Modeling Hydrocarbon Plume Dynamics in Shallow Groundwater of the Rey Industrial Area, Iran: Implications for Remediation Planning

Abstract

The rapid expansion of the petrochemical industry has led to significant environmental issues, including groundwater and soil contamination from hydrocarbon spills. This study investigates the movement and dispersion of hydrocarbon contaminants in the Rey industrial area in Tehran (Iran) using a two-dimensional finite element model. The results indicate that the oil plume exhibits slow migration, primarily due to low soil permeability and high hydrocarbon viscosity, leading to localized contamination. High-density pollution zones, such as TORC and REY7, are characterized by persistent hydrocarbon accumulation with minimal lateral migration. The findings emphasize the limited effectiveness of natural attenuation alone, highlighting the need for targeted remediation measures in high-density zones to accelerate contamination reduction. This study provides insights into the dynamics of hydrocarbon pollution and supports the development of effective remediation strategies.

1. Introduction

The rapid expansion of the petrochemical industry has emerged as a significant global environmental threat, mainly due to leakage from aging storage and transportation infrastructure [1,2,3,4]. In industrial hotspots such as Tehran’s Rey industrial area, petroleum-based contaminants undergo complex physicochemical changes that hinder natural attenuation. Upon migrating through soils and aquifers, they cause immediate and long-term damage to ecosystems [5,6]. A semi-arid region with limited water resources faces special challenges caused by these persistent pollutants.

Petroleum hydrocarbons, including crude oil and refined products such as gasoline, diesel, and kerosene, are among the most prevalent organic contaminants in groundwater systems. Their environmental persistence stems from a triple threat: low aqueous solubility, chronic toxicity, and degradation pathways constrained by subsurface conditions [7,8]. The Rey site has undergone decades of accumulated non-aqueous phase liquids (NAPLs) creating a layered contamination legacy of light LNAPLs floating on water tables and dense DNAPLs sinking through vadose zones [9,10,11]. Groundwater plumes continuously absorb dissolved-phase toxins from DNAPL reservoirs in deep strata, complicating remediation.

Quantifying the extent of hydrocarbon leakage remains challenging, even in geologically simple environments [12]. In addition to field-based observational studies, numerical simulations of subsurface methane flow and transport are critical for refining conceptual models and improving the assessment of remediation strategies [13]. These models help address uncertainties that observational approaches alone cannot resolve, particularly in relation to fluid flow and contaminant transport to groundwater. However, it is still difficult to translate theoretical models into practical remediation strategies, especially for low-permeability environments such as Tehran’s dense clay soils, where high-viscosity hydrocarbons inhibit natural flushing.

This study presents a two-dimensional finite element model specifically calibrated to the Rey industrial area’s unique hydrogeological profile. The findings provide a quantitative framework for targeting intervention in high-density kill zones such as the seven oil companies in Rey industrial area (REY7) and Tehran Oil Refining Company (TORC), thereby challenging passive remediation paradigms.

2. Study Area

2.1. Geological and Hydrogeological Setting

Tehran’s geological setting, shaped by the collision of the Alborz Mountains and the Iranian plateau, creates a tectonically unstable landscape that amplifies contamination risks. The Tertiary and Quaternary formations underlying the metropolitan region consist primarily of Eocene volcanic sequences, including tuff, andesite, and agglomerates. These formations are intersected by active fault systems such as the Masha, North Tehran, and Rey faults. These faults not only pose seismic hazards but also fracture bedrock, enabling contaminants to migrate downward into aquifers. To the south, the Rey industrial area rests on late Eocene volcanic layers overlain by thick alluvial deposits eroded from the Alborz Mountains. Borehole data reveals a deceptively uniform subsurface: 10 m of low-permeability clay overlie 10 m of silt, forming a hydraulically connected but slow-draining system. Core samples near pumping stations confirm negligible vertical heterogeneity, with permeability coefficients as low as 2.14 × 10⁻⁸ m/s at the TORC. This low permeability is a key factor in exacerbating hydrocarbon retention in the industrial zone [14].

Historic surveys highlight structural complexities that exacerbate contamination. An east–west fault structure near Qaleh Nou was identified by a French-led study conducted in 1964 that spanned from Azim Abad to Shahre Abad. As a result of these faults, the Alborz range continues to be uplifted, and the Tehran basin continues to subside due to intrusions of gabbro igneous bodies formed during ancient tectonic collisions. This dynamic creates a dual risk: (1) low-permeability clays and silts hinder natural flushing of contaminants, while (2) fractures in Mesozoic carbonate bedrock provide conduits for DNAPLs to infiltrate deep aquifers. Rey’s layered contamination results from LNAPLs pooling at the water table and DNAPLs sinking into fractures [15].

The semi-arid climate in this region amplifies risks, since toxins leach from geological traps due to limited recharge. The alluvial plains, though topographically flat, provide sinks for volcaniclastic sediments eroded under the sea, which adsorb hydrocarbons, further complicating remediation efforts. A calibrated finite element framework was developed to address how Tehran’s dense, fault-fractured strata govern the persistence of petroleum plumes under these conditions.

2.2. Oil Contamination Background

The Rey industrial area has served as a persistent hotspot for hydrocarbon contamination, driven by decades of chronic leakage from aging infrastructure, including buried pipelines, storage tanks, drainage systems, and qanat channels. Initial contamination incidents were documented in the 1970s, with severe ecological and public health consequences emerging by the 1980s. Despite early remediation efforts by Iranian authorities, inadequate infrastructure and persistent leakage exacerbated contamination, necessitating renewed investigations in subsequent decades [14].

In response, the TORC initiated groundwater monitoring campaigns, collecting samples and tracking hydraulic head fluctuations to assess pollution spread. Pumping systems were deployed to extract contaminated water, but these measures proved insufficient to curb advancing plumes. By 2002, the escalating crisis prompted TORC to collaborate with the Japan Cooperation Center for Petroleum (JCCP), launching a formal feasibility study to evaluate remediation options. In 2003, n-hexane extractable material (HEM) tests, soil oil analysis, and measurements of oil layer thickness were used in accordance with Japanese environmental protocols. Results confirmed significant pollutant migration aligned with local groundwater flow gradients, underscoring the inadequacy of existing mitigation strategies [15].

In 2006, a comprehensive survey of seven industrial facilities revealed widespread contamination. Water samples from monitoring wells and Qanats routinely exceeded permissible limits for agricultural use, with suspended solids surpassing 100 mg/L and methyl tert-butyl ether (MTBE) concentrations peaking in southern sectors. Spatial analyses (Figure 1, Figure 2 and Figure 3) delineated distinct contamination zones: high-density plumes near TORC and REY7, where oil concentrations reached 900 mg/L, and peripheral low-density zones (<300 mg/L) influenced by secondary dispersal mechanisms.

High-density zones correlate with direct leakage from pipelines and tanks, while low-density areas reflect secondary pathways such as oil-laden irrigation water and qanat-mediated transport. Figure 3 highlights aged infrastructure and contaminated surface drainage as key contributors to peripheral pollution. A hydrogeological analysis indicates that secondary zones are sustained by persistent migration from high-density cores, creating an interconnected contamination network. These findings challenge passive remediation paradigms, advocating instead for immediate intervention in high-density zones (e.g., TORC, REY7) and stringent controls on secondary dispersal pathways to mitigate long-term environmental and public health risks.

3. Methodology

3.1. Numerical Modeling

This study employs a two-dimensional finite element model to simulate the flow and transport of hydrocarbon contaminants in the Rey industrial area with a dimension of 6000 by 6000 m. The model incorporates the physics of two-phase flow (oil and water phases) and accounts for the transport of each phase through advection and dispersion processes.

In the event of an oil spill occurring above the water table of an unconfined aquifer, the oil infiltrates the unsaturated zone under the influence of gravitational forces. As the oil moves vertically, it also disperses laterally due to capillary forces and hydrodynamic dispersion [17,18,19]. However, in coarse-grained soils, the lateral dispersion can often be disregarded, as capillary forces are minimal. As the oil front advances, it leaves residual oil trapped within the pore spaces, which is attributed to surface tension effects. If the oil spill is sufficiently extensive, the oil would eventually reach the saturated water zone. At this stage, the flow of oil is influenced by its density relative to that of water. Heavy oil, with a density exceeding 1 g/cm³, displaces water and continues its vertical movement until it encounters a layer of low permeability. Upon reaching this impervious layer, the oil spreads laterally due to its own pressure and gravitational forces. Conversely, lighter oil, which has a density less than that of water, is intercepted by the capillary fringe zone. Within this zone, the oil spreads laterally, forming a lens until it attains residual saturation. This study focuses on modeling the recovery of an immiscible oil lens that is floating on the water table. The downward flow of oil towards the water table and its subsequent lateral spreading constitutes a complex process, involving a three-phase flow system. In addition to the vertical migration of non-aqueous oil, some oil may volatilize, creating a gaseous plume in the unsaturated soil zone. Furthermore, some non-aqueous oil and its vapors may dissolve in water, complicating the study of the entire system. Consequently, to develop a mathematical model, certain simplifying assumptions must be established [20,21].

In this study, we assumed the aquifer is homogeneous and isotropic. The porosity remains constant throughout the domain, and the pore sizes are uniform. The transition zone between maximum and minimum fluid saturations is represented by a distinct front. The aquifer exhibits rigidity, and the fluids are incompressible. The relative permeabilities of both oil and water are assumed to be constant; however, this assumption does not hold true in the unsaturated zone where fluid saturation varies significantly, yet it is applicable in a relatively narrow zone, where oil spreads laterally. Furthermore, oil is considered a conservative substance, thus it does not dissolve in water, volatilize, react with soil, or undergoes biodegradation. In the unsaturated zone, air is at atmospheric pressure, leading to the neglect of air flow, which simplifies the three-phase flow system to a two-phase flow system. Within the oil lens, oil is mobile while water remains immobile at residual saturations; conversely, below the water table, only the water phase is mobile. The Dupuit–Forchheimer assumptions are applicable, which suggest that flow is horizontal and valid, where vertical head gradients and flow components are minimal compared to horizontal flow.

Based on the mass-equilibrium equation of the oil phase in a porous medium, we simulated the distribution of oil lens and migration with groundwater flow [17,18].

$$\nabla .{\rho }_o{q}_o+{\displaystyle \frac{\partial \left({\rho }_o{S}_{o}n\right)}{\partial t}})=0,$$

(1)

where $q_o$ is the specific discharge of the oil phase; ${\rho }_o$ is the density of the oil phase; $S_w$ and $S_o$ are the degrees of saturations of water and oil phases, respectively; n denotes the porosity, and t represents the time. In a two-phase system, $S_w+S_o=1$. Darcy’s law can be expressed by the specific oil discharge, q_o, in terms of oil potential [17,18]:

$$q_o=-K_o\nabla {\varphi }_o=-K_o\nabla \left(h-{\displaystyle \frac{P_{ao}}{{\rho }_{o}g}}\right)\approx -K_o\nabla h,$$

(2)

where ${\varphi }_o$ is the potential head of oil, $K_o$ represents the hydraulic conductivity of oil, $P_{ao}$ is the capillary pressure between air and oil phases, h is the elevation of the lens surface and, $g$ is the gravitational acceleration. In Equation (2), we ignore the capillary pressure gradient according to the sharp interface suppositions.

By averaging Equation (1) along the vertical in the oil from η(r,t) to h(r,t) and stating the thickness of lens as L(x,y,t) = h(x,y,t) − η(x,y,t) and assuming that the oil saturation within the oil lens is constant, we yield $S_{oo}={S\prime }_{oo}$, replacing Equation (2), and reordering as [17,18]:

$$\nabla .\left({\rho }_o{K}_{o}L\nabla L\right)+\nabla .\left({\rho }_o{K}_{o}L\nabla \eta \right)=\pm {\rho }_o{Q}_o\delta \left(x-\xi \right)\delta \left(y-\zeta \right)+n{\rho }_o\left(S_{oo}-S_{oun}\right)\left(\partial L/\partial T\right)$$

(3)

where $S_{oo}$ and $S_{oun}$ are degrees of oil saturation in the oil lens and in the unsaturated soil, respectively, and η is the elevation of the water table. $Q_o$ is the rate of oil leak/pumpage at the point (ξ,ζ). Here, $\delta$ defines the Dirac delta function. Under hydrostatic conditions, when a finite layer of oil exists on the water surface—representing the oil–water interface—the pressures of the oil and water phases, $P_o$ and $P_w$, can be expressed in terms of their respective potential heads, ${\varphi }_o$ and ${\varphi }_w$.

$$P_o={\rho }_{o}g\left({\varphi }_o-\eta \right) P_w={\rho }_{w}g\left({\varphi }_w-\eta \right)$$

(4)

where ${\rho }_o$ and ${\rho }_w$ are the densities of oil and water, respectively, and, η is the elevation of the water table as measured from a datum. By ignoring the capillary pressure between the water and oil phases, for example, $P_o=P_w$, the potential water head is obtained from Equation (4):

$${\varphi }_w=\left(1-{\displaystyle \frac{{\rho }_o}{{\rho }_w}}\right)\eta +{\displaystyle \frac{{\rho }_o}{{\rho }_w}}{\varphi }_o.$$

(5)

The elevation of the air–oil interface, h (i.e., oil surface), as evaluated from the data, is the sum of the oil lens thickness L and η (i.e., h = η + L). By ignoring the capillary pressure between the air and oil phases (i.e., ${\varphi }_o=h$), the specific discharge of water $q_w$ can be expressed by Darcy’s law as follows:

$$q_w=-K_w\nabla {\varphi }_w=-K_w\nabla \eta -K_w{\displaystyle \frac{{\rho }_o}{{\rho }_w}}\nabla L$$

(6)

where $K_w$ is the hydraulic conductivity of water. The last term on the right-hand side of Equation (6) indicates the water flow due to a change in the thickness of the oil. At the maximum thickness of the oil lens, this term becomes negligible and does not contribute to additional water flow. Presuming that the density ${\rho }_o$ and hydraulic conductivity $K_o$ of oil are constants according to assumptions of sharp interface, Equation (3) can be rewritten as follows:

$$\nabla .\left(L\nabla L\right)+\nabla .\left(L\nabla \eta \right)=\pm {\displaystyle \frac{1}{K_o}}{Q}_o\delta \left(x-\xi \right)\delta \left(y-\zeta \right)+\left[{\displaystyle \frac{n{S}_{oo}-n{S}_{oun}}{K_o}}\right]{\displaystyle \frac{\partial L}{\partial t}},$$

(7)

where $S_{oun}$ and $S_{oo}$ are degrees of saturation in the unsaturated soil and oil saturation in the oil lens, respectively. The composition and reordering of Equations (6) and (7) and the supposition of constant velocity of groundwater flow give the following:

$$\left(1-{\displaystyle \frac{{\rho }_o}{{\rho }_w}}\right)\nabla .\left(L\nabla L\right)-{\displaystyle \frac{q_w}{K_w}}\nabla L=\pm {\displaystyle \frac{1}{K_o}}{Q}_o\delta \left(x-\xi \right)\delta \left(y-\zeta \right)+\left[{\displaystyle \frac{n{S}_{oo}-n{S}_{oun}}{K_o}}\right]{\displaystyle \frac{\partial L}{\partial t}}.$$

(8)

In Equation (8), quadratic terms such as $\nabla L.\nabla L$ can be ignored for small $\nabla L$ values. Then, this equation can be rewritten by linearizing the resulting equation as follows:

$$V_L={\displaystyle \frac{K_o{L}_o\left({\rho }_w-{\rho }_o\right)}{n{\rho }_w}}$$

(9)

where $V_L$ is the horizontal velocity of the oil lens front, $L_o$ is the initial thickness of the oil lens [16]. Equation (9) is an advective–dispersive transport equation. The advection term, which is proportional to a specific discharge of water, refers to the migration of oil lens with ambient groundwater flow.

The dispersion term refers to the gravitational spreading of oil lens on the water table due to the density difference (${\rho }_w-{\rho }_o$). The center of the oil mass moves at a velocity determined by the slope of the water table ($q_w/K_w$) multiplied by the hydraulic conductivity of oil $K_o$ divided by the oil-filled pore [n ($S_{oo}-S_{oun}$)].

3.2. Model Validation

Figure 4 and Figure 5 show the location of collection wells at the study site. In the oil layer collection plan, 148 collection wells are considered. Previous investigations have shown that there is high-density oil pollution at the pumping station in the south of the TORC. Out of 148 collection wells, wells 6 and 53 were randomly selected to validate the developed model and obtain the oil layer thickness. The oil layer thickness and groundwater level at wells 6 and 53 are shown in Figure 6.

Hereafter, we identify the measured LNAPL thickness from selected wells in the study area.

Table 1. Input parameter.

Parameter	Symbol	Value
Oil permeability coefficient	Kij	2.14 × 10−4 cm/s
Void ratio	n	0.4194
Saturation fraction	S	80.7%
Initial elevation of water table	ηo	12 m
Initial thickness of the oil lens at the well	Lo	3 m

shows the input parameters for which we matched the calculated and measured LNAPL thickness. Figure 7 shows the resulting agreement between predicted and measured LNAPL thickness for the two oil collection wells. Because of the agreement between predicted and measured LNAPL recovery for the wells tested, we utilized finite element models to determine recoverable volumes and LNAPL thickness for the wells.

3.3. Numerical Implementation

FEATool Multiphysics, a MATLAB toolbox, was used to solve the governing equations numerically. The computational domain represented a cross-section of the Rey industrial area, discretized into finite elements (Figure 8). The model focused entirely on two-dimensional analysis to capture the lateral distribution of the contaminant plume. This approach assumes that vertical variations in contaminant distribution are negligible, allowing for computational efficiency while accurately representing horizontal migration. The mesh density was refined in areas with high gradients in oil saturation or capillary pressure to improve numerical accuracy. Temporal discretization was performed using an implicit time-stepping scheme.

3.4. Material Properties

The physical and hydraulic properties of the materials used in the simulation are summarized in

Table 2. Material properties.

Property	Unit	Value
Soil porosity ($\varphi$)	-	0.42
Hydraulic conductivity of the water phase ($K_W$)	cm/s	1.19 × 10−8
Hydraulic conductivity of the water phase ($K_o$)	cm/s	8.07 × 10−6
Diffusion coefficient	m2/s	1 × 10−9
Initial elevation of water table (${\eta }_o$)	m	12
Initial thickness of the oil lens at the well ($L_o$)	m	3
Oil viscosity (${\mu }_o$)	Pa·s	0.6
Oil density (${\rho }_o$)	kg/m3	800
Water viscosity (${\mu }_d$)	Pa·s	0.001
Water density (${\rho }_w$)	kg/m3	1000
The saturation of non-aqueous phase liquid (NAPL) in the NAPL lens ($S_{oo}$)	-	0.85
The residual saturation of water ($S_{ow}$)	-	0.15
The residual saturation of NAPL ($S_{oun}$)	-	0.05

. These values reflect the characteristics of the soil, hydrocarbon contaminants, and groundwater in the Rey industrial area.

3.5. Boundary and Initial Conditions

The lateral boundaries of the model domain were set with a fixed oil layer thickness of zero, ensuring no inflow or outflow of contaminants. The bottom and surface boundaries were considered impermeable, reflecting geological constraints and preventing vertical contaminant migration. Groundwater flow was modeled as southward, based on observed hydrological gradients in the Rey area.

The initial conditions for the simulation were derived from field measurements. The initial distribution of the contaminant was based on the maximum oil layer thickness observed at wells in the TORC and REY7 areas (Figure 9). Groundwater levels were initialized using the measured water table distribution.

In TORC and the seven companies in the Rey industrial area, the water level in the monitoring wells and core borings was measured. These measurement results were used for estimating the distribution of groundwater level and groundwater flow direction, as shown in Figure 10. Monitoring wells and core borings in which water level was measured are shown with the (●) sign. The groundwater flow direction was estimated to be towards the southwest.

4. Results

The numerical simulations provided a detailed evaluation of the hydrocarbon contamination dynamics in the Rey industrial area over a two-year period, capturing the temporal and spatial variations in oil plume thickness. The results reveal slow plume attenuation and limited lateral migration, with observations recorded at intervals of 7.3 days, one year, and two years.

At the initial stage of 7.3 days, the oil plume remained concentrated around TORC and REY7, where the initial contamination was most severe (Figure 11a). The maximum thickness of the oil plume exceeded 6.5 m, forming a dense core that showed minimal movement. The plume boundary, defined by a thickness threshold of 0.5 m, exhibited a relatively stable geometry, closely following the initial distribution. These findings indicate that, within the first week, the viscous nature of the hydrocarbon oil and the low permeability of the soil effectively constrained the contaminant’s movement.

After one year, the attenuation of the oil plume became more evident, particularly in the central high-thickness zones (Figure 11b). The region with plume thickness exceeding 6.5 m contracted significantly, suggesting ongoing natural processes such as dissolution and minor biodegradation. However, the overall shape of the plume showed limited lateral expansion, with the boundary of the 0.5 m thickness region remaining relatively stable. The southward migration of the plume, driven by groundwater flow, was minimal, with only slight displacements observed at the southern edge of the domain. This reflects the weak hydraulic gradient and the soil’s restrictive permeability, which hinder significant contaminant transport.

By the two-year mark, the oil plume demonstrated further reductions in thickness across all regions (Figure 12). In the high-contamination zones near TORC and REY7, the plume thickness diminished considerably, with areas exceeding 6.5 m becoming highly localized. The boundary regions, where the thickness initially approached 0.5 m, showed further contraction, indicating gradual attenuation over time. Despite these reductions, the overall movement of the plume remained subdued, with negligible lateral or southward migration observed. The results show the persistent and localized nature of the contamination, with high-density zones continuing to dominate the subsurface dynamics.

The results also highlight the relatively slow dynamics of the plume’s boundary over time. Compared to the rapid thinning of the central regions, the boundary changes occurred at a much slower rate, reflecting the limitations of natural groundwater flow in driving the significant dispersal of the contaminant. Overall, the simulations reveal a contamination system characterized by low mobility, localized persistence, and gradual attenuation.

5. Discussion

The simulation results provide key insights into the physical processes governing hydrocarbon contaminant transport in the Rey industrial area. The slow attenuation of the oil plume, particularly in the high-density central regions near TORC and REY7, can be attributed to the combined effects of the soil’s low permeability and the high viscosity of the hydrocarbon oil. These properties significantly constrain the contaminant’s mobility, resulting in a highly localized plume that undergoes minimal lateral expansion over time. The observed plume behavior aligns with theoretical expectations for viscous non-aqueous phase liquids (NAPLs) in fine-grained soils, where capillary forces and viscous resistance dominate over advective transport mechanisms.

The limited southward migration of the plume, despite the direction of groundwater flow, highlights the critical role of the contaminant’s physical properties in controlling transport dynamics. The weak hydraulic gradient and the high viscosity of the hydrocarbon oil effectively suppress advective flow, confining the contaminant to the central regions. This localized behavior reduces the immediate risk of contamination spreading to downstream areas but also prolongs the persistence of the pollution, posing long-term challenges for environmental management.

The slow contraction of the 0.5 m thickness region suggests that diffusion and dispersion play only minor roles in the contaminant’s transport. This behavior emphasizes the importance of implementing containment and remediation strategies that address the limited mobility of the contaminant under existing site conditions. Techniques such as skimming wells for oil recovery, combined with in situ bioremediation to enhance natural attenuation, could effectively manage the contamination, particularly in the high-density zones.

While the study provides valuable insights, it is important to acknowledge the limitations of the modeling approach. The reliance on two-dimensional analysis assumes negligible vertical dynamics, which may oversimplify interactions between subsurface layers. Vertical processes, such as buoyancy-driven flow or capillary retention, could play a significant role in the redistribution of the contaminant, particularly in heterogeneous stratigraphy. Future studies should incorporate three-dimensional modeling to capture these effects and provide a more comprehensive understanding of the subsurface system.

6. Separation of Contaminated Zones and Proposing Remediation Strategies

Two distinct pollution zones can be identified, each requiring specific remedial measures: (1) high-density oil contamination zones, where oil layers are present in monitoring wells, and (2) low-concentration oil contamination zones, where no oil layer is observed in the wells. The first zone is primarily concentrated around TORC and REY7 within the broader Rey area, while the second zone is more widely distributed.

6.1. Oil Collection Well Installation Plan in the Southern Area of REY7

We propose the installation of an oil collection well in the southern area of REY7. High-density oil contamination has been identified in the monitoring wells at TORC and REY7. Based on the investigation of the entire Rey area, low-concentration oil pollution appears to be widely distributed, while high-density contamination is more localized. However, there is a possibility that another high-density oil contamination zone exists south of REY7. Therefore, the implementation of remediation measures, similar to those applied at TORC and REY7, is recommended. Specifically, the installation of an oil collection well in this area is considered an appropriate response.

6.2. Pollution Measures in Whole Rey Area

As can be seen from the results of the soil oil content analysis, oil pollution has spread over a vast area around the TORC. This includes agricultural fields, TORC’s forest, residential and industrial buildings, the oil sludge pit and other fields. A schematic view on the distribution of land in the TORC area is shown in Figure 13. The majority of the area (46%) is covered with agricultural land.

As in the TORC area, final remediation measures should commence only after the oil layer above the groundwater table has been removed. Therefore, the first step in the remediation process involves the installation of a pump-and-treatment system at designated locations to extract the free-phase oil. Subsequently, appropriate techniques can be selected for the complete removal of pollutants from soil and groundwater. Several remediation methods are applicable to the affected area. In general, two main strategies are used for pollution removal: (1) the destruction or transformation of contaminants, and (2) the extraction or separation of contaminants from the polluted media. These strategies can be implemented either ex situ or in situ.

Economic considerations are key factors in decision-making for the remediation of contaminated areas. However, in agricultural fields and developed areas, additional factors must also be taken into account. In contrast, for forested areas and zones with residential or industrial buildings, it is evident that in situ remediation methods are the most appropriate. In the oil sludge pit area, excavation is required to enable the application of ex situ techniques. For agricultural land, replacing the contaminated soil with new fertile soil is a potential option; however, this approach involves the excavation of a large volume of soil over an extended period, which may lead to resistance from local farmers. The proposed remediation methods for each land type (Figure 13) are outlined below.

• Agricultural fields: The replacement of contaminated soil with new fertile soil seems practical. The application of any in situ methods is very difficult because of the frequent irrigation that change many remediation parameters. However, the in situ thermal method can be used only in the non-cultivation seasons.
• TORC’s forest: In situ bioremediation is the best choice. Specifically, the existence of the rhizosphere provides an appropriate condition for soil bioremediation.
• Residential and industrial buildings: A controlled and limited in situ bioremediation system can be implemented for residential areas.
• Oil sludge pit fields: In these lands, which include the oil sludge disposal and evaporation ponds, the first action is to improve the waste handling method. Possible leakages from evaporation ponds must be resolved, and further oil sludge disposal in these lands must be prevented.

In situ bioremediation is one option for the remediation process. Another approach applicable to the broader area is natural attenuation, where pollutant concentrations decrease over time due to processes such as groundwater flow-driven dispersion, natural biodegradation by indigenous soil microorganisms, and phase changes between soil, water, and air. Scientifically speaking, the contaminants do not disappear entirely, but their concentrations gradually decline. If effective remediation measures are implemented in the most heavily polluted zones, particularly the TORC and the Rey industrial area, the remaining areas may be managed through natural attenuation. It is generally assumed that the low-density contamination originates from the high-density zones. Therefore, remediation efforts in the low-density areas should focus on halting oil migration from the high-density zones and preventing any further influx of pollutants.

7. Conclusions

This study demonstrates that the oil plume in high-density pollution zones within the Rey industrial area exhibits limited mobility due to the low permeability of the soil and the high viscosity of the hydrocarbon contaminants. Simulation results highlight the localized nature of the contamination, with minimal lateral migration and attenuation primarily governed by natural processes such as dissolution and biodegradation

High-density contamination zones, particularly around TORC and REY7, are characterized by persistent hydrocarbon accumulation. In contrast, low-density zones are influenced by oil migration via surface drainage and qanat channels. These findings underscore the limited effectiveness of natural attenuation alone and highlight the need for targeted remediation measures in high-density areas to accelerate contaminant reduction.