Seasonal Dynamics of Raw Water Quality and Pollution Load in the Sepaku River: Implications for Treatment Costs in Nusantara, Indonesia’s New Capital

The sequence of our studies on drinking water for Indonesia’s New Capital City (IKN) can be mapped under the conceptual framework of Input–Throughput–Output of the drinking water supply system (Figure 1) [1]. The framework of the drinking water supply system employs a systemic lens to understand water provision, linking upstream raw water sources (input), technological and institutional processes (throughput), and service delivery to communities (output).

At the input stage, studies examined raw water availability and long-term sustainability forecasting [2], environmental carrying capacity based on water balance, and the quality of raw water. At the throughput stage, research addressed technological, infrastructural, and sustainable distribution design integrating hydraulic efficiency and cost considerations [2, 3] and the comparative evaluation of Multi-Utility Tunnel versus conventional methods for IKN infrastructure. At the output stage, the framework emphasizes drinkable water provision coverage as a stochastic outcome, highly dependent on (1) competitive tariff setting, (2) subsidy mechanisms, (3) demand rate, and (4) the drinking water condition in terms of quality, quantity, and continuity. While institutional and governance challenges were analysed through a comparative study of Jakarta and Kuala Lumpur, culminating in a proposed institutional model for IKN’s piped water system [4].

The trajectory research from 2023 to 2025 demonstrates a coherent and evolving research sequence, progressively covering input–throughput–output dimensions of sustainable water provision for IKN. Collectively, this sequence demonstrates a holistic trajectory from assessing raw water sources to designing efficient distribution systems and ultimately to ensuring equitable and sustainable service delivery. The integration of hydrological, engineering, and financial perspectives strengthens the applicability of the findings to real-world policy and infrastructure planning.

Figure 1. Input–Throughput–Output of drinking water supply system

Source: Studies [1, 4, 5]

A critical knowledge gap remains regarding variation in raw water quality and pollution load, which affects the overall cost and sustainability of drinking water provision. The primary research question guiding this study is: How does variation in raw water quality and pollution load along the Sepaku River influence water treatment complexity and production costs, and what integrated management strategies are required to ensure affordable and sustainable drinking water provision for Nusantara?

The objectives of this study are fourfold: (1) to evaluate the variation of raw water quality status of IKN water sources using physical, chemical, and biological parameters; (2) to quantify the pollution load dynamics between dry and rainy seasons; (3) to estimate the treatment costs associated with varying pollution levels, including the role of chemical consumption; and (4) to analyse the implications for sustainable water provision coverage, recognizing its stochastic dependence on tariff, subsidies, demand, and water condition (quality, quantity, continuity).

By addressing these objectives, this research extends the input–throughput–output trajectory and provides actionable insights for designing cost-effective and resilient water supply strategies in IKN, with broader implications for urban water management in tropical regions.

3.1 Results

3.1.1 General conditions

Flow rates at the sampling sites varied between 0.1 and 27 m³/s. Water samples were collected at depths adjusted to the river water surface, ranging from 0.5 to 3 meters, to represent the actual water column conditions.

3.1.2 Quality status of raw water sources

The results of field measurements for both the Sepaku Semoi Dam (Location A) (Table 5) and Sepaku River Intake (Location B) are presented in Table 6.

Table 5. Sepaku Semoi Dam (Location A)

Table 6. Intake of Sepaku River (Location B)

3.1.3 Water pollution index

Table 7 summarizes the pollution index (PI) values for each sampling station in February and December, including maximum and average $C_{i j}$/$L_{i j}$ ratios and the resulting water quality classifications based on Regulation No. 27/2021 (see Figure 5).

Table 7. Pollutant index (PI)

Figure 5. Pollution index (PI) of raw water sources

3.1.4 Water sources pollution load

Using Eq. (2), the results of the water pollution load calculation are presented in Table 7. Specifically, the results for physical parameters are shown in Table 8, for chemical parameters in Table 9, and for biological parameters in Table 10.

Table 8. Physical characteristics of raw water

Table 9. Chemical characteristics

Table 10. Biological characteristics of raw water

3.1.5 Chemical implications to the cost of production

Based on observations and data collection, the comparative costs across utilities are as follows: Danum Taka shows a chemical cost of IDR 422/m³ with a total production cost of IDR 1,447/m³, reflecting lower chemical expenses but limited by operational inefficiencies and high water losses. Tirta Mahakam has a chemical cost of IDR 386/m³ and a total cost of IDR 5,180/m³, where higher organic and microbial loads increase chlorine demand, and overhead costs dominate. For the IKN projection [1], the estimated chemical cost is IDR 404/m³ with a total production cost of IDR 3,313.5/m³ [1], benefiting from economies of scale and improved intake management, though still influenced by seasonal peaks in TDS and nitrate.

3.1.6 Simulation of cost sensitivity

Using the average values, a simulation was carried out assuming chemical costs vary proportionally with water quality, while energy, maintenance, and other overhead remain constant [17, 18]. The annual production of IKN Projection is equal to 157,680,000 m³/year (see Table 11).

Table 11. Chemical costs simulation

This simulation shows that a ±30% shift in chemical demand due to raw water quality fluctuations may alter total annual production costs by approximately ±IDR 19 billion, underscoring the strong sensitivity of chemical expenses to pollution loads.

3.2 Discussion

3.2.1 Physical parameters (temperature, TDS, TSS, colour, salinity)

The water temperature standard is about 26.3 - 33.1℃, obtained from a 3℃ deviation of air temperature (Minister of Health Regulation No. 2 of 2023 concerning Implementation of Government Regulation No. 66 of 2014 concerning Environmental Health). The increase in river temperature due to climate change has an impact on the surrounding ecosystem [19], especially the dissolved oxygen (DO) level, which is one of the important parameters that describes the condition of raw water pollution. Surface temperature will reduce the DO concentration in India by 2.3% [20]. In addition to temperature, DO is also influenced by atmospheric pressure, temperature, salinity, water turbulence, photosynthesis, respiration, and waste [21, 22].

Table 8 presents physical parameters. The TDS load showed a substantial increase at several locations during December. For instance, at S5, the value rose sharply from 239,846.40 in February to 2,622,160.51 in December. A similar trend was observed at S6 and S3, where TDS loads reached 2,350,322.44 and 366,048.46, respectively. This increase in TDS concentration is strongly influenced by intensified human activities near the sampling sites, particularly in densely populated areas with high domestic inputs [23].

TDS is often used in determining the quality of water for consumption [24] because it describes various dissolved solid materials, including organic compounds and colloids [25]. TDS needs to be considered in planning and development [26] to provide clean water to IKN. Both river flows show significantly different TDS values. Flow 2 tends to be more stable below the quality standard, unlike flow 1. The highest TDS values were found at S7 and S8 of 1,735.00 mg/l and 15,112.67 mg/l in February measurements. The high TDS value due to construction causes solids from the riverbed to be lifted to the surface. Although the value decreased in December alongside the completion of construction, Outlet City 1 (S9) still tended to experience an increase in TDS due to the shipping activity, loading, and unloading at the pier that affected the accumulation of mud from the riverbed [10].

The TSS load also displayed a rising trend during the rainy season (December). At S3, TSS reached 3,293,440.82, the highest among all stations. Elevated values were also recorded at S7 (754,790.40) and S8 (68,584.32). High TSS levels can cause turbidity, thereby reducing water quality. The increase is likely linked to ongoing construction activities around the river, which disturb sediments.

The TSS value also fluctuated during the measurement temperature, especially when entering the rainy season. It is aligned with research [27] that the TSS value is also affected by temperature, although with a small impact.

The colour parameters above the quality standards are in line with physical observations during measurements, where all sampling locations are brownish and turbid. Another parameter is salinity, which is below the quality standards during measurements in December. Salinity describes the condition of the river ecosystem and is needed to control components in the water body [28].

Salinity loads generally followed a similar seasonal pattern, increasing during December. At S5, salinity rose significantly from 178.33 ppt in February to 1,679.62 ppt in December. However, an opposite trend was observed at stations S6 to S9, where salinity decreased in December, indicating localized hydrological or land-use influences.

3.2.2 Chemical parameters (pH, BOD, COD, nitrate)

The chemical parameters revealed spatial and seasonal variability. pH remained relatively stable within the acceptable range of 6–9, though slightly more alkaline conditions were observed in December at S6, S7, and T1.

BOD values showed extreme spikes at S5 (187,837.06 in December) and S9 (37,091.52 in February), reflecting intense organic pollution from domestic wastewater inputs. Other points, such as S7 and T1, also recorded elevated BOD in December, indicating that runoff during the rainy season carried more biodegradable material into rivers [30].

Similarly, COD values peaked at S5 in December (638,813.95) and S7 in February (975,325.78), indicating heavy chemical and organic loads likely due to industrial and domestic waste discharges [31].

Nitrate concentrations were particularly high at S5 (68,584.32 in December) and S3 (65,658.47 in December) [32], far above quality standards. Both stations are near residential zones, where nitrate pollution is often linked to sewage leakage and fertilizer runoff. This pattern is consistent with research [32], which found that centralized residential areas contribute to high NO₃⁻ levels. Even with dilution during the rainy season, hotspots at S8 and S9 still exceeded thresholds, suggesting that rainfall was insufficient to reduce nitrate levels.

Table 9 presents the chemical characteristics of river water across the sampling stations in February and December. The pH values generally remained within the acceptable range of 6–9, showing relatively stable conditions across all stations. However, a slight increase was observed in December, particularly at S6, S7, and T1, indicating more alkaline conditions, possibly due to dilution effects from rainfall or reduced organic input.

The BOD showed strong spatial and seasonal variations. Extremely high BOD values were recorded at S5 (187,837.06 in December) and S9 (37,091.52 in February), reflecting intense organic pollution likely originating from domestic and settlement activities. Other stations, such as S7 and T1, also recorded notable increases in December, suggesting seasonal runoff transported more biodegradable organic matter into the river.

Similarly, the COD displayed critical spikes at several points, notably S5 in December (638,813.95) and S7 in February (975,325.78), indicating substantial chemical and organic pollutants in the water column. These exceptionally high COD loads point to heavy domestic and possibly industrial contributions, particularly at stations located near densely populated or high-activity areas.

The nitrate (NO₃⁻) concentrations also revealed elevated levels at multiple stations. For example, S5 (68,584.32 in December) and S3 (65,658.47 in December) exceeded typical environmental thresholds. These hotspots correspond to residential areas where nitrate pollution is often linked to sewage leakage, fertilizer runoff, and other anthropogenic discharges. By contrast, some stations (e.g., T2 in February and T3 in December) exhibited relatively low nitrate loads, consistent with less human activity in their surroundings.

Overall, the chemical indicators suggest that both organic and nutrient pollution are significant issues in the study area, with December generally showing higher values than February. This seasonal difference highlights the influence of rainy season runoff and land-based activities in amplifying the pollution load. The consistently high BOD, COD, and NO₃ values at certain stations underscore the need for targeted mitigation strategies, particularly in densely populated and high-activity zones.

3.2.3 Biological parameters (total coliform)

Total coliform was the only biological indicator measured. The highest value was observed at S3, with an increase from 483,840.00 MPN/100 mL in February to 14,929,920.00 MPN/100 mL in December. Elevated coliform counts were also recorded at S5 and S6, underscoring that untreated domestic waste significantly contaminates the river [9].

While some points showed reductions during the rainy season due to dilution, almost all locations remained well above the maximum standard, especially S3, S4, S6, and S7. The link between coliform levels and physical–chemical parameters is evident: optimal bacterial growth occurs at pH ~7 and temperature around 37℃ [10]. In addition, increases in BOD and COD promote bacterial proliferation, highlighting serious health risks from microbial pollution.

Table 10 presents the biological characteristics of river water, where total coliform was the only microbiological parameter analysed in this study. Total coliform is a widely recognized indicator of microbiological pollution and reflects contamination from both human and animal waste.

Total coliform was the only biological parameter measured in this study. The highest total coliform value was recorded in S3, reaching 14,929,920.00 MPN/100mL in December, a significant increase compared to February at 483,840.00 MPN/100mL. In addition, S5 and S6 also showed high values. This indicates that domestic waste has not been properly processed before entering the river. High coliform content is a significant indicator of water pollution in urban and densely populated areas [11].

Overall, pollution load values increased during the rainy season (December), particularly for TDS, TSS, and total coliform at stations S3, S5, S6, and S7. This seasonal rise reflects the combined effects of intensified runoff, domestic discharge, and construction-related disturbances, which amplify the pollutant load carried into the river. Chemical indicators such as BOD, COD, and nitrate further confirm the strong influence of anthropogenic inputs, especially in densely populated and high-activity areas. Researcher [32] also stated that river water quality tends to deteriorate in the rainy season due to increased water discharge carrying suspended particles and various other pollutants.

3.2.4 Integrated assessment: Pollution index and pollution load

The PI assessment classified water quality status across stations from lightly polluted to moderately polluted. February conditions generally showed lighter pollution, whereas December values tended to worsen, with several points (e.g., S3, S4, S7, S8, S9) falling into the moderately polluted category. These fluctuations underscore the influence of seasonal rainfall, runoff, and local land use.

Pollution Load analysis further quantifies pollutant volumes entering the river system. The combined use of PI (qualitative status) and Pollution Load (quantitative load) provides a robust framework to evaluate water quality deterioration, identify hotspots, and prioritize interventions. This dual approach ensures that both the status and magnitude of pollution are captured.

All sampling locations indicated that the river water was classified as lightly to moderately polluted based on the PI results. Flow 2 (T1–T4) tended to show more stable conditions compared to Flow 1 (S1–S9), which exhibited greater fluctuations across seasons. This suggests that Flow 2 is relatively safer to be considered as a raw water source for IKN. However, the increasing PI values observed at most Flow 2 stations also indicate that the pollution load is gradually rising over time. While several sampling points demonstrated improvements or maintained their water quality, others experienced worsening conditions. These dynamics are influenced by the river’s natural self-purification capacity and the role of meanders that enhance quality along the stream [10]. Therefore, integrated water quality management across both flows is crucial, particularly in the context of growing population pressures and the relocation of Indonesia’s capital to Nusantara (IKN)) [9]. Such management should include periodic monitoring of key water quality parameters [33], further investigations, and strategic planning for pollution control.

3.2.5 Implications of quality status on water treatment costs

The exceedance of water quality standards for parameters such as TDS, TSS, BOD, COD, nitrate, and coliform indicates that advanced treatment is required for raw water in Nusantara. Even sampling points classified as lightly polluted may necessitate additional treatment steps—such as coagulation, filtration, and disinfection—that directly increase operational costs. Seasonal peaks observed during December highlight the need for flexible treatment capacity to accommodate higher pollutant loads. Failure to adapt treatment to these fluctuations would not only compromise public health but also escalate long-term costs for utilities.

Water quality status has a direct influence on both the complexity and cost of treatment. The more pollutants present, the more advanced and costly the treatment process required. Parameters exceeding standards (e.g., TDS, TSS, COD, NO₃, and coliform) necessitate additional processes beyond conventional sedimentation and filtration. Technologies such as ozonation, activated carbon, and enhanced disinfection may be needed to achieve potable water quality.

Previous studies [4] emphasize the importance of selecting treatment technologies that balance effectiveness with affordability, ensuring access for the wider community. Treatment costs are highly sensitive to raw water chemistry because chemicals (coagulants, lime, chlorine) represent one of the largest variable cost components [4]. The quality of the raw water source affects the cost of water treatment due to the need for chemicals, including coagulants. Worse raw water source quality needs more chemicals and advanced processing technology.

Poorer water quality directly increases chemical dosage requirements, which in turn raises operating costs. For example, elevated TDS and TSS drive higher alum/lime demand, while BOD and COD require greater chlorine dosing. Excessive nitrate and coliform further amplify disinfection requirements.

Comparative analysis confirms this relationship:

1. Tirta Mahakam faces the highest chemical burden due to urban catchment inputs (COD, coliform), resulting in chemical costs of ~IDR 386/m³ within a total production cost of IDR 5,180/m³.
2. Danum Taka incurs slightly higher chemical costs (~IDR 422/m³) but lower total production costs (IDR 1,447/m³), though still vulnerable to seasonal TSS and BOD surges.
3. IKN projection benefits from economies of scale and improved intake location, with average chemical costs of ~IDR 404/m³ and total costs of ~IDR 3,313.5/m³. However, rainy season fluctuations in TDS, nitrate, and coliform remain a critical challenge.

The simulation further demonstrates that chemical costs alone can shift total production expenses by approximately ±IDR 19 billion/year under different raw water quality scenarios. This finding underscores that upstream pollution control—through wastewater management and catchment protection—is just as important as treatment plant optimization. Every reduction in pollutant load directly translates into lower chemical expenditure, greater financial sustainability, and more affordable tariffs for consumers in Nusantara.

Conceptualization and methodology: Nicco Plamonia (N.P.), Raissa Anjani (R.A.), and Khaerul Amru (K.A.); Investigation: Riardi Pratista Dewa (R.P.D.), R.A., and K.A.; Discussion and writing: R.A., K.A., Budi Kurniawan (B.K.), Ikhsan Budi Wahyono (I.B.W.), Muhammad Komarudin (M.K.), Mohammad Zaidan (M.Z.), Bambang Winarno (B.W.), Syaefudin (S.), and Wahyu Purwanta (W.P.); Editing and review: Teddy W. Sudinda (T.W.S.), Nicko Widiatmoko (N.W.), Hidir Tresnadi (H.T.), Tito Eko Parato (T.E.P.), and Muktiyono (M.). All authors have read and agreed to the published version of the manuscript.