Access to clean water is a fundamental requirement for public health, urban resilience, and sustainable development. In Indonesia, the provision of drinking water services is institutionally supported by regional water utilities, including PERUMDA, which are responsible for ensuring quality, quantity, continuity, and affordability of water supply. However, many utilities still face persistent inefficiencies in distribution systems, particularly through Non- Revenue Water (NRW). NRW refers to water that has been produced and distributed but does not generate revenue because of physical losses, commercial losses, or unbilled authorized consumption. This condition weakens utility performance because it reduces income, increases operational costs, and limits the capacity to expand services. At the global level, NRW has become an important indicator of water utility efficiency and infrastructure sustainability. WHO and UNICEF (2025) emphasize that safely managed drinking water remains unevenly distributed, making water-loss reduction increasingly important for achieving SDG 6. Santos (2024) explains that NRW affects not only technical performance but also the economic sustainability of water utilities. In developing countries, high NRW is commonly associated with aging pipes, inaccurate metering, weak monitoring systems, and unauthorized consumption. Therefore, reducing NRW is not only a technical agenda but also a strategic requirement for improving financial resilience and service reliability.

The Indonesian water sector faces a similar challenge, as NRW in many PDAM and PERUMDA systems remains above the expected performance threshold. Luthfianto et al. (2025) reported that the average NRW level of Indonesian water utilities reached 33.72% in 2021, exceeding the RPJMN 2020– 2024 target of 30%. In the case of PERUMDA City X, internal operational data showed an average NRW of 34.16% in 2025, with a wide inter-installation disparity ranging from 15.53% to 59.51%. This variation indicates that NRW is not caused by a single uniform factor, but by different technical, administrative, and operational conditions across service zones. Such a condition requires a systematic prioritization approach before control strategies can be designed and implemented effectively. Previous studies have examined NRW from different perspectives, including water balance analysis, leakage control, pipe deterioration, and district-based monitoring. Zafira and Nurhayati (2023) highlighted the relevance of District Meter Area and SCADA-based monitoring for improving NRW control. Nanda et al. (2024) showed that factor prioritization is useful for identifying the most influential causes of NRW in regional water utilities. Shell et al. (2025) found that physical losses remain a major contributor to water loss in Indonesian distribution systems. However, these studies have not fully integrated dominant-factor weighting, causal interpretation, and priority-based control strategies within one structured decision-making framework.

Therefore, this study aims to identify and weight the dominant causal factors of NRW at PERUMDA City X using the Analytical Hierarchy Process. The study also seeks to analyze the relationship among priority factors so that the causes of NRW can be understood as an interconnected system rather than isolated problems. Furthermore, this research formulates priority-based NRW control strategies supported by simulation projections of potential reduction impacts. The contribution of this study lies in combining AHP-based prioritization with practical intervention planning for regional water utilities. The findings are expected to support more targeted, efficient, and evidence- based NRW reduction programs.

Non-Revenue Water and Water Utility Performance

Non-Revenue Water is a critical indicator of the technical, financial, and managerial performance of water utilities. It represents treated water that enters the distribution system but does not generate revenue because of physical leakage, apparent losses, or unbilled authorized consumption. Recent studies show that high Non-Revenue Water reduces utility income, increases operational expenditure, and weakens the capacity of water companies to reinvest in infrastructure. Naradipta and Slamet (2024) found that Non-Revenue Water in PERUMDA Delta Tirta Sidoarjo reached 31.67% in 2023 and caused substantial annual revenue losses. This finding confirms that Non-Revenue Water should be treated not only as a technical problem, but also as a strategic issue affecting institutional sustainability. H1: Non-Revenue Water is influenced by technical, administrative, operational, and unbilled authorized consumption factors in regional water utilities

Technical Loss Theory

Technical losses are generally associated with physical water losses caused by pipe leakage, pipe bursts, corrosion, unstable pressure, and aging infrastructure. In water distribution networks, deteriorated pipes tend to increase leakage frequency because structural weakness reduces pipe resistance to internal and external pressure. Wibowo and Slamet (2024) explain that pressure management is essential in controlling physical leakage because higher pipe pressure can increase the amount of water lost through existing leaks. Spedaletti et al. (2022) also demonstrated that monitoring water flow through the District Metered Area approach can help identify water losses more accurately and support energy efficiency. Therefore, technical factors are theoretically expected to become the most dominant causes of Non-Revenue Water when distribution infrastructure is old, poorly monitored, or hydraulically unstable.

Operational Control and District Metered Area Theory

Operational control determines how quickly water utilities detect, localize, and respond to losses within the distribution network. District Metered Area is widely recognized as a practical operational approach because it divides a distribution system into smaller zones that can be monitored through inlet flow measurement, pressure analysis, and minimum night flow evaluation. Mahardini and Tangahu (2023) reported that water loss in District Metered Area Kalipuro 5 decreased after applying monitoring, pressure management, repair, and active leakage control. Puspa et al. (2025) also emphasized that District Metered Area optimization requires hydraulic modeling, logger installation, Supervisory Control and Data Acquisition integration, asset security, and updated geographic information system data. These studies support the argument that the absence of District Metered Area can delay leakage detection and make Non-Revenue Water control less efficient. H4: The absence of District Metered Area weakens operational control and increases the difficulty of reducing Non-Revenue Water.

Priority-Based Decision Framework for Integrated NRW Control

The importance of this study lies in its effort to connect factor identification, priority weighting, and control strategy formulation into one integrated decision framework. Many previous studies have focused on water balance, Infrastructure Leakage Index, pressure management, or District Metered Area evaluation, but fewer studies have combined these technical insights with structured priority analysis. Fathani et al. (2025) showed that water balance and performance indicators are useful for formulating Non-Revenue Water reduction strategies, but such approaches still require prioritization when several causes appear simultaneously. Vaquet et al. (2024) argued that leakage detection in water distribution networks is complex because leakage behavior is dynamic and difficult to identify, especially for small leaks. Therefore, an Analytical Hierarchy Process-based framework is important because it helps water utilities transform complex field problems into ranked priorities and practical intervention strategies. H6: Priority-based control strategies derived from Analytical Hierarchy Process results can support more targeted and integrated Non-Revenue Water reduction.

Figure 1. Conceptual Framework

Research Design and Location

This study employed a quantitative descriptive approach to identify and prioritize the dominant factors contributing to Non-Revenue Water (NRW) at PERUMDA City X. The design was selected because the study aimed to transform expert judgment and field-based findings into measurable priority weights using the Analytical Hierarchy Process (AHP). Primary data were collected through field observations conducted over 42 working days across seven Water Treatment Plants from January 19 to March 20, 2026. In addition, in- depth interviews and AHP questionnaires were administered to the Technical and Development Manager, who was selected as the key expert due to institutional authority, technical competence, and direct responsibility for distribution network management. Secondary data were obtained from the official 2025 NRW report, operational data from each installation, and infrastructure profile documents of PERUMDA City X.

Population, Sample, and Sampling Technique

The population of this study consisted of the water distribution system, operational units, and managerial decision-makers involved in NRW control at PERUMDA City X. The observed technical population included seven Water Treatment Plants representing different service zones and NRW characteristics. The respondent sample consisted of one key expert selected using purposive sampling, a non-probability sampling technique commonly applied when the research requires specialized knowledge and decision-making authority. This sampling technique was considered appropriate because AHP relies on informed expert judgment rather than large respondent numbers. The selected expert represented the managerial and technical perspective needed to assess pairwise comparisons among NRW causal factors.

Research Procedure

The research was carried out in several sequential stages. First, preliminary identification was conducted by reviewing NRW literature, official utility documents, and operational data from PERUMDA City X. Second, field observations were conducted for 42 working days to understand the actual conditions of the seven Water Treatment Plants and their distribution networks. Third, the main factors and sub-factors were validated through expert interviews and then structured into an AHP hierarchy. Fourth, pairwise comparison data were collected using the AHP questionnaire and processed to obtain local and global priority weights. Finally, the priority results were interpreted to formulate control strategies and simulate the potential reduction of NRW under integrated implementation.

NRW Profile at PERUMDA City X

The official Non-Revenue Water (NRW) data for PERUMDA City X in 2025 are presented in Table 2. The data show that the average NRW level reached 34.16%, indicating that more than one-third of distributed water did not generate revenue. The highest NRW was recorded at Bulotadaa I with 59.51%, while the lowest was found at Dungingi with 15.53%. This wide range confirms that NRW conditions differed substantially among installations. These findings support H1, which states that NRW is influenced by a combination of operational, technical, administrative, and unbilled authorized consumption factors across service zones.

Priority Weighting of Main Factors

The main factor weighting using the Analytical Hierarchy Process (AHP) is presented in Table 3. The Technical Factor obtained the highest priority weight of 0.558, making it the most dominant factor in NRW causation. The Administrative Factor ranked second with a weight of 0.263, followed by the Operational Factor with 0.122. Unbilled Authorized Consumption ranked lowest with a weight of 0.057. The Consistency Ratio of 0.0654 is below the acceptable threshold of 0.10, indicating that the expert judgment was consistent and reliable.

The values in Table 3 show that technical problems contribute more than half of the total priority weight. This result confirms H2, which states that technical factors have the highest priority weight among the main causes of NRW. The relatively high administrative weight of 0.263 also supports H3, because illegal connections, meter problems, and recording errors remain relevant sources of apparent losses. The operational factor, although ranked third, still has strategic importance because network monitoring and District Metered Area implementation influence the speed of leakage detection. The low weight of Unbilled Authorized Consumption indicates that this factor exists but is not the primary driver of NRW in PERUMDA City X.

The local weight results show the most urgent sub-factor within each category. The absence of District Metered Area supports H4 because it weakens operational control and makes leakage localization more difficult. Network leakage strongly supports H2 because it represents the most important technical sub-factor contributing to real losses. Illegal connections support H3 because they directly reduce billable consumption and increase apparent losses. All Consistency Ratio values are below 0.10, indicating that the pairwise comparison results for each sub-factor group are acceptable for further interpretation.

Causal Relationship Among Dominant Factors

The ranking results indicate that the dominant causes of NRW form an interconnected causal pattern. Pipe age/corrosion weakens the physical condition of the distribution network and increases the probability of leakage. Network leakage then reduces pressure stability and may create additional hydraulic stress within the system. Illegal connections intensify this condition because unmeasured consumption can disturb pressure balance and increase water losses in affected zones. This causal interaction supports H6, which states that priority-based control strategies are needed because NRW is driven by interrelated factors rather than isolated causes. The absence of District Metered Area also plays an important role in the causal structure. Without District Metered Area zoning, leakage cannot be localized quickly and accurately. This condition causes leakage duration to become longer and increases the volume of water lost before repair actions are taken. The same limitation also affects the detection of illegal connections because utilities lack reliable zone-level monitoring. Therefore, H4 is also supported because the absence of District Metered Area weakens operational control and reduces the effectiveness of NRW management.

Hypothesis Summary

The results of this study provide empirical support for all proposed hypotheses. H1 is supported because the NRW profile shows that water loss varies across installations and is influenced by several causal dimensions. H2 is supported because technical factors obtained the highest main factor weight of 0.558 and network leakage became the highest-ranked sub-factor. H3 is supported because illegal connections ranked second globally and represented the most important administrative factor. H4 is supported because the absence of District Metered Area became the strongest operational sub-factor and was identified as a negative multiplier in NRW control. H5 is supported because the AHP process produced consistent results, as indicated by all Consistency Ratio values below 0.10. H6 is supported because the simulation demonstrates that priority-based integrated strategies can reduce NRW substantially. The combined intervention programs are projected to reduce NRW by 12.8 to 17.4 percentage points. This result confirms that control strategies should be designed according to ranked causal priorities rather than implemented uniformly across all factors. Overall, the findings show that AHP is effective for transforming complex NRW problems into structured, measurable, and actionable decision priorities.

The dominance of the Technical Factor, with a weight of 0.558, confirms that physical infrastructure conditions are the most important source of NRW in PERUMDA City X. This result is consistent with leakage theory, which explains that pipe deterioration, pressure fluctuation, and network leakage increase real losses in distribution systems (Liang et al., 2021). Vaquet et al. (2024) argue that leakage in water distribution networks is difficult to detect because it is dynamic, time-dependent, and often hidden within complex hydraulic behavior. The present study strengthens that argument by showing that network leakage became the highest-ranked sub-factor, with a global weight of 35.88%. Thus, H2 is supported because technical factors were empirically proven to have the highest priority weight among the main causes of NRW (Fadli and Rahma, 2023). The result is also consistent with Indonesian studies that identify physical water loss as a major contributor to high NRW. Rahman et al. (2022) found that PERUMDA Delta Tirta Sidoarjo experienced NRW of 31.67% and suffered substantial annual revenue loss due to physical and non-physical water losses. Fathani et al. (2025) also reported that physical water loss dominated NRW in the Pilang Raya service area of PERUMDA Tirta Giri Nata Cirebon, where water loss performance was categorized as highly inefficient. The similarity between those studies and the present research lies in the central role of infrastructure leakage as a major NRW driver. However, the difference is that this study not only identifies the magnitude of water loss, but also ranks causal factors using AHP and links the ranking to priority-based control strategies (Lutfi et al., 2023).

The importance of administrative control in this study is also comparable with the findings of Prastiwi et al. (2024), who examined NRW control through standard operating procedures in PERUMDAM Tirta Tarum Karawang. Their study found that several NRW control procedures were already implemented, but the system remained relatively passive because action was often taken after public complaints. The present study confirms a similar managerial challenge because illegal connections and damaged meters appear among the dominant sub-factors requiring proactive control. The difference is that Prastiwi et al. (2024) focused on procedural effectiveness, while this study converts administrative causes into measurable AHP priority weights. This distinction makes the present research more useful for determining which administrative problems should be handled first under limited resources (Hendri et al., 2023). The AHP results confirm that decision-making in NRW control requires structured prioritization rather than general intervention planning. Macchiaroli et al. (2023) state that the Analytical Hierarchy Process is useful for ranking water infrastructure investments because it can incorporate technical, managerial, and sustainability considerations into one decision framework. The present study supports that theoretical position because all Consistency Ratio values were below 0.10, indicating that the expert judgments were logically consistent. This finding supports H5, which states that AHP provides a consistent decision-making framework for ranking dominant NRW causal factors. The attractive contribution of this study lies in its ability to translate a complex field problem into measurable priority weights that can directly guide practical intervention planning (Yunita et al., 2022).

This study concludes that Non-Revenue Water (NRW) at PERUMDA City X is mainly driven by technical, administrative, and operational factors. The Analytical Hierarchy Process (AHP) results show that the Technical Factor is the most dominant factor, followed by Administrative, Operational, and Unbilled Authorized Consumption factors. The three most influential sub-factors are network leakage, illegal connections, and pipe age/corrosion, which together account for 70.71% of total causal priorities. These findings indicate that NRW reduction should be implemented through priority-based strategies, including District Metered Area development, active leakage detection, phased pipe replacement, illegal connection enforcement and legalization, pressure management, and periodic meter calibration. The implementation of these integrated strategies is expected to support more targeted NRW control and improve the operational and financial performance of PERUMDA City X.