Reactivation of public–private partnership (PPP) models for rural water supplies in Rwanda: strengthening operation and maintenance systems

The government of Rwanda established a mid- to long-term national development program known as ‘VISION 2020,’ to achieve universal access to safe drinking water for its population by 2020 (GoR 2012). In 2015, 52% of the rural population in Rwanda had access to a basic water source, while only 4% had access to a safely managed water source (WHO & UNICEF 2021). It was recognized that the vulnerability of operation and maintenance (O&M) of the system was one of the most significant challenges facing rural water supplies in Rwanda. To improve the O&M capacity of rural water supply systems, the ownership of these systems in Rwanda was transferred to the district-level government in 1987. However, a field study commissioned by the World Bank found that half of Rwanda's rural piped water supply systems were non-functional due to inadequate management and insufficient cost recovery (Prevost et al. 2010). Subsequently, the O&M of rural piped water supply systems transitioned from community-based management to management by private operators (POs) with the introduction of public–private partnership (PPP) in 2004.

The management of PPPs can be divided into several types, such as directly managed or delegated by either the public or private sector. PPPs for rural water supply can be categorized into four forms: (1) an implementation contract, wherein the operator receives a fee for operation and regular maintenance, while the publicly owned water company continues to be accountable for other responsibilities, such as undertaking new investment; (2) an affermage/lease contract, wherein the operator generates income but is obliged to pay a predetermined O&M fee to the asset owners and responsibility for operating and maintaining existing assets plus commercial and management responsibilities are transferred to the operators; (3) a construction and operation contract, wherein the water supply system is constructed, rehabilitated, and sometimes even designed, while the operation is managed under a management or lease arrangement; and (4) a concession contract, wherein the scheme is operated as a concession – the operator assumes full responsibility and has the exclusive right to operate, maintain, and carry out investment, bearing all costs and retaining all revenues for the duration of the contract (Triche et al. 2006; Kleemeier 2010; Kleemeier & Lockwood 2015; Frone & Frone 2018). The PPP for the rural piped water supply in Rwanda is implemented with reference to the affermage/lease agreement model. Despite implementing PPP schemes, in many cases in Rwanda, water supply services were stopped without proper rehabilitation and adequate maintenance of water facilities because of the lack of capabilities of the POs and the key actors in the water service. Moreover, the districts that owned the facility and were responsible for overseeing the PO lacked technical and operational management capabilities. The specific issues identified were as follows: first, the vulnerability of the current rural water supply O&M system. This included the lack of a clear delineation of the roles of each actor within the facility O&M system, the lack of established methods and criteria for selecting POs, the inadequate management of contracts between counties and POs, and the insufficient management of financial resources related to facility maintenance and repairs. The second issue related to the absence of a feasible water quality management system for the current state of the rural piped water supply in Rwanda. This included the PO's inability to conduct water quality analysis and assess the safety of water supply quality and residual chlorine management capacity. The third issue concerned the inadequate implementation capacity of various stakeholders, including asset owners (district) for water supply, asset managers (district water and sanitation officers), and insufficient capacity of POs in implementing water supply services. The fourth issue revolved around the lack of comprehensive management of water supply facility information (registry of water supply facility and pipeline maps). The fifth issue is related to the Department of Rural Water and Sanitation Services (RWSS) of the Water Supply and Sanitation Agency (WASAC), which is the regulatory organization for rural water supply, which lacked a proper data management structure within the organization, often resulting in individual staff members managing the data individually. The last issue highlighted that the existing rural piped water supply system was not equipped with flow meters, chlorination facilities, and other equipment and devices necessary for effective O&M.

Several studies have investigated PPP for rural water supplies. Effah & Chan (2013) identified and categorized the risks associated with PPP contracts based on a literature review and six case studies conducted in the Ghanaian water sector. Among the common risks, practitioners should be aware of included weak regulatory and monitoring systems, insufficient financing, the absence of risk allocation mechanisms, inexperience with PPPs, public opposition, and late or nonpayment of bills. Machete & Marques (2023) analyzed 185 water supply and sanitation projects financed by the World Bank between 2015 and 2021 and found that the most prevalent risks were fiduciary, institutional capacity for implementation and sustainability, environmental and social, and political and governance risks. Platon et al. (2014) stated that the most important economic and financial sector risks in PPPs were tariff, credit, regulatory, macroeconomic, and political interference risks. The most important commercial risk was demand risk (non-payment risk) for water and sewerage sanitation services.

Moreover, it was noted that PPP schemes must be well designed, regulated, and carefully implemented to avoid transferring much of the risk incurred in the water sector to the end users of water supply services, namely the poor customers (Frone & Frone 2013). Nyanyofio et al. (2022) examined the governance and implementation of PPPs based on management contract models with respect to the rural water supply in Ghana. The study was conducted through in-depth interviews with key management executives and focus group discussions with residents of selected communities experiencing water scarcity. The results suggested that although the PPP projects improved residents' access to clean and affordable water, fostering local economic and social development in the catchment area, they encountered challenges, such as lack of funding, political interference, limited infrastructure capacity, and pollution, which threatened their success.

This study aimed to examine the factors that can enhance the sustainability of PPPs and present the methodology, implementation results, future challenges, and recommendations for enhancing rural water supply services through a PPP project. This project was implemented by the RWSS of WASAC, with support from Japan International Cooperation Agency (JICA) from 2015 to 2019.

The project employed six steps as measures to mitigate the issues associated with PPP schemes, as listed below.

• (I) The first step involved developing the rural water supply O&M system. The activities included determining the institutional framework for rural water supply management through the formulation of a national guideline that details the specific roles of each actor, the establishment of a standardized PO selection methodology (i.e. implementing a license system, establishing water supply clusters for PO participation, and finalizing templates for tender documents), strengthening contract management by preparing a model contract form between districts and POs, and monitoring POs using a monthly report form. Moreover, financial resources were secured for facility repairs, and management methods were established by opening dedicated water accounts for each district, standardizing the facility royalty fees calculation method, and establishing the monitoring system for the management of facility royalty fees.

• (II) Second, a practical drinking water quality management framework was established, including a water quality monitoring and management system along with the maintenance of free residual chlorine. The water quality monitoring and management system followed the four core principles, namely reliability: water quality analysis was conducted by a reliable third-party organization; feasibility: a minimum of 14 water quality analysis parameters were selected, and measurements were conducted twice a year; responsibility: POs were responsible for ensuring the quality of drinking water; and responsibility burden: analysis costs were borne by the beneficiary (reflected in the water tariff). Furthermore, to enhance overall water quality management skills, the WASAC staff were trained to improve their skills as instructors in water quality management, while POs and district staff were trained to improve their practical skills in water quality management.

• (III) Third, the management capacities of the water supply facilities were improved by training the staff of the RWSS, four model districts, POs, and water user committees. The WASAC staff received individual and on-the-job training to strengthen their expertise and teaching skills to provide technical assistance to the staff of districts and POs. Strengthening the management capacity of districts and POs for water supply facilities was key to promoting the PPP model. Thus, the Project focused its training activities on improving their practical skills, including the development of training materials.

• (IV) The fourth step involved developing inventory data and mapping of water supply systems. The development of inventory data for water supply systems included data collection, data entry, and mapping of water supply facilities in 27 districts across the country. The following are the items in the inventory data developed through the Project:

• Target facilities: water sources, public taps, water kiosks, reservoir tanks, pumping stations, washout chambers, air valve chambers, valve chambers, starting chambers, collection chambers, break pressure, and pipelines.

• Attribute data: functionality (fully functional/partially functional/abandoned), construction year, year of rehabilitation, materials, specifications of the facilities, type of facility, number of users, and location data.

• Outputs: Geographic information system (GIS) mapping of the water supply systems and inventory data of water supply facilities by MS-Excel.

• (V) The fifth step involved strengthening the data management system. Typically, staff members access the organization's shared server to store data, and the server administrator handles regular backups. However, this method incurs a high cost of purchasing and operating the server, along with the time and effort required for virus protection and regular backups. To address these issues, WASAC adopted a method to share data using a free online storage service from Google Inc.

• (VI) Finally, necessary equipment, such as water meters and disinfection facilities for the O&M of existing rural water supply systems, were installed at the eight pilot sites. Pilot work was conducted to install simple water flow meters and chlorine dosing systems at model sites in each of the four districts. This implementation aimed to ensure that the staff of WASAC and the personnel in charge of the four model districts acquired the necessary knowledge and understanding of the installation processes, facilitating the dissemination of the necessary equipment to all piped water supply facilities.

A comparison of the status of water supply management in the project area before and after the project period is presented in Table 1. It includes specific achievements.

The first activity was the elaboration of the O&M framework for rural water supply systems, as shown in Figure 3. At the start of the project, a basic O&M framework was in place by the Rwandan side, as shown in Figure 3 (upper panel). The framework was based on a PPP model in which a district as the facility owner signs a delegated contract with a PO or a community-based organization. RWSS of the WASAC, the supervisory agency, provided technical assistance to the districts and the water supply operators, and RURA monitored the water supply services as a regulatory body.

However, the model did not clarify the actors' responsibilities, and there was no licensing system for POs, model delegated contracts, monthly reporting system, or royalty system from POs to the districts. The project team clarified the issues of the initial framework through a baseline survey and elaborated necessary actions to ensure the functionality of the framework (Figure 3, lower panel). The following results were achieved through the refinement of the O&M framework and were assumed to have contributed directly to the reduction in the amount of downtime in the rural water supply systems.

The first achievement was the establishment of a licensing system for POs. In 2015, 25 entities, including community-based organizations, were operating and maintaining the water supply systems in the four model districts; however, there were large differences in the operators' capabilities. A licensing system was introduced to exclude operators that did not meet the criteria for participating in the O&M bidding processes. This weeded out POs with low capacity and improved overall operator capacity to respond to breakdowns in the water supply facilities, such as timely repair of water infrastructure as stipulated in the contract, the establishment of strategic stores of necessary maintenance spare parts, and hiring technicians. The second achievement was the establishment of a royalty system from the POs to the districts. There was no royalty system in 2015. Thus, a new system was implemented by which POs would pay 10% of the revenue generated from the O&M of the water supply system to the districts, so that the districts could use the royalty payments to budget for infrastructure major repairs and rehabilitation. The royalty payment rate had increased to 89% by the second half of the project, as shown in Figure 4. The third achievement was the separation of bank accounts. Previously, revenues from water supply systems were deposited into a single account in the districts, where revenues blended with other budgets to become a general fund. Thus, the budget was not exclusively used for the development of the water sector at the end of the fiscal year. The districts opened dedicated accounts for the water sector, which created an environment in which the water sector could use the account as a specific funding source. With these improvements, budget execution for water supply facility repairs increased significantly compared to 2015.

The second activity was the basic data management required for the O&M of rural piped water supply systems. When the project started in 2015, there was no information on the water supply systems, including facility configuration, quantities, pipeline extensions, and instrumentation quantities. The development of a database reduced the time required to plan for repairing, rehabilitation, or upgrading the facilities. Thus, database management may also have contributed to the reduction in the amount of downtime. These were typical factors that contributed to the reduction in downtime. In addition, strengthening the capacity of POs, districts, and WASAC through in-house training, on-the-job training, and WASAC's technical assistance to the districts and POs was also considered a contributing factor in reducing downtime.

Specific indicators of improvement through the project, other than the reduction in the amount of downtime, are described as follows. In the model districts, 71 water supply systems were managed by licensed POs under a delegated contract. Moreover, three of the five clusters in the model districts (60%) applied for project-formatted model contracts. The average submission rate of the monthly reports from POs was 95%, enhancing the district's ability to manage PPP models as an asset owner.

In 2018, POs paid 41.3 million Rwandan Franc (RWF) in royalty fees, with an average annual collection rate of 89%. The average water tariff collection rate from users in 2018 improved to 91% from the baseline of 81% in 2015 in the model districts. The project activities led to a significant improvement in the detection rate of free residual chlorine, an indicator of water supply quality safety, rising from 12% at baseline (2015) to 72% in 2019. For database development, the project provided technical training for district water supply and sanitation support engineers of the WASAC engaged in on-site data collection and completed the rural water supply inventory and the GIS map database for 27 rural districts (1,064 systems). Moreover, the creation of an O&M manual for the 63 water supply systems in the model districts managed by POs was completed by the project's conclusion. Among the risks associated with PPPs identified by Effah & Chan (2013) were the weakness of the regulatory and monitoring system that could have been reduced through the development of guidelines and improved data management; the absence of a risk allocation mechanism could have been reduced through contract refinement and the development of a model contract form; and residents' opposition and delay of water tariff payment or nonpayment could have been improved by enhancing the maintenance capacity of POs including water quality improvement and service improvement through reduced downtime. Inexperience with PPPs can be mitigated by sharing the experience gained in model districts with those outside the project area. Finally, securing funding for PPPs may become more accessible by demonstrating sound PPP operations and management to funders or donors.

In the project's approach to vulnerable people, each district, as the asset owner, compiles a list of socially vulnerable people, and POs implement exemptions and/or preferential treatment for water payments. Such measures had already been established in Rwanda, eliminating the need for the project to establish a separate policy of safety nets for the vulnerable. The project's model contract form outlines specific activities, mandating the districts to create a list of socially vulnerable people exempt from tariffs, and ensures that all residents, including the poor, have the right to access safe water.

As noted above, there was a need for comprehensive activities from guidelines to be developed at the national level to monthly reports by POs on the ground. While modifications are necessary to adapt the model to local characteristics, the established models should be disseminated throughout Rwanda in the future.

It should be noted that the management of wastewater by POs has not been implemented at the date of publication, but the WASAC initiative was moving toward wastewater management at the rural level and some villages have already developed wastewater management systems with their own frameworks.

Although justifications for the approaches considered for the project have been provided, further recommendations for future activities for PPPs in the rural piped water supply include the following. First, consideration should be given to extended technical support for screening and monitoring of POs on the provincial level. Second, a national-level appointment of an auditor for PPPs on the rural piped water supply should be considered who should audit the O&M performance of the sample of the systems throughout the country on an annual basis. The results of the annual audit would allow the government to exclude those POs not performing well and enable it to learn lessons from the successfully rated POs.

PPPs in the rural pipe water supply involve various stakeholders, including central government, local governments (districts), POs, and residents using the facilities. Therefore, in many cases, simply implementing a PPP system is not sufficient to successfully manage the entire water supply facility. Sustainability requires coordination mechanisms among stakeholders, clarification of their roles, and the development of strategies and guidelines tailored to the current situation. Moreover, capacity building of the stakeholders according to their roles ensures the smooth functioning of PPPs. As a result of various project activities, the annual downtime of water supply services in the model districts was reduced by 41%, which was inferred to be mainly related to the establishment of a licensing system for POs, an improved reporting mechanism and royalty payment strategy, dedicated bank accounts for water, and the development of an asset inventory. In addition, 71 water supply systems became managed by licensed POs under delegated contracts, and the detection rate of free residual chlorine increased from 12 to 72%. This case study on revitalizing the PPP model for village piped water supply in Rwanda is an example of a methodology that can be used for the sustainability of the PPP models.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.