Water quality management by enhancing assimilation capacity with flow augmentation: a case study for the Yamuna River, Delhi

In the waterbodies, dissolved oxygen (DO) depletes due to organic pollutants from wastewater. DO and biochemical oxygen demand (BOD) indicate the presence of organic substances (Basant et al. 2010). The oxygen-demanding contaminants' natural purification depends on the required oxygen to be concentrated, the oxygen necessary to secure the ecosystem qualities fit for species with designated standards, and the aquatic system's BOD purifying extent, which is assessed by reduction and replenishment of oxygen (Chapra et al. 2021). Rivers have their assimilation capacity, and it is necessary to acquire knowledge about the disposed pollutant loads disposed from diverse sources that rivers could receive without retrogression of their indigenous state (Oliveira et al. 2012). Mathematical modeling can validate waste load in a water body by establishing the cause–effect relationship between contaminant load and water quality. Hence, assimilation capacity could be evaluated by several simulation models (González et al. 2014). These frameworks are used as decision-making tools for wastewater management policies (McIntyre & Wheater 2004). The simulation models correlate the water quality after disposing of wastewater into a water body (Cox 2003). Water quality models can also predict the reciprocation of the aquatic system with different scenarios. The modeling outcomes are effective managing tools for assisting the river quality administrator in evaluating realistic water body conservation strategies and aspects of pollutant loading uncertainty. The present study aspires to assess the assimilation capacity of a severely polluted river stretch, Yamuna River, Delhi. The model QUAl2Kw predicts the river quality and assesses the assimilating efficiency of the pollutant load. Kannel et al. (2007) appraised the river conditions of the Bagmati River, Nepal, with QUAL2Kw, and the framework represents the observed data and is highly sensitive to water depth. Neilson et al. (2013) studied nutrient criteria and waste load analysis using the model QUAL2Kw and set nutrient criteria for the rivers in Utah. Zare Farjoudi et al. (2021) worked on the Zarjub River, Iran, to reduce the cost of waste load treatment and determination of self-purification capacity and found this framework as a suitable tool. The QUAL2Kw framework was used for the Cetrima River, Portugal, enriched with nutrients. It was observed that the framework predicted the river quality parameter with limited data availability and evaluated the waterbody's conditions in modifications of the states (Oliveira et al. 2012). QUAL2Kw was used to simulate the load-carrying ability of the Kali Surabaya River, and the pollutant load for BOD and chemical oxygen demand (COD) was larger than the purification capacity of the river (Aliffia & Karnaningroem 2019). The seasonal variation of assimilation capacity for the Karun River, Tehran, was determined using this model and showed that the different scenarios adopted for modeling, which were reducing wastewater flow, wastewater concentrations, and increasing the flow, enhanced the river characteristics (Moghimi Nezad et al. 2018). Although there are several models to predict the pollutants' fate, due to easy accessibility, the ability to simulate maximum contaminants, and the availability of uncertainty analysis, QUAL2Kw is the most suitable tool for the calculation of the load-carrying ability of a water body (Darji et al. 2022). Hence, this study has used this framework to predict the assimilative capacity of the Yamuna River, Delhi, and water quality management of the severely polluted stretch of Yamuna, Delhi, by flow augmentation and reducing the BOD and COD pollutant load. The urban reach of Delhi is one of the most contaminated river stretches in India, and Delhi contributes 79% of the pollutant load (Joshi et al. 2022). This segment carries wastewater from different industries and municipal sewerage in Delhi (Parmar & Singh 2015). This segment of the Yamuna River is polluted from 22 outfalling drains within the 22-km region from Wazirabad to Okhla (CPCB 2006). Before entering Delhi, the river shows medium water quality. After entering Delhi, due to discharging a massive BOD load and lack of fresh water, the river segment becomes a sewerage line (Upadhyay et al. 2011). Hence, it is crucial to maintain the ecological health of this reach. The wastewater with little or no treatment has deteriorated the river reach (CPCB 2006). Uninterrupted wastewater input with excessive organic pollutants from different sources decreases river quality. The DO concentration becomes low when the river maintains low flow and receives huge wastewater flow (Gain & Giupponi 2015). Hence, flow augmentation is required to manage the water quality and increase the DO of such a polluted reach. The DO concentration of this river reach shows a sharp decline to zero or is undetectable after discharging wastewater from the Nazafgarh drain, which is the prime contributor to waste load (Parmar & Singh 2015). Several studies appraised this segment's river quality (Kumar et al. 2019). A little work has been done on the pollution-carrying capacity of this reach. CPCB (1982) did the assimilation capacity of the Delhi reach of Yamuna River for COD and chloride, and four major contributing drains were considered. However, the study was only related to the discharging pollutant load from the point sources. Parmar & Keshari (2014) used QUAL2E to study the waste load allocation for this stretch and recommended that flow augmentation was unsuitable for this reach. However, these studies did not include assessing the assimilation capacity by varying BOD and COD with the model QUAL2Kw and improving the load-carrying capacity with flow augmentation. Verma et al. (2022) suggested that this river reach required a combination of management options including load reduction, flow augmentation, and external aeration. The present study aims to understand the requirement of flow augmentation to enhance the assimilation capacity of the Yamuna River, Delhi, with suggested effluents standards and hence to manage the desired water quality of the river.

The Yamuna River flows from Yamunotri, India, and is the lengthiest tributary to the Ganga River adjoining Prayagraj after traversing 1,376 km from the origination. Before getting into Delhi at Palla, the river crosses 348 km, and the length of Delhi's reach is 48 km. At Palla, the river water shows desirable water quality conditions with low BOD and sufficient DO concentration (Joshi et al. 2022). After reaching Wazirabad, around 23 km from upstream (Joshi et al. 2022), most indigenous water withdraws and supplies to Delhi, and the perennial river contains little or no fresh water. From the Wazirabad barrage to Okhla upstream, the river feeds 16 main drains containing around 3,000 megaliters per day of wastewater with 265 tonnes per day of BOD load (Delhi Pollution Control Committee (DPCC) 2020). National Green Tribunal (NGT, 2014) also reported that the national capital of India leads to pollution of the Yamuna River Delhi stretch by drains containing domestic and industrial sewage. Due to the disposal of untreated and partially treated wastewater from different sewage treatment plants through these drains, the river stretch becomes mostly anoxic after the Wazirabad barrage. It contains high oxygen-demanding substances, microorganisms, and nutrients. The study includes Delhi's 22-km urban river reach between the Wazirabad barrage and the upstream of Okhla barrage and 16 main outfalling drains between these distances. The climatic condition of this area varies between hot in summer and cold in winter. The average summer temperature is 32 °C, with a maximum temperature of 45 °C. At the same time, the average temperature in winter is 12–13 °C, and the lowest temperature is around 2 °C (Arora & Keshari 2021). The monsoon period starts from late June to September, and the highest average rainfall was approximately 515 mm in August (Joshi et al. 2022). During this time, wastewater dilutes with rainwater and improves river quality. Hence, variation in water quality was observed during the monsoon period. Figure 1 shows the Yamuna River, Delhi, with outfalling drains.

The study used data from the DPCC, which is responsible for collecting data for the Delhi reach and all the drains outfalling between the distance. The monitoring stations included in this study covered five stations DPCC S1 (Wazirabad downstream), S2 (Inter-State Bus Terminal), S3 (Income Tax Office ), S4 (Nizamuddin bridge), and S5 (Okhla Upstream). The coordinates of these stations are shown in Table 1. The DO, BOD, COD, and pH water quality data were collected for March 2021 and April 2022. For the Delhi region, March–May is the low flow period due to negligible rainfall, known as the pre-monsoon period. The 16 outfalling drains were taken as point sources, and data were collected from DPCC. Due to data constraints, only four parameters were collected and simulated. The model QUAL2Kw was selected for this study, and the average data of March 2021 were used for calibration and those of April 2022 were used for confirmation.

QUAl2Kw model was used in this study to simulate BOD, COD, DO, and pH. This framework divided the reach into unequal, properly mixed segments of the same hydrological and water quality conditions (Kang et al. 2020). The 22-km river reach was divided into 14 segments depending on the confluence of drains as point sources. Due to the instability of the model, readings from drains outfalling in a minimal distance were taken in one segment. The model capabilities and descriptions are found in the user manual of QUAL2Kw. The framework is suitable for the river to reach with more or less constant pollutant loads and flow (Oliveira et al. 2012). The input data comprised headwater flow and water quality data, 16 outfalling drains wastewater flow, and quality data as point sources. These point sources carry domestic and industrial wastewater and discharge from sewage treatment plants. Delhi receives deficient rainfall; hence, surface runoff is very low. Besides this, some diffused sources of pollutant loads are used for cattle bathing, washing clothes, and bathing people. The model was calibrated using low flow and dry period data for March 2021. Geometrics and hydraulics data are shown in supplementary Table S1. Calibration was done repeatedly until the predicted values came closer to actual conditions. The Manning constant and slope of the river were taken as 0.05 and 0.0002, respectively (Singh & Ghosh, 2001). The Manning equation was used because of limited data availability. The BOD values were taken as fast carbonaeous biochemical demand and COD as generic constituents. The constituents included in the model are flow, temperature, BOD, DO, COD, pH, conductivity, and alkalinity. Due to data constraints, nutrient data were not included in this study. For the slow-moving river with shallow depth, the O'Conner–Dobbins equation was used to calculate reaeration constants (Paliwal et al. 2007). The BOD and DO deal with the mechanism of sedimentation and settling, but due to low oxygen availability, 25% settleable BOD (CPCB, 82) decomposes in anoxic conditions and hence no trade of DO (Paliwal et al. 2007). Again, product methane rises upward, and due to buoyant forces, settled substances resuspend (Kazmi & Hansen 1997). Due to high turbidity, sunlight is obstructed, and hence, phytoplankton activities are negligible. The DO variation due to photosynthesis and respiration is insignificant for this reach (Parmar & Keshari 2014). An exponential model was chosen for oxygen inhabitation for carbonaceous biochemical demand, and the calculation step was set to 5.625 for model stabilization. Except for the monsoon period, flow conditions are almost the same throughout the year (CPCB 2006). The non-monsoon flow prevails most of the year, and critical flow is essential to determine the assimilation capacity. For calibration and validation, 1 m³/s flow is assumed at the upstream point. The headwater flow and quality are shown in supplementary Table S2. The point source input values are shown in supplementary Table S3. The Yamuna River, Delhi, is polluted with loads from different nonpoint sources. Although groundwater recharge is negligible for this area, pollution from nearby slum areas, cattle bathing, and agricultural runoff should be considered (Kazmi & Hansen 1997). In this study, 1 mg/l of distributed BOD load was adjusted after a 5 km distance. The model was run until simulated values agreed with observed values. The framework was auto-calibrated for a population size of 100 and 50 generations, and simulation was done with new datasets for April for confirmation. The root mean square error was calculated to verify the calibration result with validation results.

The QUAL2Kw was applied to assess the assimilation capacity of this polluted stretch. Hence, four cases were studied, generating 41 scenarios varying the BOD and COD load with flow augmentation. Supplementary Table S4 shows input BOD and COD loads with point source flow. The head water flow was increased at 10 cumecs intervals for different scenarios. The scenarios were generated to achieve the water quality suggested for this river stretch, i.e., Class C by the CPCB. For this criterion, river water should maintain DO greater than 4 mg/l and BOD less than 3 mg/l. The upstream flow was increased to maintain this requirement by adjusting different BOD and COD loads. Flow augmentation of 10 cumecs increment was done upstream for developing scenarios of four cases shown in supplementary Figure S1.

Figure 4 shows that with existing flow and pollutant load, DO concentration decreased to 0 throughout the river reach, and BOD concentration was also above 60 mg/l after outfalling of D1. Hence, pollutant load is also required to be reduced with flow augmentation at the upstream.

In strategy 3, the flow was increased from 10 cumecs up to 120 cumecs for scenarios s1–s12 of the four cases, and BOD and COD loads kept changing, as shown in Table S4. In case 1, 12 scenarios were generated, increasing flow from 10 to 120 cumecs. BOD and COD were kept at 10 and 50 mg/l, respectively, for all point sources, as effluent standards were set for this river stretch by the National Green Tribunal (NGT) of India.

The QUAL2Kw model assessed the assimilation capacity of Yamuna's most polluted stretch. The model is appropriate for this reach as it can be simulated with low data availability. Thus, it is ideal for decision-making tools like India, where limited data are available. This study revealed that the river's assimilation capacity was low due to high BOD and low DO levels. The wastewater enters the river from 16 drains and also diffused sources. Najafgarh drains added the highest wastewater quantity with elevated BOD and COD levels; thus, after adjoining this drain, the river's water quality fell to inferior. These conditions prevailed over the 22 km of this reach. In this study, the upstream flow increment with a reduction of BOD and COD was studied. In strategy 1, wastewater from all drains was curtailed, and the desired standard of reach was not found. Improvement of the assimilation capacity of this river is a very challenging job, as there is less upstream water with low DO and high BOD. In strategy 3, four cases were established with 41 scenarios with an increment of flow and reductions of BOD and COD. These cases suggest that load reduction and flow increment can improve the assimilation capacity of the river reach. This reach required substantial load cutting with flow dilution to enhance the water quality. Both the remedy options are very complicated and economically unfeasible. The study also revealed that COD and BOD are responsible for DO deterioration. It is also noted that the nitrogenous substances would improve the estimation of DO. As these drains carry domestic water containing nitrogenous waste, it has been suggested that regular monitoring of ammonium, organic nitrogen, and nitrate nitrite is also required.

All relevant data are available from an online repository or repositories. https://www.dpcc.delhigovt.nic.in/home/monthly_analysis_report.

The authors declare there is no conflict.