Next Article in Journal
Human Health Risk Assessment of Phenolic Contaminants in Lake Xingkai, China
Previous Article in Journal
Disentangling the Source of Uncertainty in Monthly Streamflow Predictions: A Case Study of Riu Mannu di Narcao Basin, Sardinia Region, Italy
Previous Article in Special Issue
Seasonal Distribution of Microbial Community and n-Alkane Functional Genes in Diesel-Contaminated Groundwater: Influence of Water Table Fluctuation
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Evaluating Natural Attenuation of Dissolved Volatile Organic Compounds in Shallow Aquifer in Industrial Complex Using Numerical Models

by
Muhammad Shoaib Qamar
1,
Nipada Santha
2,
Sutthipong Taweelarp
3,
Nattapol Ploymaklam
4,5,
Morrakot Khebchareon
4,5,
Muhammad Zakir Afridi
1 and
Schradh Saenton
1,2,5,*
1
M.S. Program in Environmental Science (CMU Presidential Scholarship), Environmental Science Research Center, Faculty of Science, Chiang Mai University, 239 Huaykaew Rd., Chiang Mai 50200, Thailand
2
Department of Geological Sciences, Faculty of Science, Chiang Mai University, Chiang Mai 50200, Thailand
3
Department of Geotechnology, Faculty of Technology, Khon Kaen University, Khon Kaen 40002, Thailand
4
Department of Mathematics, Faculty of Science, Chiang Mai University, Chiang Mai 50200, Thailand
5
Advanced Research Center for Computational Simulation (ARCCoS), Chiang Mai University, Chiang Mai 50200, Thailand
*
Author to whom correspondence should be addressed.
Water 2025, 17(13), 2038; https://doi.org/10.3390/w17132038
Submission received: 22 April 2025 / Revised: 26 June 2025 / Accepted: 3 July 2025 / Published: 7 July 2025
(This article belongs to the Special Issue Application of Bioremediation in Groundwater and Soil Pollution)

Abstract

A VOC-contaminated shallow aquifer in an industrial site was investigated to evaluate its potential for natural attenuation. The shallow groundwater aquifer beneath the industrial site has been contaminated by dissolved volatile organic compounds (VOCs) such as trichloroethylene (TCE), cis-1,2-dichloroethylene (cis-DCE), and vinyl chloride (VC) for more than three decades. Monitoring and investigation were implemented during 2011–2024, aiming to propose future groundwater aquifer management strategies. This study included groundwater borehole investigation, well installation monitoring, hydraulic head measurements, slug tests, groundwater samplings, and microbial analyses. Microbial investigations identified the predominant group of microorganisms of Proteobacteria, indicating biodegradation potential, as demonstrated by the presence of cis-DCE and VC. BIOSCREEN was used to evaluate the process of natural attenuation, incorporating site-specific parameters. A two-layer groundwater flow model was developed using MODFLOW with hydraulic conductivities obtained from slug tests. The site has an average hydraulic head of 259.6 m amsl with a hydraulic gradient of 0.026, resulting in an average groundwater flow velocity of 11 m/y. Hydraulic conductivities were estimated during model calibration using the PEST pilot point technique. A reactive transport model, RT3D, was used to simulate dissolved TCE transport over 30 years, which can undergo sorption as well as biodegradation. Model calibration demonstrated a satisfactory fit between observed and simulated groundwater heads with a root mean square error of 0.08 m and a correlation coefficient (r) between measured and simulated heads of 0.81, confirming the validity of the hydraulic conductivity distribution. The TCE plume continuously degraded and gradually migrated southward, generating a cis-DCE plume. The concentrations in both plumes decreased toward the end of the simulation period at Source 1 (located upstream), while BIOSCREEN results confirmed ongoing natural attenuation primarily by biodegradation. The integrated MODFLOW-RT3D-BIOSCREEN approach effectively evaluated VOC attenuation and plume migration. However, future remediation strategies should consider enhanced bioremediation to accelerate contaminant degradation at Source 2 and ensure long-term groundwater quality.

1. Introduction

Groundwater is an essential source of fresh water, crucial for sustaining household activities, industrial and agricultural operations, and ecological systems [1]. Globally, more than 1.5 billion people depend on groundwater as their main source for drinking, irrigation, and industrial purposes [2]. In addition to the amount of water needed, the quality of water is extremely important, especially for drinking water [3]. Contamination of groundwater from a range of pollutants, whether occurring naturally or resulting from human activities, has heightened concern about potential future water scarcity [4].
To improve the production processes, industries have exacerbated contamination by releasing numerous chemicals and organic compounds into the environment, and spills or inadequate disposal can lead to infiltration into the soil and groundwater. Volatile organic compounds (VOCs) in groundwater are a major environmental concern, with intervention essential to address with the risk of pervasive a VOC plume caused by anthropogenic activities [5]. Volatile organic compounds (VOCs), widely recognized for their toxicity and carcinogenic properties, pose significant risks to both the environment and human health [6]. Their prevalent production, usage, and inadequate disposal practices have led to their frequent presence as contaminants in soil and groundwater [7]. Once absorbed into the subsurface, dissolved VOCs interact with soil and sediment through various sorption processes. The most significant mechanism is hydrophobic partitioning into soil organic matter, which is determined by the organic carbon partition coefficient (Koc) and is strongly linked to the compound’s octanol–water partition coefficient (Kow). Soil with a higher organic content exhibits a greater adsorption capacity, which reduces the migration of VOCs in groundwater. Additionally, adsorption onto mineral surfaces (e.g., clays and oxides) and under varying redox conditions can also influence the VOC fate, though to a lesser extent.
Addressing the contamination of volatile organic compounds, such as trichloroethylene, cis-1,2-dichloroethylene, and vinyl chloride, in shallow groundwater aquifers presents considerable challenges, as traditional remediation methods are costly and complex. Natural attenuation is recognized as an effective remediation technique in various environmental remediation strategies [8,9]. Monitored natural attenuation (MNA) relies on natural processes such as biodegradation, dispersion, advection, sorption, and dilution to minimize the contaminant level in groundwater. In practical terms, natural attenuation may require months or even decades to completely purify a contaminated aquifer depending on the concentration of the contaminants. In industrial areas, regulatory frameworks, public health awareness, and technological resources enable more proactive and science-based remediation efforts, often including advanced monitoring and modeling tools. In contrast, developing countries may face constraints such as a limited institutional capacity, inadequate environmental regulations, and competing economic priorities, which delay or even prevent the remediation of contaminated aquifers. Moreover, geopolitical issues such as transboundary aquifer management or water conflicts can hinder collaborative efforts in data sharing and policy enforcement. As a result, the effectiveness of natural attenuation strategies and engineered remediation approaches may vary significantly across regions, emphasizing the need to integrate hydrogeological assessments with socio-political realities when developing groundwater protection policies.
Numerical groundwater and transport models have proven to be an effective tool to address a variety of groundwater-related problems. Simulation of groundwater is divided into flow modeling and contaminant transport modeling. The most commonly used groundwater flow model is MODFLOW, which solves the governing equation of groundwater flow through the finite difference method [10]. In the past three decades, reactive transport simulations have drawn enormous amounts of attention in recognizing the behavior of transport and transformation of chlorinated ethenes in aquifers under microbial processes [11]. As an extension of MODFLOW, RT3D incorporates chemical reactions and processes into the simulation, which is essential for understanding the degradation of contaminants, transformations, or interactions with the environment. One of the simpler models commonly used for natural attenuation is BIOSCREEN [12], which can simulate solute transport and degradation processes, incorporating advection, dispersion, and biodegradation mechanisms to predict contaminant attenuation over time and space, respectively. Although comprehensive numerical modeling tools have been utilized in previous studies to simulate key subsurface processes such as advection–dispersion, adsorption, and biodegradation, the use of BIOSCREEN is justified as a complementary tool to support preliminary screening-level analysis and model validation. BIOSCREEN provides a simplified approach to contaminant transport in groundwater along the flow path under a set of well-defined assumptions. By utilizing first-order decay and instantaneous reaction models, BIOSCREEN provides a robust framework for assessing the natural attenuation potential of both dissolved hydrocarbons and chlorinated solvents, respectively. This tool is particularly useful for sensitivity analysis and for comparing the output of more complex numerical simulations.
Despite extensive research on natural attenuation of chlorinated ethenes, a significant gap remains in understanding the long-term behavior of these compounds within groundwater aquifers exposed to persistent contamination over multiple decades. Previous studies often rely on short-term laboratory experiments, like batch and column tests, which fall short of capturing the complexity of in situ hydrogeochemical reactions at the field scale. Additionally, the relationship between site-specific microbial populations and VOC degradation pathways remains underexplored, despite the widespread use of numerical models for groundwater systems. In particular, the function of metabolic processes is in varying redox conditions and the transition from aerobic to anaerobic zones during prolonged attenuation. Thus, this study bridges the gap by merging real-world aquifer data, microbial interactions, predictive modeling, future contamination levels, and plume movement. This approach will offer novel perspectives on the viability and efficacy of natural attenuation of a VOC-contaminated shallow aquifer remediation strategy, as well as the potential for enhanced bioremediation in industrially impacted aquifers.
This study aims to assess the natural attenuation processes of volatile organic compounds within a shallow groundwater aquifer contaminated for over 30 years by industrial activities, focusing on the transport patterns, degradation rates over time, and future behavior of the contamination of plume using numerical models. The industrial site affected by the long-term discharge of chlorinated ethenes provides an opportunity to evaluate real-world hydrogeochemical changes under natural conditions. This study ensures a more accurate understanding of the processes occurring at the field scale by eliminating errors associated with laboratory-scale analyses and by applying models.

2. Materials and Methods

2.1. Site Description and History

An industrial site in Thailand has been affected by a historical TCE leakage incident that occurred over 30 years ago. The improper storage and disposal practices resulted in the percolation of TCE and its degradation by-products such as cis-1,2-Dichloroethylene (cis-DCE) and vinyl chloride (VC) into the shallow aquifer. This contamination has expanded over the past few years, creating a plume that poses a threat to surrounding water consumers and ecosystems. An aerial digitized map of the industrial area is shown in Figure S1.
TCE, cis-DCE, and VC in shallow aquifers were monitored in the study area from 2011 to 2016 and, again, in 2022, demonstrating temporal and spatial variations in contaminant distribution using a cut-off value that exceeded the maximum contaminant level, as shown in Figure 1, Figure 2 and Figure 3. The TCE concentration ranged from 0 to 1.4 mg/L, cis-DCE from 0 to 2.7 mg/L, and vinyl chloride from 0 to 0.02 mg/L. The maximum contaminant levels (MCLs) of TCE, cis-DCE, and VC are 0.005, 0.07, and 0.002 mg/L, respectively [13].
The concentration of VOCs in the industrial area is exceeding the maximum contaminant level and poses a significant risk of cancer. The estimated cancer prevalence for TCE at its maximum concentration of 1.4 mg/L is 0.2337 (23.37%) from a lifetime exposure, and a concentration of 2.7 mg/L of cis-DCE is associated with a 0.4326 (43.26%) cancer risk [14]. These high risks emphasize the urgent need for remedial measures to address these contaminants in the shallow aquifer, as they pose significant public health threats.

2.2. Soil Sampling

The soil samples were collected from two boreholes with depths of 15 and 25 m to assess the physical and chemical properties of the soil layers, as shown in Figure 4. A chemical analysis of soil was performed to evaluate organic matter and total organic carbon, as evidence of potential microbial activity and ability to retain contaminants. Soil samples were analyzed to determine the moisture content, grain size distribution, specific gravity, and porosity using standard geotechnical methods. Table 1 provides a list of methods for studying chemical and physical properties of the soil.

2.3. Groundwater Sampling and Chemical Analysis

Groundwater samples were collected quarterly, spanning the period from 2011 to 2016 and 2022, from 10 groundwater wells using low-flow pumps. The locations of groundwater sampling wells and groundwater flow direction are shown in Figure 4. Water levels in the existing groundwater wells were measured once a month using a water measuring tape device. The chemical analysis of collected groundwater samples for volatile organic compounds (TCE, DCE, and VC) was conducted using gas chromatography coupled with mass spectrometry (GC-MS) to ensure a high sensitivity and precision. In addition to VOC analysis, anions such as nitrate, phosphate, sulfate, and chloride were analyzed using ion chromatography (IC). The concentration of metals such as manganese, iron, sodium, calcium, potassium, and magnesium was analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES). Finally, the alkalinity was measured using an on-site titration test kit, allowing for immediate evaluation.

2.4. Slug Tests

The slug test [18,19] was conducted in five of the wells (deep and shallow) in the study area to measure the permeability coefficient of the shallow aquifer by using the Hvorslev technique (Figure 4). Groundwater was pumped out from the shallow and deep wells, and the water level was measured every minute until the water level returned to the normal level before pumping. The test results were determined as hydraulic conductivity (K).
Hvorslev’s Equations (1) and (2) for a partially penetrating well not in contact with an impermeable boundary are as follows:
ln H 0 ln H = 2 K r L t r c 2 ln L / 2 r w e + 1 + L / 2 r w e 2
r w e = r w K z / K r
where H is displacement at time t, H 0 is initial displacement at t = 0, K r is radial (horizontal) hydraulic conductivity, K z is vertical hydraulic conductivity, L is screen length, r c is nominal casing radius, r w is well radius, t is elapsed time since the initiation of the test, and r w e is effective well radius or equivalent well radius.

2.5. Pail Test

In this study, two observation wells (OB-1 and OB-2) with a depth of 25 m were selected for the pail test due to the availability of a specific strain, fair permeability, and highest concentration of VOCs. During the pail test, groundwater (80–200 L) was pumped out from these observation wells into a container. Carbon sources, nutrients, and electron acceptors, with a ratio of 100:10:1, were then added to the water to enhance the biodegradation process. The solution was thoroughly mixed and circulated to ensure adequate oxygen and solution were injected slowly back into the aquifer. For the addition of nutrients, the C:N:P ratio was calculated as 200:16.9:3.4 g for glucose, urea, and di-ammonium phosphate (DAP). This bioremediation process underwent regular monitoring between 3 and 6 months while groundwater was sampled once a month.

2.6. BIOSCREEN

BIOSCREEN intends to simulate transport and degradation using analytical solutions of advection–dispersion, as shown in Equation (3). BIOSCREEN has been widely used in site investigations and remediation planning due to its reliability and ease of use, particularly for sites contaminated by petroleum spills and leaking underground storage tanks. BIOSCREEN utilizes the Domenico solution to model one-dimensional solute transport with biodegradation, enabling it to be effective in assessing groundwater contamination.
C x = C 0 2 e r f c x v x / R 2 D x t / R + exp v x x D x e r f c x + v x t / R 2 D x t / R exp λ x v x
where C(x) is contaminant concentration at distance x (mg/L), C 0 is initial contaminant concentration (mg/L), v x is groundwater velocity (m/d), D x is longitudinal dispersion coefficient (m), λ is a first-order decay coefficient (per year), R is retardation factor, and erfc (x) is complementary error function. A summary of the BIOSCREEN input parameters is provided in Table 2.
The contaminated zone was designed as a thickness of 25 m in the saturated zone, divided into five zones over widths of 13, 21, 16, 21, 21, and 13 m, with input concentrations ranging from 0.022 to 0.03 mg/L based on field data. The modeled area dimensions were constructed at 166 m length and 100 m width, with a 30-year simulation period to show the long-term natural attenuation potential. The estimated plume length was 140 m, and field data, such as concentrations at different distances from the source (0, 16, 33, 50, 66, 83, 100, 116, 133, 150, and 166 m), were used to validate the model. BIOSCREEN used centerline and array modes to generate contaminant concentration profiles along the flow centerline. The findings, including plume extent, concentration profiles, and attenuation rates, were compared to field data to determine the natural attenuation potential of the groundwater aquifer accurately.

2.7. Numerical Model Setup

2.7.1. Groundwater Flow Model

This section outlines the methods used to build a groundwater flow simulation model. A model helps to combine different data types, such as hydrogeological conditions and hydrologic stresses that influence groundwater systems. MODFLOW [10,18,20] was used in this study. A steady-state groundwater flow was assumed to observe long-term average hydrogeological conditions in the study area. This assumption is confirmed due to the relatively stable water table and absence of significant seasonal or transient pumping influences during the modeled period. Steady-state conditions allow for a simplified simulation of groundwater flow while maintaining sufficient accuracy in assessing contaminant transport behavior over an extended time scale. The groundwater flow equation in which MODFLOW estimates the hydraulic head, h, is shown in Equation (4). The general governing equation for three-dimensional groundwater flow under steady-state conditions is given by
x K x h x   +   y K y h y   +   z K z h z     ±     W   =     0
where h is hydraulic head (m), Kx, Ky, Kz is hydraulic conductivity in the x, y, and z directions (m/d), and W is volumetric flux per unit volume representing sources and sinks.

2.7.2. Conceptual Model

The model conceptualization incorporates key parameters such as the measured groundwater heads, hydraulic conductivity, and well depth values derived from field measurements. The northern and southern boundaries of the model domain were replaced with specified head boundaries to indicate lateral inflow and outflow of groundwater. Boundary conditions were assigned based on observed hydraulic head data and hydrogeological characteristics of the industrial area. A specified head of 259.0 m was applied on the upgradient, representing the horizontal inflow zone, while a head of 258.3 m was assigned to the downgradient to simulate discharge conditions. Furthermore, a new coverage setup was added to the model named a groundwater well. The model utilized accurate data, such as depth, coordinates, and measured hydraulic head, to enhance accuracy and ensure a more accurate representation of groundwater flow. The K values from the slug test were interpolated into a finite-difference grid, which was used as an initial parameter value for subsequent simulation.

2.7.3. Model Setup

In the numerical model, the model domain was arranged in a 3D grid with 40 rows and 40 columns in the x and y directions (a uniform model grid size of approximately 10 × 10 m2), with two layers in the z direction. This resolution was chosen as a compromise between computational efficiency, inverse modeling run-time, and solution accuracy, based on preliminary simulations. It was found that repeated simulations with finer resolutions (e.g., 60 × 60 and 80 × 80 grids) did not significantly impact the results compared in terms of plume width, peak concentration, and gradient steepness. The results remain consistent across the grid, which indicates that the 40 × 40 resolution is sufficient.
The model included two layers to capture the vertical stratigraphy of the shallow aquifer. The first layer represented the upper portion of the aquifer, with a depth of 0 to 15 m, a top elevation of 259.6 m, and a bottom elevation of 244.6 m. The second layer represented the lower portion, with a depth of 15.5 to 25 m, a top elevation of 244.6 m, and a bottom elevation of 234.6 m. According to the field data, initial hydraulic conductivity values were set at 0.03 m/d for the first layer and 0.06 m/d for the second layer. The vertical anisotropy (Kv/Kh) for both layers was originally set as 0.1 and will be adjusted during the automatic model calibration process. This will ensure that our model simulation accounts for vertical heterogeneity. Steady-state simulations were conducted to determine groundwater flow patterns under constant conditions. A summary of the model discretization and some model parameters are shown in Table 3.

2.7.4. Model Calibration

Model calibration is the process of systematically adjusting the model parameters within limits to determine the optimal fit between hydraulic heads and flows. A highly parameterized parameter estimation, the PEST [21] tool, was used for model calibration to enhance model reliability and accuracy. PEST with the pilot point technique allows the model to simulate groundwater flow in aquifers with spatially variable hydraulic conductivity. A total of 100 pilot points were inputted to the model, as shown in Figure S2, and PEST was used to estimate the values at each point. The hydraulic conductivity from the slug test remains unchanged when using the fixed pilot point technique. An interpolating technique that facilitates model heterogeneity will then be applied to spatially interpolate these pilot point values in the model. The vertical anisotropy (Kv/Kh) of the model for both layers was also calibrated using PEST to allow a vertical heterogeneity of hydraulic conductivity.

2.7.5. Contaminant Transport Model Simulation

The MODFLOW package utilized RT3D [22] to simulate the transport of VOCs in the aquifer. The model was developed to simulate both advective and dispersive transport processes as well as the chemical reactions that lead to the degradation of contaminants. The advection–dispersion used in RT3D for contaminant transport in groundwater is shown in Equation (5).
C t = x D x C x + y D y C y + z D z C z x ( v x C ) y ( v y C ) z ( v z C ) + R ,
where C is contaminant concentration (mg/L), t is time (days), D x , D y , and D z are hydrodynamic dispersion coefficients in x, y, and z directions (m/d), v x , v y , and v z are seepage velocities in x, y, and z directions (m/d), and R is source/sink term (including reactions such as degradation, adsorption, etc.).
Initially, the dispersion package used a longitudinal dispersivity of 30 m (manually calibrated) along the flow path, the ratio of transverse horizontal dispersivity to longitudinal dispersivity (or α T , H / α L ) of 0.1, and the ratio of vertical horizontal dispersivity to longitudinal dispersivity (or α T , V / α L ) of 0.01. These dispersivity values are essential for determining the spatial and temporal dynamics of the contaminant plume and its transport in a more realistic way.
A constant concentration boundary condition was applied over a localized area at the center of the model domain to represent a continuous point source, consistent with a leaking above-ground storage tank scenario. The contaminant source was assumed to remain active over a defined period, adequate to simulate the change and migration of the plume. This approach is widely employed in groundwater modeling when the historical source strength and timing are not known. The spatial extent of the source was determined based on observed VOC concentrations in nearby wells, and the duration of the source activity was altered during model calibration to match the observed plume length and concentration distribution. This parameterization provides a precise estimation of source behavior while enabling an effective evaluation of natural attenuation processes.
In the simulation, there are two source zones of TCE that contribute to the contamination. The source and sink section of the model was characterized by a constant concentration of TCE, which was initially set to 2 mg/L at source 1 and 4 mg/L at source 2, corresponding to the expected concentration at both source zones. It should be noted that TCE concentrations in observation wells at both source zones have been monitored intermittently, with observed values varying over time and ranging from approximately 0.3 to 3.4 mg/L. Unfortunately, the available temporal concentration data were not sufficiently detailed to support the development of a time-varying source concentration profile for the transport simulation. As a result, we adopted a simplified approach in the RT3D model by applying constant source concentrations of 2 mg/L and 4 mg/L for Source #1 and Source #2, respectively. This assumption may lead to an underestimation or overestimation of the mass flux from the source zones at certain times, and we have noted this as a limitation of the current modeling approach.
The simulation targeted numerous results for a 10,950-day (30-year) period, but this simulation period is sufficient to observe the significant degradation process and dispersion of TCE under natural groundwater conditions. The Generalized Gear solver in the chemical reaction package effectively solves the kinetic reactions that govern the biodegradation of TCE and DCE, and it is widely used to simulate the decay of organic contaminants in groundwater modeling.
After the model configurations were completed, the RT3D package was executed to simulate the transport and degradation of TCE over the 30-year timeframe. The simulation depicted advection, dispersion, and biodegradation, providing a comprehensive overview of the contaminant’s behavior in the aquifer under natural conditions.

2.8. Assessment of the Natural Attenuation of the Site

Historical contaminant trends, microbial analysis, and geochemical indicators will be used to assess the degree of natural attenuation at the study site. When the contaminant concentrations exhibit a consistent decline over multiple sampling periods due to advection, dispersion, and biodegradation, it will highlight the significance of natural attenuation. To identify the favorable conditions for natural attenuation, the presence of electron acceptors and redox indicators will also be evaluated to determine the degradation process at the study site. The microbial community analysis will be conducted through 16S rRNA sequencing to identify known dechlorinating bacteria, such as Dehalococcoides and Geobacter. Once these microorganisms are in a sufficient quantity, natural attenuation can be considered an effective long-term approach. Furthermore, the first-order decay will be calculated from BIOSCREEN using the half-lives of TCE through manual calibration. The calculated decay rate will be compared to literature data if the observed attenuation is within the expected ranges for the natural degradation process.

3. Results

3.1. Soil Properties and Composition

The soil properties such as grain size distribution, moisture content, bulk density, porosity, specific gravity, soil organic matter (OM), and total organic carbon (TOC) showed the same results in both wells, as shown in Table 4.

3.2. Microbial Analysis and Pail Test

3.2.1. Microbial Analysis

A total of 216 genera were identified and further categorized into four main microbial groups analyzed from soil samples collected within the study area. Proteobacteria was identified as the primary group of microorganisms in study area, as shown in Figure 5a. The soil sample from a depth of 25–5 m contained 94 genera, with the dominant species and genera being Sphingomonas spp. (16.58%), Ramlibacter sp. (7.93%), and Massilia spp. (4.24%). The soil sample from a depth of 3–25 m comprised 101 genera, with most of the species being Methylomonas sp. (10.65%), Sediminibacterium sp. (10.03%), Streptomyces sp. (9.03%), and Staphylococcus sp. (7.13%). The soil sample from a depth of 6–15 m included 37 genera, with the dominant species being Methylomonas sp. (45.73%), Shigella sp. (12.16%), and Brodyrhizobium sp. (10.31%). The soil sample from a depth of 3–15 m contained 71 genera, with the dominant species being Methylomonas sp. (10.01%) and Bradyrhizobium sp. (8.72%). Soil samples of 2–15 m depth showed 86 genera, with the predominant species being Acinetobacter sp. (63.29%), Hydrogenophaga sp. (11.31%), and Acidovorax (8.09%), as shown in Figure 5b. Four significant microbial species were shown to be capable of breaking down chloroethene by anaerobic organohalide respiration: Clostridium, Enterobacter, Geobacter, and Propionibacterium. Aerobic metabolic or co-aerobic pathways identified 15 species with the same potential: Bacillus subtilis, Bacillus megaterium, Bacillus anthracis, Bacillus korlensis, Burkholderia, Methylomonas, Nocardioides, Comamonas, Mycobacterium, Pseudomonas (uncultured bacterium), Pseudomonas oryzihabitans, Pseudomonas (uncultured gamma proteobacterium), Rhodococcus, Ralstonia, and Stenotrophomonas.

3.2.2. Pail Test and Site’s Natural Electron Acceptors

Key parameters such as pH, dissolved oxygen (DO), nitrate, sulfate, phosphate, oxidation–reduction potential (ORP), iron, alkalinity, and hardness were analyzed in two observation wells (OB-1 and OB-2) to evaluate the hydrochemical characteristics. Dissolved oxygen and nitrate act as important electron acceptors under aerobic or slight aerobic condition directly supporting microbial respiration, whereas sulfate and iron concentrations indicate the presence of reducing conditions, which are beneficial to anaerobic microbial communities capable of reductive dechlorination. Table 5 provides the concentrations of these parameters in observation well OB-1 and observation well OB-2 over the 4-month period of the pail test.
Prior to the addition of organic substrates to stimulate indigenous microorganisms in OB-1 and OB-2, the concentrations of VOCs and basic electron acceptors—namely dissolved oxygen and sulfate—were found to be relatively high while nitrate was low. After one month, ORP and the oxygen level declined rapidly, except for dissolved iron, which increased. This increase can be explained by the reduction of insoluble ferric iron (Fe3+) compounds—such as FeOOH(s), Fe2O3(s), or Fe(OH)3(s)—to soluble ferrous iron (Fe2+) upon receiving electrons. This suggests that, in addition to TCE undergoing degradation via the reductive dechlorination pathway, microbial communities in the area may also be utilizing ferric iron as an electron acceptor to facilitate TCE degradation.
Nevertheless, TCE concentrations remained relatively high during the first month following nutrient addition, indicating that microorganisms initially utilized the readily available basic electron acceptors, primarily dissolved oxygen and nitrate in redox processes. By the second month, a notable decrease in TCE concentrations was observed, accompanied by an increase in dissolved iron and a reduction in sulfate levels. This trend suggests a microbial shift in electron acceptor utilization—from basic electron acceptors to those associated with TCE degradation, such as iron and sulfate. In the third month, a slight increase in TCE concentrations was detected, implying that microbial communities may have reverted to using oxygen and/or nitrate, resulting in a reduced rate of TCE degradation.

3.2.3. Hydraulic Conductivity Measurements

The hydraulic conductivity (K) measurements demonstrated a substantial variation, with one deep well (DW-1) having a significantly higher K value than other wells, which demonstrated relatively similar results (Table 6). This divergence may indicate heterogeneity in the subsurface, where the deep well may intersect a more permeable geological layer, such as coarse-grained sediments or fractured rock. The observed variation in hydraulic conductivity (K), particularly the significantly higher value at deep well DW-1 compared to adjacent wells, reflects underlying geological heterogeneity and possibly anthropogenic factors. Geologic heterogeneity could result from localized zones of coarser-grained sediments such as sands or gravels embedded within finer matrix materials like silts or clays, possibly deposited during fluctuating paleo-hydrological regimes. DW-1 may intersect a preferred flow path, such as a fracture zone, which enhances vertical or lateral permeability. This heterogeneity has critical implications for plume migration predictions: areas with higher K can act as conduits for faster contaminant transport, increasing the risk of downgradient receptor exposure, while low-K zones may promote contaminant retention or diffusion-limited degradation. The relatively similar K values in the other wells indicate a more stable sedimentary composition that is likely dominated by materials with fine to medium grains.

3.3. Natural Attenuation Assessment Using BIOSCREEN

The BIOSCREEN results indicate that natural attenuation processes such as dispersion, dilution, and biodegradation actively reduce the contaminant level of TCE (trichloroethylene) along the flow path. The primary mechanism for reducing the concentration of TCE was biodegradation using a first-order decay with a manual calibration of half-life of 4.8 years, which is equivalent to the first-order degradation constant (k) of 0.004 d−1. Moreover, hydrogeological processes such as advection and dispersion also contributed to the attenuation process over time. The BIOSCREEN model simulation using a first-order decay accurately replicated the field data, as determined by the close correlation between the predicted plume mass reductions and spatial distributions depicted in Figure S3. Adsorption and biodegradation reduce VOC concentrations through physical retention and biological transformation, significantly influencing contaminant plume evolution. By simplifying these attenuation mechanisms, the model more accurately reflects field observations and provides insights into the intrinsic capacity of the aquifer system to attenuate VOC contamination naturally.
The initial mass of contamination in Source #1 was 1173 kg in the source zone and slightly decreased to 1166.4 kg, showing that some degradation occurred since the beginning of the period, as shown in Figure S4. When considering the plume without biodegradation, the mass of contaminants would have been 6.5 kg, representing the amount that would be expected to remain in the plume through purely physical processes such as advection and dispersion, whereas the actual mass in the plume is only 2.5 kg, indicating that 4.1 kg (62%) of contamination mass has been removed by the biodegradation process in the study area. This is further supported by the first-order decay model, which shows that degradation is a continuous, slow process, rather than an immediate reaction. The plume has a current groundwater volume of 0.092 Mm3 and a flow rate of 0.00863 Mm3/y through the source zone, indicating groundwater movement dynamics that may lead to dilution and dispersion of the contaminants.

3.4. Groundwater Flow Model Results

The steady-state model calibration was simulated by calibrating hydraulic conductivity parameters to derive a close match between observed and simulated heads (Table 7). The model successfully reproduced the hydraulic head distribution, showing that the hydraulic conductivity values accurately represent the aquifer’s hydraulic conductivity field. The calibration process minimized the differences between observed and simulated heads, suggesting the model’s reliability for groundwater flow simulation. This model appears to be relatively reliable, with most groundwater wells falling within a 95% confidence interval. The values of root mean square errors (RMSEs) and normalized root mean square errors (NRMSEs) are 0.0898 m and 18.33%, respectively. Figure 6a shows that the steady-state hydraulic head distribution suggests that groundwater generally flows from the north to the south. The scatter plot of observed and simulated hydraulic head values shows that the overall trend is consistent with the expected trendline (Figure 6b).
The hydraulic conductivity distribution in the groundwater model revealed distinct variations across both layers, influencing the flow dynamics (Figure 7). In the first layer, horizontal conductivity (HK) indicates a permeable zone that enables groundwater movement. HK increases in the second layer, reaching a maximum value of 0.152 m/d. This higher conductivity in the deeper layer implies a more transmissive formation, potentially leading to increasing vertical flow interactions between layers. These conditions can significantly impact contaminant migration, particularly for volatile organic compounds (VOCs), as they may experience increased downward transport due to the higher hydraulic connectivity. After calibration, the values of vertical anisotropies (Kv/Kh), originally set as 0.1 for both layers, were 0.197 and 0.124 for layers 1 and 2, respectively. The spatial variability in hydraulic conductivity, along with vertical anisotropy, leads to grid-to-grid variations in groundwater flow patterns, which in turn influence the fate and transport of dissolved VOCs within these aquifers.

3.5. Contaminant Transport Model

In the RT3D setup, two contaminant sources were added based on field data. Source 1 represents the primary source of contamination, where TCE was initially released and persisted over time, and Source 2 emerged when the storage containers were moved to the other side, near GW-8, leading to a secondary contamination area. The estimated parameter values from manual calibration are a longitudinal dispersivity of 30 m and first-order biodegradation rates of sequential reaction of 0.009, 0.00055, and 0.001 d−1, respectively. The typical first-order biodegradation rates for TCE range from 0.01 to 0.05 d−1, for DCE from 0.0001 to 0.005 d−1, and for vinyl chloride from 0.0005 to 0.02 d−1 [23]. The rates observed in this study are within these ranges, which are consistent with the inherent variability in biodegradation rates influenced by factors such as microbial activity, and redox conditions. The degradation rate of TCE, which is higher than that of DCE and vinyl chloride, corresponds to the more efficient microbial pathways for its degradation [24]. In contrast, DCE and vinyl chloride tend to degrade more slowly, particularly in anaerobic conditions, where microbial communities may be less effective at degrading these VOCs [25].
The comparison of simulated and measured VOC concentration data is crucial for validating the accuracy of the RT3D model. The concentration vs. time plots of TCE were generated for GW-2, GW-4, GW-5, GW-6, and GW-7 to evaluate the accuracy of RT3D model (Figure S5). The measurements were obtained from field sampling data, and the simulated concentrations were obtained from RT3D output. In GW-7 and GW-4, the simulated concentration corresponds to the measured values, indicating that the selected model parameters, such as reaction rate constants and transport properties, provide a reasonable representation of field conditions. Meanwhile, in other wells, some deviations between simulated and measured concentrations are observed. Despite minor deviations, the simulated and measured values fall within the same order of magnitude, further supporting the model’s reliability.
The concentration vs. time plots of cis-DCE were generated for GW-2, GW-4, GW-5, GW-6, and GW-7 (Figure S6). In GW-2 and GW-6, the simulated concentration closely replicates the field data, indicating that the selected model parameters are sufficient fo capturing the behavior of the plume migration. The simulated data showed a decrease in concentration over time but the measured concentration of cis-DCE exhibited fluctuation. Additionally, seasonal changes in redox conditions and microorganisms may influence degradation rates, resulting in variation in observed data.
In the RT3D model simulation results, Source 1 demonstrates a gradual decrease in the plume size over time, indicating that the natural attenuation processes are effectively reducing the contamination levels in this area. This is supported by the presence of groundwater well data from Source 1, which shows a consistent reduction in TCE concentrations within the plume (Figure 8). On the other hand, Source 2 exhibits minimal natural attenuation, as the concentration of TCE remains relatively stable. This suggests that the absence of wells in this area has limited the effectiveness of attenuation processes. With no further extraction or in situ remediation interventions occurring in this zone, the natural attenuation processes may not be as effective in Source 2. Furthermore, the TCE plume is seen moving across the boundary of the modeled area, indicating that the source may be releasing contaminants into adjacent regions that are not part of the current monitoring area.
In Source 1, RT3D simulation demonstrates a continuous reduction in cis-DCE concentrations over time, representing active natural attenuation processes in the aquifer, as shown in Figure 9. The cis-DCE plume initially increased due to the transformation of TCE; however, over time, it also decreased, suggesting further degradation, possibly into vinyl chloride. The simultaneous reduction in plumes suggests that natural attenuation mechanisms, including microbial degradation, sorption, and dilution, are effectively mitigating the contamination of VOCs. The cis-DCE plume in Source 2 exhibits persistent contamination, indicating that natural attenuation is not significantly reducing its concentration in this area. Unlike TCE, which has migrated beyond the boundary, the cis-DCE plume remains within the study area but exhibits minimal degradation over the period. This persistence suggests that the conditions in Source 2 may not be favorable for reductive dechlorination, potentially due to limited groundwater well data or the absence of necessary electron donors to support biodegradation.
TCE mass reduction due to degradation in Source #1, calculated from RT3D, was 3.93 kg after 30 years, with a total TCE mass emanating from the source zone amounting to 1175.12 kg. While the relative mass reduction levels due to degradation estimated by both BIOSCREEN (see Section 3.3) and RT3D were comparable, BIOSCREEN slightly overestimated biodegradation. This discrepancy may be attributed to BIOSCREEN’s one-dimensional framework, which considers a unidirectional groundwater flow and lacks a transverse dispersion. As a result, BIOSCREEN produced higher plume concentrations, leading to an elevated first-order decay rate compared to RT3D.

4. Discussion

The analysis of soil properties in both of the wells (15 m and 25 m) demonstrated their significant influence on groundwater contamination dynamics and the process of natural attenuation. A slight variation in moisture content, bulk density, specific gravity, porosity, organic matter content, and total organic carbon impacted VOC mobility, adsorption, and biodegradation potential. Microbial analysis indicated that Proteobacteria were the predominant group in the industrial area, significantly contributing to the natural attenuation of VOCs. Key genera such as Geobacter, Burkholderia, and Comamonas contributed to anaerobic and aerobic degradation pathways, facilitating the breakdown of TCE and cis-DCE [26]. The microbial community underscores the biological potential for intrinsic bioremediation, supporting the efficacy of natural attenuation processes in the shallow groundwater aquifer [27]. The composition and functional activity of microbial communities in a shallow aquifer is highly sensitive to seasonal variations and anthropogenic activities, such as nutrient influx, agricultural runoff, or wastewater infiltration. Seasonal fluctuations in temperature, oxygen availability, and groundwater recharge can shift redox conditions, which in turn modulate the abundance and activity of key degraders. Higher groundwater recharge during wet seasons may increase oxygen levels and favor aerobic degradation pathways, while dry seasons can promote anaerobic consortia. Similarly, nutrient influx, particularly nitrate or organic carbon, may stimulate microbial growth but also induce competitive or inhibitory effects that alter VOC biodegradation efficiency. These dynamics can lead to temporal variability in attenuation rates, complicating predictions and requiring continuous monitoring to ensure model assumptions remain valid. The concentrations of groundwater key parameters indicate favorable conditions for natural attenuation, with redox conditions such as iron, manganese, and nitrate suggesting the presence of microbial processes leading to the degradation of the contaminant. The alkalinity and hardness values emphasize the buffering capacity of the aquifer, which influences metal solubility and contaminant mobility. These findings highlight the complex interplay between geochemical conditions and microbial activity, which collectively determine the effectiveness of natural attenuation in mitigating groundwater contamination.
In BIOSCREEN model simulation, TCE degradation did not occur through instantaneous reaction models that rely on electron acceptors such as oxygen, nitrate, sulfate, or observed ferrous iron and the electron acceptor did not change significantly due to the abundance of electron acceptors in the shallow aquifer. The first-order decay model improved this slower attenuation, as it accounts for continuous, steady reductions in TCE concentration over time rather than requiring instantaneous changes in geochemical conditions. The contaminant plume captures 0.092 Mm3 of groundwater, with a flow rate of 0.00863 Mm3/y, suggesting relatively slow groundwater movement. This low flow rate indicates that natural attenuation processes, including dilution, dispersion, and biodegradation, play a crucial role in controlling contaminant migration. The BIOSCREEN model inherently lacks the capability to capture spatial heterogeneities in hydraulic conductivity, porosity, or redox zones, which are often critical in complex field settings like the study site. Such simplifications may lead to under- or overestimation of contaminant plume dimensions, attenuation rates, and the influence of local geologic variability. In contrast, two- and three-dimensional numerical models (e.g., MODFLOW coupled with RT3D) provide a more realistic representation of flow and transport, especially in heterogeneous aquifers with multiple sources, varying boundary conditions, and stratified lithology [28].
The groundwater flow model effectively simulates the movement of groundwater within the shallow aquifer, providing an understanding of hydraulic behavior and potential contaminant transport. The model results indicate a predominant flow direction from north to south, aligning with study site hydrogeology, and highlight variations in flow velocity across different zones. The higher hydraulic conductivity facilitates faster groundwater movement nearby the groundwater wells, which may enhance contaminant dispersion. The minor deviation occurs primarily due to the observed head values falling within a narrow range around 258 m, with only minor changes in the decimal digits. Due to the small study area, the hydraulic gradient is minimal, and the head distribution remains nearly uniform throughout all of the wells. Consequently, even minor discrepancies between measured and calculated hydraulic head values appear more noticeable, and the model accurately simulated the general groundwater flow within the study area [29]. The accurate calibration indicates that the model can be used to predict future groundwater behavior and assess potential impacts on flow dynamics, providing a strong basis for further contaminant transport modeling.
The numerical model reveals spatial and temporal patterns of VOC natural attenuation consistent with field data. The retardation effect of adsorption was evident in the slower migration rates of hydrophobic compounds, as the plume’s leading edge progressed less rapidly than predicted by advection–dispersion alone. Furthermore, modeled biodegradation rates significantly reduced VOC concentrations, demonstrating the importance of microbial activity in attenuating contaminants over time. The integration of these processes in the model allowed the identification of zones within the aquifer where natural attenuation is most effective, such as areas with a higher organic carbon content or favorable redox conditions. This combined approach emphasizes the need to examine both physical and biological mechanisms when evaluating the potential of monitored natural attenuation as a remediation strategy.
During the simulation period, the contaminant plume has spread over time throughout the study area, with maximum concentrations observed nearby source zones. However, over time, the plume shows significant dilution and attenuation as it moves downgradient. The spatial distribution of the VOCs reveals that TCE and cis-DCE undergo varying degrees of attenuation based on their interaction with microbial populations, particularly within the anaerobic zones where organohalide-respiring bacteria are active. The time evolution of contaminant concentrations indicates that while TCE persists longer in the study area, cis-DCE is more readily reduced due to higher degradation rates. In some wells, the concentration from the field data shows occasional increases, which may be attributed to heterogeneous subsurface conditions, fluctuating groundwater flow patterns, or seasonal variations affecting contaminant migration. Additionally, desorption from soil particles, back-diffusion from low-permeability zones, or delayed arrival of contamination from upgradient sources can cause temporary concentration spikes [30]. Variability in microbial activity, influencing degradation rates, and potential sampling inconsistencies could also contribute to these fluctuations. Future research should focus on enhancing the model calibration process with additional field data nearby source 2, and conducting sensitivity analysis to refine the model and enhance its predictive capabilities for groundwater management assessments.
While monitored natural attenuation (MNA) offers a passive, low-cost remediation strategy particularly well-suited for sites with stable plume conditions and evidence of active biodegradation, its effectiveness must be evaluated against alternative methods in terms of implementation complexity, timeframes, and long-term sustainability [31]. Pump-and-treat systems, for instance, provide immediate plume containment and control, but are often energy-intensive, require long operational durations, and may require high maintenance costs, particularly in low-permeability areas [32]. In situ chemical oxidation (ISCO) can rapidly reduce contaminant concentrations, yet it requires precise delivery control and may disrupt local microbial communities [33]. Enhanced bioremediation, which involves injecting nutrients or electron donors to stimulate microbial activity, shows promising degradation rates in VOC-contaminated aquifers, but it requires detailed site characterization and careful dosing to avoid rebound effects or by-product accumulation. In contrast, MNA capitalizes on existing geochemical and microbial conditions, offering long-term efficacy with minimal disturbance, although its decadal timescale and dependence on favorable site conditions can limit its applicability where crucial risk reduction is required. Given the relatively low contaminant flux, microbial indication of VOC degradation, and stable hydrogeological setting at this study site, MNA appears to be a cost-effective and scientifically justifiable strategy, particularly when supplemented with long-term monitoring and contingency planning.

5. Conclusions

This work proposed to evaluate the potential of the natural attenuation process as a remediation strategy for volatile organic compounds (VOCs) in shallow groundwater aquifer at an industrial complex that has been contaminated for more than 30 years. The fate and transport of dissolved VOCs were simulated under site-specific conditions using a combination of numerical models such as MODFLOW for groundwater flow modeling, BIOSCREEN for preliminary screening of natural attenuation potential, and RT3D for reactive transport modeling.
BIOSCREEN results demonstrated that the contaminants are degrading gradually over time rather than being immediately removed, and the primary mechanism of degradation in this study was assumed to be initial-order decay. This finding is consistent with the study area’s conditions, where groundwater wells are located in a single industrial site. The BIOSCREEN analysis provided a reliable investigation of natural attenuation potential, which was further experienced using RT3D. The groundwater flow model was successfully calibrated using measured hydraulic heads, with a root mean square error (RMSE) of 0.08 m and correlation coefficient (r) of 0.81, indicating a suitable fit between simulated and measured data. The RT3D model indicates varying degrees of natural attenuation across the study area, with Source 1 showing a gradual decrease in TCE and cis-DCE concentrations, suggesting that natural attenuation processes, such as biodegradation, advection, and dispersion, are actively involved in mitigating VOCs. The plume is undergoing sequential degradation of VOCs (e.g., TCE → cis-DCE → VC → ethene), consistent with field observations. In contrast, Source 2 exhibits persistent contamination, with both TCE and DCE, likely due to limited groundwater wells and less favorable biodegradation conditions. Furthermore, data from 2017 to 2021 is missing or was not recorded, creating a gap in understanding the contamination trends throughout this time period. This data gap limits the ability to fully assess long-term natural attenuation processes and may have affected the accuracy of model predictions. This data gap restricts the ability to capture significant temporal trends in plume behavior, particularly potential changes in contaminant concentrations, redox conditions, and biodegradation rates that may have occurred during this period. Consequently, model calibration relies heavily on pre-2017 data, which may not accurately reflect the evolving site dynamics or the cumulative effects of attenuation processes. This increases the epistemic uncertainty associated with predictive simulations, particularly when extrapolating the future plume migration or estimating mass removal rates. The lack of continuous data also hinders accurate estimation of time-lagged responses to natural or anthropogenic influences such as changes in groundwater recharge, land use, or microbial adaptation. The model created discrepancies during the simulation process as it was not perfectly aligned with the field data. Models are simplifying reality, and uncertainties remain due to assumptions about the properties of the aquifer, degradation rates, and the availability of measured data. Future research could focus on incorporating more detailed geochemical data, examining the impact of aquifer heterogeneity, and conducting microbial studies to better recognize the role of bacteria in degrading VOCs.
This study has significant implications for the field of hydrogeology and environmental science as it demonstrates the value of integrating multiple numerical modeling tools such as BIOSCREEN, MODFLOW, and RT3D, to evaluate the natural attenuation process in a real-world scenario. Moreover, it provides a practical framework for evaluating natural attenuation as a remediation strategy at industrial sites contaminated with VOCs, highlighting the importance of site-specific data and calibration. This work also contributes to the growing body of evidence supporting natural attenuation as a sustainable and effective remediation strategy, particularly in cases where microbial activity and favorable hydrogeological conditions are present.
The natural attenuation process can take decades to effectively remediate a contaminated area. The industrial site has discovered bacteria that can degrade VOCs in shallow aquifers, and the enhanced bioremediation process can be implemented in the future to improve the remediation process. In Source 2, maintenance is required to ensure the process of bioremediation and potential changes in the concentration of contaminants or plume behavior. This could involve removing or treating the contaminated soil and groundwater in areas with high concentrations. Additionally, long-term monitoring is recommended to ensure that contaminant concentrations remain below regulatory thresholds and to detect any changes in groundwater conditions. The industrial site can effectively manage groundwater contamination by implanting these recommendations to reduce the environmental and human health risks.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/w17132038/s1, Figure S1. Aerial digitized map of study site adapted from ArcGIS® 10.8.1. Figure S2. Regular pilot points are used for parameter estimation. Figure S3. TCE plume centerline. Figure S4. TCE plume mass output. Figure S5. TCE plots of simulated and measured data. Figure S6. cis-DCE plots of simulated and measured data.

Author Contributions

Conceptualization, S.S.; field sampling, data acquisition, data curation, data analysis, data presentation, and data interpretation, M.S.Q., M.Z.A., N.S., and S.T.; model formulation, execution, and calibration, M.S.Q., N.P., M.K., and S.S.; funding acquisition, S.S.; supervision, S.S.; writing—original draft, M.S.Q.; writing—review and editing, M.S.Q. and S.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research work was partially supported by Chiang Mai University and the Chiang Mai University Fundamental Fund 2025.

Data Availability Statement

Water quality data as well as groundwater/contaminant transport model simulation input/output files are available upon request.

Acknowledgments

The CMU Presidential Scholarship is also gratefully acknowledged for supporting Muhammad Shoaib Qamar and Muhammad Zakir Afridi.

Conflicts of Interest

The authors declare no conflicts of interest. The funders had no role in the design of this study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

  1. Meng, J.; Hu, K.; Wang, S.; Wang, Y.; Chen, Z.; Gao, C.; Mao, D. A framework for risk assessment of groundwater contamination integrating hydrochemical, hydrogeological, and electrical resistivity tomography method. Environ. Sci. Pollut. Res. 2024, 31, 28105–28123. [Google Scholar] [CrossRef] [PubMed]
  2. Sankoh, A.A.; Laar, C.; Derkyi, N.S.A.; Frazer-Williams, R. Application of stable isotope of water and a Bayesian isotope mixing model (SIMMR) in groundwater studies: A case study of the Granvillebrook and Kingtom dumpsites. Environ. Monit. Assess. 2023, 195, 548. [Google Scholar] [CrossRef]
  3. Jin, X.; Wu, Y.; Santhamoorthy, M.; Nhi Le, T.T.; Le, V.T.; Yuan, Y.; Xia, C. Volatile organic compounds in water matrices: Recent progress, challenges, and perspective. Chemosphere 2022, 308, 136182. [Google Scholar] [CrossRef]
  4. Yang, Y.; Zhang, Y.; Zheng, H.; Zhang, B.; Zuo, Q.; Fan, K. Functionalized dual modification of covalent organic framework for efficient and rapid trace heavy metals removal from drinking water. Chemosphere 2022, 290, 133215. [Google Scholar] [CrossRef]
  5. Hou, D.; Li, F. Complexities surrounding China’s soil action plan. Land Degrad. Dev. 2017, 28, 2315–2320. [Google Scholar] [CrossRef]
  6. Kumar, A.; Singh, E.; Khapre, A.; Bordoloi, N.; Kumar, S. Sorption of volatile organic compounds on non-activated biochar. Bioresour. Technol. 2020, 297, 122469. [Google Scholar] [CrossRef] [PubMed]
  7. Liu, X.; Wang, Y. Identification and assessment of groundwater and soil contamination from an informal landfill site. Sustainability 2022, 14, 16948. [Google Scholar] [CrossRef]
  8. Lv, H.; Wang, Y.; Wang, H. Determination of major pollutant and biogeochemical processes in an oil-contaminated aquifer using human health risk assessment and multivariate statistical analysis. Hum. Ecol. Risk Assess. Int. J. 2019, 25, 505–526. [Google Scholar] [CrossRef]
  9. Marić, N.; Matić, I.; Papić, P.; Beškoski, V.P.; Ilić, M.; Gojgić-Cvijović, G.; Miletić, S.; Nikić, Z.; Vrvić, M.M. Natural attenuation of petroleum hydrocarbons—A study of biodegradation effects in groundwater (Vitanovac, Serbia). Environ. Monit. Assess. 2018, 190, 89. [Google Scholar] [CrossRef]
  10. Harbaugh, A.W.; McDonald, M.G. Programmer’s Documentation for MODFLOW-96, an Update to the US Geological Survey Modular Finite-Difference Ground-Water Flow Model; US Geological Survey; U.S. Department of the Interior: Reston, VA, USA, 1996. [Google Scholar]
  11. Luo, M.; Zhang, X.; Zhu, X.; Long, T.; Cao, S.; Yu, R. Bioremediation of chlorinated ethenes contaminated groundwater and the reactive transport modeling–A review. Environ. Res. 2024, 240, 117389. [Google Scholar] [CrossRef]
  12. Newell, C.J.; McLeod, R.K.; Gonzales, J.R. BIOSCREEN Natural Attenuation Decision Support System User’s Manual Version 1.3; Rep. EPA/600/R-96; EPA: Washington, DC, USA, 1996; p. 87. [Google Scholar]
  13. Agency, U.E.P. National primary drinking water regulations: Lead and copper rule revisions. Fed. Regist. 2021, 84, 40. [Google Scholar]
  14. U.S. Environmental Protection Agency; Response, R. Risk Assessment Guidance for Superfund; Office of Emergency and Remedial Response, US Environmental Protection Agency: Washington, DC, USA, 1989. [Google Scholar]
  15. ASTM D 422-63R07e2; Standard Test Method for Particle-Size Analysis of Soils. ASTM International: West Conshohocken, PA, USA, 2017. [CrossRef]
  16. ASTM D2487-17e1; Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System). ASTM International: West Conshohocken, PA, USA, 2017. [CrossRef]
  17. Walkley, A.; Black, I.A. An examination of the Degtjareff method for determining soil organic matter, and a proposed modification of the chromic acid titration method. Soil Sci. 1934, 37, 29–38. [Google Scholar] [CrossRef]
  18. Hvorslev, M.J. Time Lag and Soil Permeability in Ground-Water Observations; Waterways Experiment Station, Corps of Engineers, US Army: Washington, DC, USA, 1951. [Google Scholar]
  19. Harbaugh, A.W.; Banta, E.R.; Hill, M.C.; McDonald, M.G. MODFLOW-2000, the US Geological Survey Modular Ground-Water Model: User Guide to Modularization Concepts and the Ground-Water Flow Process; U.S. Department of the Interior: Reston, VA, USA, 2000. [Google Scholar]
  20. McDonald, M.G.; Harbaugh, A.W. A Modular Three-Dimensional Finite-Difference Ground-Water Flow Model; US Geological Survey: Washington, DC, USA, 1988. [Google Scholar]
  21. Doherty, J. PEST model-independent parameter estimation user manual. Watermark Numer. Comput. Brisb. Aust. 2004, 3338, 3349. [Google Scholar]
  22. Clement, T.; Jones, N. RT3D Tutorials for GMS Users; Pacific Northwest National Lab. (PNNL): Richland, WA, USA, 1998. [Google Scholar]
  23. Newell, C.J. Calculation and Use of First-Order Rate Constants for Monitored Natural Attenuation Studies; US Environmental Protection Agency, National Risk Management Research Laboratory: Washington, DC, USA, 2002. [Google Scholar]
  24. Yang, Y.; McCarty, P.L. Competition for hydrogen within a chlorinated solvent dehalogenating anaerobic mixed culture. Environ. Sci. Technol. 1998, 32, 3591–3597. [Google Scholar] [CrossRef]
  25. Bradley, P.M. History and ecology of chloroethene biodegradation: A review. Bioremediat. J. 2003, 7, 81–109. [Google Scholar] [CrossRef]
  26. Najmanová, P.; Steinová, J.; Czinnerová, M.; Němeček, J.; Špánek, R.; Knytl, V.; Halecký, M. Thermally enhanced biodegradation of TCE in groundwater. Water 2022, 14, 3456. [Google Scholar] [CrossRef]
  27. Wright, J.; Kirchner, V.; Bernard, W.; Ulrich, N.; McLimans, C.; Campa, M.F.; Hazen, T.; Macbeth, T.; Marabello, D.; McDermott, J. Bacterial community dynamics in dichloromethane-contaminated groundwater undergoing natural attenuation. Front. Microbiol. 2017, 8, 2300. [Google Scholar] [CrossRef]
  28. Wei, X.; Bailey, R.T.; Records, R.M.; Wible, T.C.; Arabi, M. Comprehensive simulation of nitrate transport in coupled surface-subsurface hydrologic systems using the linked SWAT-MODFLOW-RT3D model. Environ. Model. Softw. 2019, 122, 104242. [Google Scholar] [CrossRef]
  29. Stamos-Pfeiffer, C.; Larsen, J.; Powell, R.E.; Matti, J.C.; Martin, P. Hydrogeology and Simulation of Groundwater Flow in the Lucerne Valley Groundwater Basin, California; US Geological Survey: Washington, DC, USA, 2022. [Google Scholar]
  30. Borden, R.C.; Cha, K.Y. Evaluating the impact of back diffusion on groundwater cleanup time. J. Contam. Hydrol. 2021, 243, 103889. [Google Scholar] [CrossRef]
  31. Ding, D.; Jiang, D.; Zhou, Y.; Xia, F.; Chen, Y.; Kong, L.; Wei, J.; Zhang, S.; Deng, S. Assessing the environmental impacts and costs of biochar and monitored natural attenuation for groundwater heavily contaminated with volatile organic compounds. Sci. Total Environ. 2022, 846, 157316. [Google Scholar] [CrossRef]
  32. Zha, Y.; Yeh, T.-C.J.; Illman, W.A.; Mok, C.M.W.; Tso, C.-H.M.; Carrera, B.A.; Wang, Y.-L. Exploitation of pump-and-treat remediation systems for characterization of hydraulic heterogeneity. J. Hydrol. 2019, 573, 324–340. [Google Scholar] [CrossRef]
  33. Sutton, N.B.; Atashgahi, S.; van der Wal, J.; Wijn, G.; Grotenhuis, T.; Smidt, H.; Rijnaarts, H.H. Microbial dynamics during and after in situ chemical oxidation of chlorinated solvents. Groundwater 2015, 53, 261–270. [Google Scholar] [CrossRef] [PubMed]
Figure 1. Concentration contour of TCE during 2011–2016 and 2022 in the study area.
Figure 1. Concentration contour of TCE during 2011–2016 and 2022 in the study area.
Water 17 02038 g001
Figure 2. Concentration contour of cis-DCE during 2011–2016 and 2022 in the study area.
Figure 2. Concentration contour of cis-DCE during 2011–2016 and 2022 in the study area.
Water 17 02038 g002
Figure 3. Concentration contour of vinyl chloride during 2022 in the study area.
Figure 3. Concentration contour of vinyl chloride during 2022 in the study area.
Water 17 02038 g003
Figure 4. Locations of groundwater, soil sampling, and slug test wells in study area.
Figure 4. Locations of groundwater, soil sampling, and slug test wells in study area.
Water 17 02038 g004
Figure 5. (a) Microbial population groups classified at phylum level and (b) microbial species analyzed within the study area.
Figure 5. (a) Microbial population groups classified at phylum level and (b) microbial species analyzed within the study area.
Water 17 02038 g005
Figure 6. (a) Hydraulic head contour plot and groundwater flow direction (see arrows) and (b) observed head vs. model simulated heads after model calibration.
Figure 6. (a) Hydraulic head contour plot and groundwater flow direction (see arrows) and (b) observed head vs. model simulated heads after model calibration.
Water 17 02038 g006
Figure 7. Spatial distribution of hydraulic conductivities after model calibration.
Figure 7. Spatial distribution of hydraulic conductivities after model calibration.
Water 17 02038 g007
Figure 8. Snapshot of predicted TCE concentration plume.
Figure 8. Snapshot of predicted TCE concentration plume.
Water 17 02038 g008
Figure 9. Snapshot of cis-DCE concentration plume.
Figure 9. Snapshot of cis-DCE concentration plume.
Water 17 02038 g009
Table 1. Testing methods and standards of soil physical and chemical analysis.
Table 1. Testing methods and standards of soil physical and chemical analysis.
Soil PropertiesMethod of Study
Physical ParametersGrain size (USCS)
Wet sieve analysis
Hydrometer
Atterberg’s limits: liquid limit and plastic limit
ASTM D422 [15]
ASTM-D2487 [16]
Bulk densityCore method
Soil porosityCalculated from bulk density and specific gravity
Chemical ParametersOrganic matter (OM)Walkley and Black [17]
Total organic carbon (TOC)Combustion
Table 2. Summary of the BIOSCREEN input parameters.
Table 2. Summary of the BIOSCREEN input parameters.
ParametersValues
Seepage velocity (m/yr)11.09
Hydraulic conductivity (m/s)5.0 × 10−5
Hydraulic gradient (m/m)0.008
Porosity0.3718
Longitudinal dispersivity, αL (m)5.23
Transverse dispersivity, αT,H (m)0.52
Retardation factor (m)1.0
Bulk density (kg/L)1.6475
Partition coefficient, Koc (L/kg)2.42
Fraction organic carbon, foc0.0033
First-order degradation constant, k (d−1)0.004
Delta oxygen (mg/L)1.2
Delta sulfate, SO42− (mg/L)13.9
Ferrous iron, Fe2+ (mg/L)1.3
Delta nitrate, NO3 (mg/L)0.3
Table 3. Summary of the model setup.
Table 3. Summary of the model setup.
Model SetupDescription/Meaning/Unit
1. Scale of model
    ▪ width of model (UTM-E)505,746–506,144 m
    ▪ length of model (UTM-N)2,055,059–2,055,485 m
    ▪ top and bottom of model (msl.)234.6–259.6 m (amsl)
2. Model grid and layer
    ▪ grid cell size (West-East)9.95 m
    ▪ grid cell size (North-South)10.65 m
    ▪ number of columns40
    ▪ number of rows40
    ▪ total number of grid blocks or cells3200
3. Layer setting
    ▪ number of layers2
    ▪ thickness of the layerVariable
4. Units
    ▪ length and thickness of the layerm
    ▪ timed
    ▪ hydraulic conductivity (K)Spatially variable (m/d)
    ▪ vertical anisotropy (Kv/Kh) for both layers0.1 (before calibration)
5. Number of monitoring wells10
Table 4. Soil properties obtained from 15 m and 25 m boreholes.
Table 4. Soil properties obtained from 15 m and 25 m boreholes.
PropertiesWell, 15 mWell, 25 m
Min.Max.AverageMin.Max.Average
Moisture Content (%)13.7121.7518.8611.5251.9625.92
Bulk Density (g/cm3)1.591.781.711.431.791.68
Specific Gravity2.542.672.602.612.742.64
Porosity (%)29.8940.3434.3233.0845.2836.32
Organic Matter (%)0.081.850.450.020.310.11
Total Organic Carbon (%)0.051.070.350.240.420.31
Table 5. Concentration of key parameters in OB-1 and OB-2 during 4-month pail test.
Table 5. Concentration of key parameters in OB-1 and OB-2 during 4-month pail test.
ParametersUnitOB-1OB-2
Average ± SDAverage ± SD
Ironmg/L0.50 ± 0.331.242 ± 0.99
Manganesemg/L0.13 ± 0.030.872 ± 1.62
Bicarbonate Alkalinity ( H C O 3 )mg/L373.4 ± 72.71573.8 ± 33.48
Carbonate   Alkalinity   ( C O 3 2 ) mg/L<1<1
Carbonate   Hardness   ( C a C O 3 ) mg/L284.8 ± 33.55343.8 ± 13.72
Residual Chlorinemg/L0.03 ± 0.000.040 ± 0.00
Dissolved Oxygenmg/L3.37 ± 0.962.525 ± 1.17
Hydroxide Alkalinitymg/L<1<1
Nitrate   ( NO 3 ) mg/L<1<1
Non-Carbonate Hardnessmg/L5.00 ± 0.00<1
Oxidation Reduction PotentialmV378.0 ± 116.4333.2 ± 108.0
pH at 25 °C 6.92 ± 0.547.100 ± 0.25
Phenolphthalein Alkalinitymg/L<1<1
Phosphate   ( PO 4 3 ) mg/L1.63 ± 0.351.300 ± 0.00
Residual Alkalinitymg/L102.0 ± 53.56243.3 ± 29.01
Sulfate   ( SO 4 2 ) mg/L3.28 ± 2.6512.23 ± 11.65
Total Alkalinitymg/L373.4 ± 72.71573.8 ± 33.48
Total Hardnessmg/L286.3 ± 33.88343.8 ± 13.72
Table 6. Hydraulic conductivities from slug test wells.
Table 6. Hydraulic conductivities from slug test wells.
Tested BoreholesDepth (m)K (m/d)
DW-1250.12361
SW-190.03216
DW-2250.01536
DW-315.50.03792
SW-2150.01848
Table 7. Groundwater level measurement in wells.
Table 7. Groundwater level measurement in wells.
WellObserved Head (m)Computed Head (m)Residual Head (m)
GW1258.754258.802−0.048
GW2258.629258.738−0.109
GW3258.834258.740.094
GW4258.764258.7310.033
GW5258.759258.7180.041
GW6258.612258.721−0.109
GW7258.899258.7040.195
GW8258.409258.45−0.041
GW9258.489258.4350.054
GW10258.519258.539−0.02
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Qamar, M.S.; Santha, N.; Taweelarp, S.; Ploymaklam, N.; Khebchareon, M.; Afridi, M.Z.; Saenton, S. Evaluating Natural Attenuation of Dissolved Volatile Organic Compounds in Shallow Aquifer in Industrial Complex Using Numerical Models. Water 2025, 17, 2038. https://doi.org/10.3390/w17132038

AMA Style

Qamar MS, Santha N, Taweelarp S, Ploymaklam N, Khebchareon M, Afridi MZ, Saenton S. Evaluating Natural Attenuation of Dissolved Volatile Organic Compounds in Shallow Aquifer in Industrial Complex Using Numerical Models. Water. 2025; 17(13):2038. https://doi.org/10.3390/w17132038

Chicago/Turabian Style

Qamar, Muhammad Shoaib, Nipada Santha, Sutthipong Taweelarp, Nattapol Ploymaklam, Morrakot Khebchareon, Muhammad Zakir Afridi, and Schradh Saenton. 2025. "Evaluating Natural Attenuation of Dissolved Volatile Organic Compounds in Shallow Aquifer in Industrial Complex Using Numerical Models" Water 17, no. 13: 2038. https://doi.org/10.3390/w17132038

APA Style

Qamar, M. S., Santha, N., Taweelarp, S., Ploymaklam, N., Khebchareon, M., Afridi, M. Z., & Saenton, S. (2025). Evaluating Natural Attenuation of Dissolved Volatile Organic Compounds in Shallow Aquifer in Industrial Complex Using Numerical Models. Water, 17(13), 2038. https://doi.org/10.3390/w17132038

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop