Waste load allocation by integrated GMS modeling and economic evaluation for nitrate reduction in Varamin aquifer

Groundwater management with competing users is a challenge in arid and semi-arid areas (Darbandsari et al. 2020). Its quality is also vulnerable to wastewater and leaching from domestic, industrial, and agricultural activities (Wada et al. 2021). In particular, nutrient puts groundwater resources at risk and has secondary impacts on health and the economy (Kariman et al. 2018; Golian et al. 2021). Recent research indicates increased global concerns about nitrate leaching into the groundwater. Water extraction and wastewater discharges can affect its concentration in an aquifer (Shahraki et al. 2020). Water with excessive nitrate concentration can cause serious health problems for livestock and humans (Lorenz et al. 2014; Najafi Alamdarlo et al. 2016). Methemoglobinemia in infants and cancer risk are some examples (Gardner et al. 2020; Liu et al. 2022). Nitrate concentration in soil and aquifer increases in areas with a high level of agricultural activity using fertilizers and manure (Imani et al. 2015, 2017). Nitrogenous fertilizers are heavily invested in the fields by farmers to maintain adequate production and increase yields (Jamshidi et al. 2020; Jamshidi & Naderi 2023). The excess nitrate leaches into groundwater and causes contamination problems (Martínez et al. 2020). In contrast to natural attenuation processes (Akbarzadeh et al. 2015), engineered systems, such as a wastewater treatment plant (WWTP), can significantly reduce nitrate pollution (Jamshidi & Niksokhan 2016). Thus, recent research efforts have focused on understanding the pathways of nitrate (NO₃) generation and contamination in groundwater and developing eco-friendly approaches to remove NO₃ from groundwater (Huno et al. 2018). Marshall et al. (2019) examined the spatiotemporal variability of wastewater parameters, including nitrate and their transport in groundwater. They could estimate the impacts of WWTP on groundwater pollution transport over time (Marshall et al. 2019).

NO₃ is traceable and manageable through waste load allocation (WLA) policies (Menció et al. 2016; Alfarrah & Walraevens 2018). To predict NO₃ contamination in groundwater, a nitrate-based transport model was recently used to estimate four WLA scenarios. In every scenario, NO₃ concentration was influenced by discharged pollution loads. Its concentration decreased in all scenarios during the 20-year prediction period, while none of the scenarios could reduce it below the standard level (Karlović et al. 2022). Depending on the specific aquifer conditions, the types of contaminants, and related costs, each WLA scenario may control or reduce some pollution loads (Benjakul 2010; Matiatos et al. 2019). It is recommended to select the best-performing WLA systematically by evaluating its effectiveness or potential in water quality improvement (Samadi-Darafshani et al. 2021). Since pollution removal is set at a desired level in WLAs, the related cost of attaining that purity level increases as treatment improves (Jamshidi et al. 2014; Englande et al. 2015). Identifying appropriate strategies and financing for wastewater infrastructure under WLA is often challenging for decision-makers (van Afferden et al. 2015). Therefore, groundwater quality management by constructing WWTP requires a two-faceted economic-functional analysis. This should be implemented by an integrated simulation tool to estimate the fate and transport of pollutants and track the sensitivity of receiving aquifer to different emissions. For this purpose, numerical models are valuable tools in understanding groundwater systems and estimating their qualitative status under various WLA scenarios (Foster et al. 2021). Groundwater resources modeling poses challenges in understanding influential parameters and achieving proper model calibration, crucial for accurate results, and effective water resource management (Eini et al. 2020; Jafari et al. 2021). To solve WLA problems, it is necessary to understand the quality response of receiving water bodies to each scenario. Groundwater simulation is commonly carried out by two GMS software modules. The MODFLOW and MT3D modules are used for water quantity and quality modeling, respectively (Raetz 2022). For example, a new simulation-optimization method in a watershed has been devised to solve problems associated with multiple pollutants (Sadak et al. 2020). Multipollutant loads were allocated according to the locations of their sources using an optimization formula. However, groundwater systems may take a decade or more to respond to this strategy.

This study aims to compare the impacts of constructing two WWTPs (Pakdasht and Varamin) with different nitrate removal efficiencies on groundwater quality. This study uses aquifer flow simulation by MODFLOW, followed by contaminant fate and transport modeling by MT3D. These modules predict NO₃ concentration in groundwater in different WLA scenarios during 3-year (T3) and 10-year (T10) periods. Here, WWTPs have different nitrate removal efficiencies with total abatement costs. Thus, on the basis of a numerical model, this research evaluates the effectiveness of WWTPs in remediating groundwater quality and considers related costs. By these means, the optimal WLA strategy is recommended in the study area. Accordingly, this study has two notable innovations in the field. Firstly, the research employs a combined qualitative and quantitative modeling for WLA in groundwater. This approach ensures a holistic understanding of the nitrogen reduction in aquifers by different wastewater treatments. Secondly, by incorporating multiple scenarios, the study provides a cost-effective analysis that enhances our understanding of potential management strategies. The question ‘is it necessary to upgrade both WWTPs with tertiary units?’ is answered by introducing the marginal cost (MC), as a decisive index, in the applied methodology.

This study examines the impacts of WWTPs on the Varamin Plain, located in the south-eastern part of Tehran province, Iran. This plain is between the latitudes of 35.39°N and 35.07°N and the longitudes of 51.26°E and 51.55°E and covers approximately 957 km² area. Excessive water demands and unsustainable groundwater extraction in the Varamin aquifer have caused declining water levels, resulting in land subsidence and reduced groundwater table (Nayyeri et al. 2021). This plain supplies the agricultural and drinking water of two counties, Pakdasht and Varamin. Droughts in recent decades have made wastewater recharge an attractive option. It could partially compensate for water quantity for agriculture, whereas deteriorated water quality (Noghreyan et al. 2022). Here, untreated wastewater of Pakdasht and Varamin are the primary point sources (Ministry of power 2013; Nouri et al. 2020). It is planned to construct and operate two WWTPs for these two counties. WW1 locates in 35°24′0″N, 51°53′0″E (Varamin) and the WW2 blueprint is 35°28′0″N, 51°48′0″E (Pakdasht). WW1 and WW2 receive sewage from about 284,000 and 351,000 persons, respectively. Accordingly, this study seeks a cost-effective and environmentally sound WLA to answer the following question: which WWTP, WW1 or WW2, should be prioritized regarding groundwater quality variations and total costs?

GMS software (version 10.1) was used to simulate the Varamin aquifer. MODFLOW and MT3DMs were the modules that simulated groundwater level and quality, respectively. Since the study area is quite large, it has been divided into six zones based on land use and available observation wells having water quality data. This zoning could provide an elaborate framework for demonstrating the sensitivity of groundwater to the WLA scenario in short term (3 years) and long term (10 years). For simulation, the study area was gridded by 250,000 m² squares. The locations of observation wells, their temporal variations on water levels, and aquifer recharge were initially determined based on available data. Furthermore, qualitative information, such as NO₃ concentrations and discharges from emission sources, were assigned to the nodes (Ministry of Energy 2013).

The effectiveness of six WLA scenarios on aquifer quality was investigated and compared through the calibrated model. Table 2 illustrates the specifications of these WLA scenarios. Differences in WLAs are basically in the locations of WWTPs (WW1 and WW2), the zoning of treated wastewater recharge, and their NO₃ removal efficiency (25 and 50%). S0 depicts the basic status of the simulated aquifer in the study area without any projected WLA.

In each WLA scenario (S1–S6), the abated nitrogen (N) was assigned to the target residential region as equivalent to reduced NO₃ concentration (%). It is necessary as Pakdasht and Varamin counties have no WWTPs and the untreated wastewater is conventionally discharged to the aquifer. N mitigation in S1–S6 means that NO₃ concentration in treated wastewater is decreased via constructing and operating WWTPs prior to any discharge or infiltration to the groundwater. The outcomes of all WLA scenarios were then analyzed and evaluated in both the objective area as well as the study area observation wells.

It should be added that WLA scenarios were individually defined in the calibrated model and NO₃ concentration at eight observation wells was obtained by GMS outputs. The impact of each scenario on pollution abatement is estimated using the average T3 and T10 concentrations from all observation wells. Following that WLA scenarios were qualitatively ordered based on their effectiveness on groundwater quality. In this case, the most environmentally favorable WLA is the one with the lowest average NO₃ concentration in the entire plain based on NO₃ concentration at both T3 and T10. Nevertheless, a viable WLA scenario needs economic analysis as well.

In this research, we conducted a comprehensive assessment of various WLA scenarios and their implications for long-term groundwater quality and economic considerations. Our study was based on the NO₃ parameter recommended by Nzama et al. (2021), as it is considered a critical groundwater pollutant that can be managed through land-use changes and WWTPs (Nzama et al. 2021). The effectiveness of these practices was verified through simulations using MODFLOW and MT3DMs, recognized as efficient groundwater modeling techniques (Raetz 2022).

Our investigation on the Varamin Plain showed that the simultaneous implementation of two WWTPs presents relatively higher pollution abatement compared to WLA policies involving a single WWTP. However, their pollution abatement is not significant in groundwater in the short term, and therefore, implementing both WWTPs is not recommended considering the related costs. Interestingly, this finding aligns with the results of Wada et al. (2021), where a single WWTP had a more significant impact on groundwater quality. Our results also contradict the findings of Adebowale et al. (2019), with the notion that constructing multiple WWTPs yields greater efficiency, particularly in the areas surrounding the WWTPs, within a shorter time period (Adebowale et al. 2019).

In a recent study by Saadatpour et al. (2019), it was concluded that WWTPs can improve the water quality index, and increasing the investment cost leads to greater quality improvements. However, there is a specific limit beyond which the water quality index cannot be further improved, regardless of the cost incurred or equity sacrificed (Saadatpour et al. 2019). In our current study, we also found that increasing the cost of WWTPs results in greater quality improvements. However, once specific limits are reached, more investment would be futile.

It has also become evident that pollution concentration limits set by regulatory bodies must consider a combination of pollution objectives and stakeholder demands to ensure the development of user-friendly applications. By doing so, water quality violations can be minimized for up to 20 years (Dinar & Quinn 2022). Moreover, previous studies have highlighted the practicality of pollution removal in WWTPs (Jamshidi & Niksokhan 2016) within the context of WLA. These factors, coupled with our latest research findings, underscore the importance of integrating economic objectives and practical considerations to establish sustainable groundwater quality management strategies.

This study used GMS modules to simulate the impacts of 6 WLA scenarios on NO₃ reduction in two time periods (T3 and T10). Results showed that GMS is a reliable and practical tool for groundwater simulation prior to WLA analysis. The MODFLOW and MT3D modules can be used for quantity and quality modeling, respectively. According to the results, WWTPs in the study area would not be considerably effective on NO₃ abatement in less than 10 years. Here, constructing two WWTPs (WW1 and WW2) is the most leading WLA for groundwater quality enhancement. However, it requires considerable costs. In addition, improving aquifer quality is not linear to the number of constructed WWTPs. Thus, this study recommended calculating the marginal costs, as an efficiency index, for referring to the optimal WLA. By this approach, S3 (WW1 with 25%N removal), S1 (WW2 with 25%N removal), and S4 (WW1 with 50%N removal) are prioritized as cost-effective WLAs.

M.A.S. led the investigation, prepared the methodology, did software analysis, validated, analyzed, did data curation, visualized, and wrote the original draft. S.J. conceptualized the study, prepared the methodology, did analysis, supervised, did project administration, and wrote (reviewed and edited) the article. H.K.M. collected resources, did software analysis, validated, and visualized the study.

Iran Water Resource Management Co. under contract No. S.07.1401 with the University of Isfahan partially contributed to the funding of this research.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.