Cost-effective non-revenue water reduction: analysis through pilot activities in Kigali City, Rwanda

The Sustainable Development Goal target 6.1 for drinking water is that ‘By 2030, achieve universal and equitable access to safe and affordable drinking water for all’ (United Nations 2024); this requires efficient water supply with a reduction in water waste. In recent years, water supply through pipe networks has been developed in urban areas of developing countries. Owing to this, problems with inefficient water use related to the fixed tariff system have been identified. Studies in Pakistan and Tajikistan (Ogata et al. 2021; Ogata et al. 2023) report that the adoption of a fixed tariff system induced insufficient drinking water quantity and water pressure as customers kept taps open or used tap water for agriculture.

Although it is essentially difficult to calculate the non-revenue water (NRW) ratio under a fixed tariff system, Kanakoudis et al. (2015a) stated that the NRW ratio increased up to 70% of the NRW when the fixed tariff was expressed in equivalent water volume. NRW is a critical issue that must be addressed by water utilities worldwide to ensure a sustainable water supply. When treated water produced from water treatment facilities does not reach customers, it is considered NRW (Ab Malek et al. 2021). A high NRW ratio leads to higher payment rates for consumers to compensate for the inefficiency in water management and keep water utilities in operation, which is a disadvantage for the majority of consumers in low-income countries (Tumuheirwe & Lutaaya 2006). The impacts of high NRW ratios are economic (lost revenues), environmental (water and energy losses), and social (inefficient water pricing policies not based on the actual water consumption profile/patterns) (Kanakoudis et al. 2015b). In addition, the cost of improving service delivery is much lower when implemented through investment in NRW reduction than through investment in capital projects to increase supply capacity (Frauendorfer & Liemberger 2010). There have been several cases of successful NRW reduction, such as Phnom Penh Water Supply Authority in Cambodia, which reduced its NRW ratio from 72 to 6% between 1993 and 2008 (Chan 2009), Manila Water in the Philippines, which reduced its NRW ratio from 52 to 30% between 1997 and 2006 (Marin 2009), and Dhaka Water Supply and Sewerage Authority in Bangladesh, which reduced its NRW ratio from 40 to 27% between 2008 and 2013 (Farok 2017).

Numerous studies have also reported the specific methods of NRW reduction. Boztaş et al. (2019) reported that 77.4% of the total leaks in the water distribution system in Turkey were identified at the service connection, which was repaired to achieve substantial NRW reduction. Yi et al. (2017) recommended NRW reduction in Myanmar through capacity building of water supply staff to optimize the efficiency of water services, including water quality, water pressure and facility maintenance. Gonelas & Kanakoudis (2016) highlighted the importance of pressure management as one of the most effective NRW reduction strategies. The experiences of Manila Water in the Philippines showed that substantial reduction in NRW was attributed to its increased meter reading efficiency and improved billing volume by putting more resources into training meter readers as well as aggressive leak repair activities, metering of illegal connections, replacement of defective meters, and closing of illegal connections (Inocencio & David 2001). Cervancia et al. (2022) recommended documenting and monitoring ageing infrastructure, controlling inaccurate meter readings, establishing meter repair and replacement programs, and initiating regular monitoring of illegal connections as NRW reduction methods in rural Philippines. Cassidy et al. (2021) recommended reducing NRW in Chile through digitization. They showed how the use of real-time monitoring tools enabled the classification of burst incidents and assessed their impact on water loss volumes and identification of operational inefficiencies, particularly in relation to detection and repair times in small and medium burst incidents. In addition, the introduction of an integrated customer meter management tool optimized meter management, reduced apparent losses through more accurate estimation of meter errors, and enabled water utilities to replace meters based on specific lifetimes.

However, reducing NRW and preventing unplanned water supply interruptions are particularly costly for water utilities (Maziotis et al. 2023). Therefore, to implement regional NRW reduction activities, it is necessary to demonstrate its effects through small-scale pilot projects.

Previous studies have analyzed the economic benefits of NRW reduction. Idogho & Olotu (2013) calculated the economic benefits of modeling NRW data in Nigeria. They estimated that the NRW rate in the target area decreased from 50.7% in 2007 to 10.6% in 2011, resulting in an increase of 4,400 m³ in revenue water and a savings of USD 17,400. Indah & Karpriana (2021) calculated NRW as opportunity costs by gathering information from the financial, management, and equity participation reports of an Indonesian city and estimated that the potential revenue from reducing the NRW to 20% would be INR 222 billion. The WATERLOSS-DSS, a decision support system (DSS) for determining NRW reduction measures was developed and used for a pilot study in Greece and Cyprus (Kanakoudis et al. 2015c, 2016). The DSS platform includes the DSS tool, which (a) proposes a list of prioritized NRW reduction measures; (b) evaluates the network's performance variables and indicators; (c) compares and benchmarks water distribution networks’ performances; (d) manages the registry of NRW reduction measures; and (e) induces the measures prioritized for any specific system (Kanakoudis et al. 2015b). By implementing NRW reduction activities based on this system, the annual water savings were estimated to reach 1 million m³ and the economic benefits were estimated at more than EUR 700,000.

The objectives of this study were to identify the factors necessary for effective NRW reduction and establish a practical methodology to measure its cost-effectiveness through the implementation of technical cooperation projects.

This paper refers to the ‘JICA Project Brief Note, Rwanda, Project for Strengthening Non-revenue Water Control in Kigali City Water Network’ (JICA 2022) and other related documents (JICA et al. 2021, 2022), which describe the background and results of the technical cooperation project supported by JICA. In addition, this study reviewed the NRW reduction activities in other countries and compared their methodologies with those of the current study. We also updated the analysis methodology, particularly the recovery of NRW reduction costs. A description of ensuring the accuracy of customer meters for the calculation of the NRW ratio was also added.

A management team was assembled to organize the implementation of the project activities, consisting of a team of JICA experts, the WASAC director of Urban Water and Sanitation Services, the director of Commercial Services, the financial officer, and heads of each department related to NRW reduction. An action team consisting of the staff of WASAC departments relevant to NRW reduction operations was also organized. The project established four desired outcomes for the WASAC: (1) enhancement of planning capacity for NRW reduction; (2) acquisition of basic knowledge, skills, and techniques for NRW control; (3) enhancement of staff capacity to conduct NRW reduction measures by implementing the pilot project; and (4) establishment of a system to accurately measure NRW ratios. Technical capacity development of water utility staff was also recommended in previous studies on NRW (Inocencio & David 2001; Yi et al. 2017), and considered an essential element for NRW reduction.

First, a 5-year strategic plan (5YSP) was developed to determine the necessary actions for achieving NRW reduction and enhancing the planning capacity of WASAC. To formulate the 5YSP, the current status of water services was surveyed. This was followed by a series of workshops held to identify the problems that the WASAC faced in addressing NRW challenges. Based on these findings, specific activity plans for NRW reduction were developed. The 5YSP, approved by the WASAC Board of Directors, was disseminated to all its 20 branches to initiate specific activities. The implementation status at each branch was monitored through quarterly reports.

Second, training was conducted to deliver basic knowledge of NRW, technical skills related to NRW management, geographic information system (GIS) database updates, hydraulic analysis, leak detection, pipe repair, customer meters, and billed water volume management. Leak detection equipment, pipe repair equipment, and customer meters were provided for classroom training exercises. On-site training using existing equipment was also provided during the pilot project.

The activities conducted for NRW reduction included surveillance of water use by customers, customer meter inspection and replacements, water leakage surveys using leak detection equipment and repairs, installation of pressure-reducing valves (PRVs) for pressure control, and replacement of water distribution and service pipes. Training opportunities for WASAC staff were incorporated within the implementation of NRW reduction activities.

The specific steps for NRW reduction implemented in the pilot areas were as follows: Steps were taken to address physical losses (leaks), including implementation of high water pressure management; establishment of maximum hydrostatic pressure standards for facility design, construction and redevelopment of water distribution facilities in strict compliance with water pressure standards; installation and proper operation of pressure-reducing facilities, such as PRVs; strict compliance with pipe material standards; proper installation of pipes; replacement of pipes that leak frequently and old pipes; leakage surveys (minimum nighttime flow and step test); and leakage repair. To address commercial losses, WASAC maintained customer meters, conducted water meter verification based on the billed water volume analysis results, ensured the replacement of faulty meters (large errors), regularly replaced meters, optimized billed water volumes, ensured meter reading/reduced estimated billed water volume locations, investigated zero-billed water volume customers, and corrected problems.

Cost–benefit analysis was conducted for NRW reduction in the pilot areas using baseline data from June to July 2017 in the Kadobogo area (administrative area of the Kacyiru branch) and from March to April 2018 in the Ruyenzi area (administrative area of the Nyarugenge branch). The NRW reduction activities (first year) were conducted from April 2018 to March 2019 in the Kadobogo area and from October 2018 to September 2019 in the Ruyenzi area. To assess the impact of the activities, the NRW ratio was monitored in the Kadobogo area from April 2019 to March 2020 and in the Ruyenzi area from October 2019 to March 2019. The amount of NRW reduced through the pilot activities was defined as the difference between the actual monthly amount of NRW (with project) and the predicted amount of NRW (without project) in the pilot area from the baseline period to the end of monitoring. The benefit of NRW reduction activities was calculated by multiplying the amount of NRW reduced through pilot activities by the average cost per unit of water.

Based on the NPV calculations, investment in activity was judged as appropriate if the net benefits of the project were greater than the costs and if the earnings rate was relatively high. The conditions for determining whether the implementation of NRW reduction activities (project) was reasonable were NPV > 0 and benefit-to-cost ratio (B/C) > 1. The net benefit was defined as the increase in tariff revenue due to a reduction in NRW. Capital expenditure in the first year included hydraulic separation for DMA formation, installation of flow meters and chambers as preparations for NRW reduction, leakage investigation and repair, PRV installation, and pipe renewal. As for the operating expenditure, leakage survey and repair costs incurred within the project and leakage repair costs incurred independently from the project were accounted for. For the water unit price, the sales prices (2018 averages) of Kacyiru (Rwandan Franc (RWF) 567/m³) and Nyarugenge (RWF 592/m³) were adopted. The water price generally includes a tax, meter rental fee (RWF 100), and regulator fee (1 USD = 1,340 RWF as of August 2024). However, these items were deducted during the cost–benefit analysis.

Fourth, a remote monitoring system was established to automatically collect the water distribution volume data and calculate the NRW ratio for each branch based on the obtained flow data. The water distribution network in Kigali City is controlled by six branches of the WASAC, and it was not previously possible to determine the NRW rate of each branch individually. However, two of the six branches (Remera and Kanombe) have a remote monitoring system built using the flowmeter manufacturer's server through a project supported by the Netherlands. Therefore, it was necessary to build an additional remote monitoring system for automatic water distribution data collection and calculation of the NRW rate at the remaining four branches (Nyarugenge, Gikondo, Kacyiru, and Nyamirambo). To hydraulically separate the water distribution network by branch, surveys of geographic information system-based distribution network data and site locations for flow meter installation at the branch boundaries were conducted after obtaining a consensus from all related branches. An overall design for the remote monitoring system was developed based on the survey results. The monitoring system consisted of on-site data measurement and transmission equipment (35 electromagnetic flow meters, 1 ultrasonic flow meter, 2 mechanical flow meters, and 29 pressure gauges) installed at the water treatment plant, pumping stations, and distribution pipes at each branch boundary. A server installed at the Nyarugenge branch office collected and transmitted all the data through the cellular network of the telephone company. Furthermore, the server was connected to the WASAC's wide-area network system so that the data could be viewed and used at each WASAC office. The monitoring system was constructed in September 2021. Training was provided on the use of the software to calculate the NRW ratio using the data collected by the monitoring system.

The following activities were carried out to check and improve the accuracy of the customer meters in the pilot area: First, error measurements were carried out with portable test meters and the meter tolerance was set at ±5% according to WASAC standards. If the measurement error exceeded ±5%, the meter was to be replaced. A total of 1,242 customer meter accuracy measurement surveys were conducted in Kadobogo, but 35 could not be surveyed due to inaccessibility or customer absence, yielding a total of 1,207 surveyed meters (97%). Additionally, 35 disconnections for nonpayment and 88 faulty meters were identified. Therefore, error measurements were conducted in a total of 1,084 customer meters. The results showed that 239 m had an error of more than ±5%, and 845 were within the acceptable range. Thus, 327 m (239 + 88) were replaced and the final average error of a customer meter in Kadobogo was estimated as −0.69%. In Ruyenzi, the customer meter error measurement covered 413 large customers to enhance the efficiency of the accuracy survey. Therefore, six faulty meters were identified, and error measurements were carried out for 407 customer meters. As a result, 86 m showed an error of more than ±5%, and 321 were within the acceptable range. Therefore, 92 m (86 + 6) were replaced and the final average error of a meter in Ruyenzi was estimated as −0.43%.

Although the project was set for three years starting in August 2016, it had to be extended until September 2022 due to an unexpected delay in selecting a contractor for monitoring system construction, a one-year travel ban due to the COVID-19 outbreak, and the guarantee period of maintenance of the monitoring system. COVID-19 influenced not only the project period but also became an obstructive factor because of the restriction of movement between branches imposed since early 2020 to prevent infection.

The NRW reduction targets for the project were 17% in Kadobogo (from 37 to 20%) and 43% in Ruyenzi (from 68 to 25%). However, the results were only 12% in Kadobogo and 10% in Ruyenzi. The low rate of NRW reduction, compared with that in other countries (72–6% in Phnom Penh, 52–30% in Manila, and 40–27% in Dhaka), was a result of low performance in the pilot area, even considering the implementation period of <2 years.

Furthermore, the effective implementation of NRW activities (even in limited areas) can generate the costs for extending the area of NRW reduction activities. It would also act as the initial investment for activities, such as those dedicated to repairing service connections (Boztaş et al. (2019), or NRW reduction through digitization (Cassidy et al. 2021)).

Finally, as the project included the coronavirus disease 2019 (COVID-19) pandemic period, some additional activities to combat infectious diseases were included. During the pandemic, WASAC could not supply clean water continuously to its customers 24 h daily because of the frequent leakages and cutoff of water supply each time for such leakage repairs, insufficient water distribution, and inadequate maintenance of the water supply facilities and distribution network. Therefore, the following activities were carried out as emergency support to tackle COVID-19. First, materials and equipment needed to reduce water supply interruptions were provided to WASAC, including materials for repairing leaks, replacement of distribution and service connection pipes in the pilot area, and float valves for the water distribution reservoirs in Kigali City. Second, provision of emergency water supply to people with limited access to safe water was carried out for ∼5,200 households in the WASAC distribution network during 11 months from November 2020 to September 2021, by installing temporary water tanks and transporting water by water tankers to the areas with insufficient supplied water.

NRW reduction projects can improve the management of water utilities and have been implemented in a lot of countries. However, many utilities face challenges in substantially reducing the NRW ratio in a short period, and it is necessary to implement NRW reduction in a sustainable manner. This study demonstrated the success of a pilot project that implemented NRW reduction through the following activities: strengthening the planning capacity for NRW reduction; offering opportunities for acquiring basic knowledge, techniques, and skills for implementing NRW measures; strengthening the capacity of staff through the implementation of pilot projects; and building a system to accurately measure the NRW ratio. While reducing NRW ratios is inherently beneficial for water utilities, analyzing its costs and benefits to identify potential long-term improvements in utility management may incentivize more utilities to adopt NRW reduction measures. The cost–benefit analysis method presented in this study provides a valuable resource for future NRW reduction projects and research. However, this study has several limitations. The accuracy of customer meters in the pilot area was checked and the error was mitigated but there might be human error during data processing, which can influence the accuracy of the results. The sustainability of the activity is another issue. The project was completed in September 2022, but the data after March 2020 was not processed appropriately because of the complex tasks of the WASAC staff due to the COVID-19 pandemic. We hope that more detailed and practical methods for cost–benefit analysis for NRW reduction will continue to be developed in the future.

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

The authors declare there is no conflict.