83 Table 38 - Key Findings: Institutional and Structural Variables Organisational Structure Key Findings The formation of an O&M Committee increases the chance that a subproject will be in good condition. The formation of an O&M Committee and the appointment of an O&M officer O&M Committees increase the chance that a subproject will be properly maintained. The presence of an O&M Committee greatly increases the average community investment in O&M. Communities where user fees were collected had a higher chance of performing Fee Collection effective subproject maintenance. CDCs and O&M Committees are more effective at organising maintenance Organising Maintenance Activities activities than village elders or other authority figures. When all CDC members are familiar with the O&M plan, the chance of the O&M Plan subproject being in good condition increases. Technical Capacity Key Findings O&M Training O&M training has a positive effect on subproject condition. Experience in Construction Experience in construction has a positive effect on initial construction quality. CDC experience with similar subprojects increases the chance that a subproject Experience with Similar Subprojects will be in good condition. Community Investment Key Findings Subprojects that respond to a community need are more likely to be properly Community Need maintained. Transparency and accountability of the CDC are important factors in subproject Transparency and Accountability maintenance. Community members who do not benefit from the subproject due to lack of Community Contributions to O&M subproject capacity require additional financial incentives to participate in maintenance activities, driving up the cost of O&M for labour-based subprojects. Encouraging community awareness of the importance of the subproject and its Subproject Awareness benefits for the community is essential to ensuring sustainability. For monetary and material maintenance contributions, lack of funds was the main impediment to contribution. For labour contributions, lack of time was Impediments to Contribution identified as the main impediment, followed by lack of funds and lack of skills. Funding may be more of an issue for labour because in certain cases the CDC must resort to paid labour. Community Type Key Findings The level of wealth or poverty does not appear to have any impact on subproject Level of Wealth or Poverty condition or maintenance activities performed in sample communities. Communities with properly maintained subprojects had a slightly lower average number of projects per community than communities with poorly maintained Average Number of Subprojects subprojects. The average number of subprojects built has little effect on subproject condition. 84 The number of households in a given community has no visible effect on subproject condition, most likely due to the increased size and complexity of subprojects that are built for larger communities. Number of Households Communities where subprojects were poorly maintained had a slightly higher average population than communities where subprojects were properly maintained. Subprojects built in ethnically homogeneous communities have a greater chance Ethnicity of being maintained than subprojects in ethnically heterogeneous communities. Subproject Type Key Findings Estimated O&M cost (at the village level by CDC members) has no effect on Estimated O&M Costs maintenance activities or subproject condition, although more costly to maintain subprojects are overall in worse condition than less costly projects. As communities must pay for a minimum of 10% initial subproject cost, the Average Construction Cost higher the investment the more likely they are to properly maintain the subproject. More technically complex subprojects from the sample group were better Technical Complexity of Subproject maintained than less complex subprojects. The initial quality of construction also has an effect on maintenance activities. Quality of Initial Construction Good-quality subprojects are 28% more likely to be properly maintained than poor quality subprojects. Subprojects implemented under NSP II have a higher chance of being better NSP I vs. NSP II maintained than NSP I subprojects. SUBPROJECT O&M COST Based on the three O&M scenarios outlined in the section above, communities cannot realistically be expected to pay for O&M for complex assets including tertiary roads and large scale water supply networks. Although operating the subprojects can be affordable for most communities, putting money aside for replacement and reconstruction costs is not. Without funding help from line departments or other sources, large scale subprojects are not sustainable in the long term. If funding is not provided from other sources such as line departments, the main problem for subproject sustainability relative to O&M cost seems to be lack of money put aside in the case of an eventual breakdown. Subprojects will require substantial yearly investments from community members if they are expected to pay for their replacement upon failure. Collecting this money will be difficult and will require serious efforts on the part of CDC members and the FP to convince community members that they must contribute funds every year even though the subproject may be functioning properly. Furthermore, community members will not contribute these funds if they do not have a significant amount of trust in the CDCs capacity to manage funds, as well as in its accountability and transparency. The table below, Table 39, presents key findings related to subproject O&M cost by subproject type, outlining which subprojects communities are likely able to maintain on their own based on cost of maintenance. 85 Table 39 - Key Findings: Subproject O&M Cost Subproject O&M Cost Key Findings Canals, Culverts, Karezs, and Protection Where labour is paid for by the CDC, cost to operate and maintain canals and Walls karezes becomes too expensive for community members. The costs of replacing motors, valves, turbines, and pen stock pipes may be too Micro-Hydro Plants expensive for certain communities, depending on possible O&M monetary contributions and size of the MHP. While most communities are capable of paying for replacement batteries, the eventual cost of replacing an entire solar panel system (i.e. the panels Solar Panels themselves) is too expensive for the assessed communities based on reported O&M contributions. Maintenance for roads requires a serious labour investment from community members, including replacing gravel moved off the side of the road and cleaning side drains. If roads are not maintained they will require re-gravelling. Re- Tertiary Roads gravelling is much too expensive for community members to undertake on their own, potentially driving up the cost of O&M up to millions of Afghanis depending on the length of the road. The cost of replacing pipes over the life-span of a water supply network is too expensive for most communities to undertake. Even if money is put aside every Water Supply Networks year, it will not cover the eventual replacement costs of the entire system. In certain cases, the operating budget in the form of generator fuel or city electricity fees was too expensive for communities. CONCLUSIONS AND STRATEGY RECOMMENDATIONS GENERAL CONCLUSIONS The communities each exhibited different levels of organisational, technical, and financial capacity to implement and maintain subprojects independently, with some requiring no further involvement from the FP, other than distributing the grant for subproject construction, and some in dire need of assistance. In practice, based on the key findings above, the majority of communities reside somewhere in between those extremes, and would benefit from some level FP assistance to improve their capacity to effectively steward sub-project assets. Moreover, very rarely are communities able to collect enough resources to cover the cost of O&M. Thus, as most communities will likely require some level of assistance to harmonise their respective capacity with that set forth in the NSP Operations Manual, the role of the FP becomes crucial in ensuring that institutions and practices tailored to the particular reality of each community are created, including adequate and ongoing training and comprehensive planning to prepare for O&M activities. Indeed, the FP must be involved not only before and during subproject implementation to ensure that essential mechanisms are put into place, but should also continue to follow-up with communities who require additional guidance in terms of organising labour, collecting resources, or planning for short and long-term maintenance. 86 STRATEGY RECOMMENDATIONS The following strategy recommendations are directed at NSP and as a corollary, the FPs, and based on the key findings presented above. The recommendations are intended to improve subproject construction condition and maintenance activities, to mitigate O&M costs at the community level, and to help ensure that subprojects are maintained and managed independently by communities, as per NSP guidelines. Recommendations are separated into 3 phases relative to the subproject lifecycle, as shown below in Figure 75: pre-project planning, subproject implementation, and post-project completion. The pre-project planning phase includes all activities up to subproject selection, the implementation phase all activities related to development of organisational structures such as the CDC and various committees, training, and development of the O&M plan, up to completion of subproject construction, and the post-project completion phase includes activities that should take place after the subproject has been constructed. Figure 75 - Subproject Implementation Timeline PRE-PROJECT PLANNING Selecting a community for NSP should involve an initial assessment by the FP of the size of the community, potential contributions to O&M, previous subprojects constructed, and the technical capacity of community members. Where communities have no members with experience in construction or expertise in maintaining similar subproject, plans should be made for more intensive training and coaching. Where possible, subprojects with high O&M costs should be planned on a large scale for multiple communities to mitigate cost per community member. In addition, community members should be approached at the pre-project planning phase to determine capacity to contribute to subproject O&M in terms of labour, materials and funds. These contributions should then be tested against actual O&M costs, which should be calculated to include the eventual replacement or breakdown of the subproject (i.e. costs should include a yearly fee that communities must contribute in order to prepare for the replacement of the subproject in the case of an eventual breakdown), and based on the model provided in this report. 87 Subprojects eventually proposed to the community should meet the above criteria in terms of O&M costs, technical capacity, and community investment, and subprojects that respond to the needs of the entire community should be prioritised to reduce demand for paid labour. Finally, if tertiary roads are selected, links should be established with the relevant line ministry to ensure contributions to O&M as in no scenario will community members be able to afford maintenance costs for roads. Alternatively, only basic access roads (un-gravelled roads) should be selected. To facilitate this, NSP should provide more detailed community selection guidelines for FPs, including a “Community Assessment and Suitability” form that must be completed and approved before any grants are distributed. SUBPROJECT IMPLEMENTATION During subproject implementation, the following steps should be taken in order to ensure good subproject condition and effective maintenance. First, the FP should ensure that an O&M committee has been formed within the CDC, ideally with members experienced in construction and repairs and with similar subprojects, and an O&M officer should be appointed. Following the creation of these institutional bodies, a comprehensive O&M plan should be developed with the CDC and with the participation of all community members. Specific O&M tasks should be discussed and outlined in the plan, with associated costs in terms of labour, materials and especially funds, with all community members signing on to contribute what they can to O&M activities. Ensuring buy-in from all community members and clearly assigning O&M responsibilities at this stage should help to mitigate potential conflicts relating to O&M activity contributions at a later date. As part of this process, a user fee structure should be put into place with varying degrees of contributions based on individual household wealth, where relevant. Subproject construction should be closely monitored by the O&M committee and the O&M officer, and should a contractor be selected this process should be supervised by the FP and the CDC. O&M committee and CDC members should also be involved in the actual construction process, working alongside FP engineers in order to gain a more complete understanding of the subproject, particularly for more technically complex subprojects; supervision and increased involvement will also help to increase initial construction quality, a key factor for the performance of future effective maintenance activities. Finally, all relevant key components of a subproject should be constructed, such as side-drains and culverts for tertiary roads, cemented blocks for MHP supply pipes, and top rings for wells, to minimise future maintenance costs. Comprehensive and ongoing training should be provided to the O&M committee and CDC members, and maintenance activities should be communicated and explained to all community members. If, during the pre-project planning phase, no community members have experience with the subproject, a more intensive and comprehensive training schedule should be implemented. As O&M Committees and Officers are required under NSP guidelines, to ensure that the FP facilitates the creation of the above institutional bodies, NSP should penalise FPs that do not follow through with these necessary steps in subproject implementation. Furthermore, an outline of a more detailed O&M 88 plan, including yearly costs and contributions divided amongst community members, should be developed by NSP and distributed to FPs, and a more detailed training schedule should be developed for each subproject type, with more rigorous and basic training for communities with no experience with similar subprojects. Finally, NSP should develop a series of strict guidelines to ensure that subproject construction is monitored and managed by communities, with any required assistance from the FP, and that CDC and O&M members are involved throughout the entire implementation phase. POST-PROJECT COMPLETION Currently, little to no contact between the FP and communities occurs after subproject completion. Ideally, communities were expected to manage all O&M activities independent of support from the FP, although results from the assessment in this report show that many subprojects are poorly maintained due to lack of funds collected, capacity, or willingness. Thus, the FP should follow-up with communities, filling any gaps in training or expertise required by the CDC and community members. Even if the process described above is closely followed, problems with O&M may occur, particularly when no members have previous experience maintaining a technically complex subproject. The FP should revisit communities on a monthly basis after subproject completion to address any developing issues or concerns. These visits could be reduced and eventually phased out if the community demonstrates effective independence in subproject maintenance and management. In the long-term, subprojects should be linked with relevant line departments, and an official (most likely an engineer) from these departments could inspect subprojects on regular basis, preparing reports on O&M needs and sharing best practices amongst communities. Regular inspections could also provide an opportunity to build an asset database for large scale assets (water supply networks, micro-hydro plants, and tertiary roads) at the district level, contributing to effective country-wide budgeting for O&M costs. NSP should plan for FPs to follow up with communities, and in some cases (for tertiary roads, MHPs and WSNs) a percentage of the initial grant should be set aside for short-term O&M costs. In addition, training and assistance should be provided as required by the FP in the months following the subproject. NSP could visit communities to ensure that proper support is being provided post-construction. Finally, based on the results discussed in this report, it is unlikely that communities will be able to pay for and manage O&M in the long-term, particularly for larger, expensive and more technically complex subprojects. Thus, NSP should work with line ministries in order to pass over management of maintenance activities. 89