«Signature: _ April 13, 2012 Vijay Selvan Ram A Deterministic and Probabilistic Analyses of the Carbon Tetrachloride Contaminant Plume in Groundwater ...»
33 organic compounds, of which 64% were reliably identified, were found at the former UCIL formulation plant. Among the compounds identified were four isomers (alpha-, beta-, delta- and gamma-) of hexachlorocyclohexane (HCH), which were all
products in the production process at UCIL. 400 meters north of the main UCIL site are the Solar Evaporation Ponds (SEPs), which collect waste after they have been neutralized and treated, and with the exception of mercury, the concentrations of other heavy metals were below background. An organic analysis found 15 compounds of which approximately 20% were identified; yet only a few organic compounds were identified reliably when compared with other sites (Labunska et al. 1999).
They collected groundwater samples from twelve drinking water wells both to the north and south of the UCIL site (Figure 3) and detected high levels of chloroform, carbon tetrachloride and chlorinated benzenes in wells close to the northern boundary of the plant (Labunska et al. 1999). A later study carried out in 2004 also found similarly high levels from a drinking water hand-pump located near the southern boundary of the plant (Labunska et al. 2004).
Pollution Monitoring Laboratory (PML) In 2009, scientists from the Pollution and Monitoring Laboratory (PML), in association with the Centre for Science and Environment (CSE) collected soil and water samples from inside and outside the UCIL factory. These samples included water samples collected from handpumps, borewells and dugwells from residential areas around the UCIL plant, and from as far as 3.5 km away. Waste samples from waste storage sheds, soil samples from within the factory premises, the SEPs, and surface water samples show contamination of HCH isomers and heavy metals. They found results similar to Greenpeace, with groundwater samples being heavily contaminated with chlorinated benzene compounds and organochlorine pesticides (Johnson et al., 2009).
NEERI, an institute devoted to the intersection of environmental concerns and human public health, performed three rounds of sampling and geophysical investigations on UCIL premises to determine the extent of contamination in both soil and groundwater.
Soil sampling within UCIL premises, outside the UCIL premises, and groundwater sampling within and outside UCIL premises reveal a contamination profile that is less alarming than the two previously described studies. The geophysical investigations reveal the possibility of chemical contamination at three sites (Site I, Site III and Site V) (Figure 4). From the analysis of groundwater sampled from borewells within the UCIL premises and outside, pesticide contamination was found in 5 wells in the vicinity of UCIL premises along a Northeast axis (NEERI 2010).
Study Objectives This study will attempt to characterize the contaminant plume for CCl 4 in groundwater and create future projections for the contaminant plume at sites along the axis of the direction of groundwater flow. More specifically, the aims for the study will be to 1) characterize the hydrogeology of the site as reported in the literature 2) estimate the source concentrations of CCl 4 3) run a deterministic model using single-point estimates to project the contaminant plume out in space and time, and 4) run a probabilistic model to project a range of likelihoods that the concentration of CCl 4 will exceed drinking water standards set by the World Health Organization (WHO) and the United States Environmental Protection Agency (US EPA).
As Labunska et al. report in their 1999 Greenpeace report, the groundwater under the site flows in the northeast direction. There is an abundance of human settlements
groundwater for those populations that draw their drinking water from this source. The studies outlined above indicate a lengthy profile of contamination of both the soil and groundwater, which includes a number of pesticides and their production intermediates.
However, for the purposes of this study, CCl 4 will help to elucidate some measure of exposure-risk for those residents who drink from this groundwater.
CCl 4, a volatile organic compound (VOC) was chosen because of its reported distribution in the groundwater along the northeast axis. According to Labunska et al., CCl 4 is found in high concentrations in wells along the northern plant boundary, while those concentrations in wells along the southeast corner of the site are lower, though still significantly elevated (Labunska et al. 1999). In addition, the toxicological profile of carbon tetrachloride is well characterized, providing a good justification for assessing their risk.
In order to evaluate the behavior of CCl 4 and its’ associated levels in the groundwater, the software Analytical Contaminant Transport System (ACTS) will be used to study the steady state and time dependent behavior of CCl 4. The ACTS software utilizes environmental models commonly found in the literature for analytical purposes.
It can be used as a screening-level tool for environmental managers and policy-makers to obtain some general knowledge of the system and its behavior. In order to utilize ACTS, hydrogeologic parameters and contaminant properties must be measured, in addition to characterizing the aquifer and contaminant plume. Because this study relies on values from the literature regarding Bhopal and the UCIL site, where uncertainty exists or
methods, producing a range of possible outputs (e.g. a range of contaminant levels at a specified distance after a specified period of time) (Anderson et al., 2010).
Chemical and Contaminant Properties of Carbon Tetrachloride Chronic exposure to CCl 4 can have an adverse effect on liver functioning.
Breakdown products have been shown to attack cell proteins, and in severe cases liver cells can be destroyed leading to a decrease in liver function. If exposure is short-term, these effects can be reversed. Kidneys are also sensitive to CCl 4. It can cause a buildup of water and waste products in the body with potentially fatal consequences if the exposure is too high. There are also neurological symptoms associated with high exposure to CCl 4.
In the most severe cases, coma and permanent damage to nerve cells can result. Animal studies by the Department of Health and Human Services (DHHS) have shown CCl 4 to be possibly carcinogenic. IARC classifies CCl 4 in Group 2B (possibly carcinogenic to humans), while the EPA classifies CCl 4 as a probable human carcinogen (ATSDR 2005).
The World Health Organization (WHO) has set maximum contaminant level guidelines for drinking water for a range of chemical contaminants. The guideline for carbon tetrachloride is 0.004mg/L (WHO 2011). The EPA has set 0.005 mg/L for carbon tetrachloride as the maximum contaminant level (EPA 2011).
Modeling Approach and Assumptions The aquifer under the UCIL pesticide manufacturing facility and the surrounding area was characterized within ACTS as a confined, two-dimensional, saturated, infinite model with constant dispersion coefficients. In order to define the modeling domain, the
Southwest to Northeast, with a maximum length of 2 kilometers, and the y-axis centered on the UCIL facility, extended to include the whole facility with a width of roughly 500
meters. The following assumptions were made for modeling purposes:
1. The contaminant plume will be characterized as flowing in the reported direction of groundwater flow (Labunska 1999, NEERI 2010), with the location of the source remaining constant. Groundwater flow is assumed to be steady and uniform.
2. The surficial aquifer is characterized as infinite, meaning the boundaries of the aquifer are assumed to be far enough away that they do not affect the dispersion coefficient.
3. Because of the limited amount of data regarding soil and groundwater properties specific to the site, certain hydrogeologic parameters (porosity, bulk density of soil, etc) were estimated based on their typical values as reported in the literature.
4. It is assumed that the concentration of CCl 4 is constant and no source remediation has been performed. This is due to the high level concentrations of the contaminants found at the site approximately 15 years after the closing of the site (Labunska et al. 1999) Model Input Parameters and Source Definition for Deterministic Simulations Parameters pertaining to groundwater flow and contaminant characteristics are obtained from a variety of sources including relevant literature articles, existing site field data and measurements taken remotely through Google Earth (Table 1). All of the values for the deterministic analysis are shown in Table 2. The model was calibrated by manual adjustment of input parameters, which most closely approximated the measured concentrations found in field measurements generally located along the axis of
In order to estimate the source concentration of CCl 4, the model was calibrated by varying the concentration at the inflow boundary, assuming the source originated in 1984.
By varying the concentration at the inflow boundary, and adjusting the field and chemical constants, it was determined that 4,000mg/L most closely approximated the concentration measured at the monitoring well used for calibration (Labunska et al. 1999). 4,000mg/L is approximately 5% of the solubility limit of CCl4.
Following calibration, a deterministic scenario was simulated using a constant source scenario under the assumption that there has been no source removal or remediation of the site.
Model Input Parameters for Probabilistic Simulations In order to conduct probabilistic analyses of contaminant transport, two-stage Monte Carlo simulation was performed by generating probability density functions (PDFs) for three input parameters. In the first stage, PDFs were generated for each parameter that had a degree of uncertainty, in other words, the values were recorded in the literature or cited in field measurements as ranges. 10,000 realizations were simulated to characterize the variability in the parameters. In the second stage, another 10,000 realizations of the model were run within ACTS using the inputs of the PDFs for each parameter, thereby creating a range of probabilities of contaminant concentrations in the groundwater for specified times and distances.
The two parameters chosen for two-stage Monte Carlo simulations are the Darcy groundwater velocity and the retardation coefficient. The Darcy groundwater velocity underwent Monte Carlo simulations by varying proxy parameters that are required to
parameters and their distributions are presented in Table 3. Probabilistic simulations will be run for three points of interest, 500m, 1000m and 2000m northeast of the center of the UCIL site for three different time periods. These areas represent dense population centers.
Calibration Results In order to determine the parameters that best estimates measured concentrations of carbon tetrachloride, the model was calibrated by varying the input parameters that achieved results that most closely matched concentrations found in a well roughly 300 meters northeast of the center of the production facility. The concentration at the inflow boundary that most closely matched the concentration measured at IT9030 and IT9032 was 4,000mg/L (Table 4, Figure 3).
Deterministic Model Simulations After having calibrated a set of parameters to appropriately model the system, simulations were conducted using single-value and distributed-value inputs to model the contaminant plume at a maximum distance of 2000 m with different time end-points.
Given the initial source concentration, the deterministic model output shows a normalized concentration of 3.80 x 10-22, which is 1.52 x 10-15mg/L at a location 2000 meters from the source concentration in 2084. To put the results in context, the outcome of interest was the location of the plume, which approached the U.S. EPA MCL, which is represented in the figure as a normalized concentration of 1.25 x 10-6 (Figure 5a-c).
The model was run to simulate the plume at 10, 30 and 50 years since the closing
of the level that represents the EPA level MCL is at a distance of approximately 850 meters from the source. Given a constant source scenario, after 10, 30 and 50 years, distances beyond roughly 850m are not projected to contain concentrations of CCl 4 above 0.005mg/L.
Probabilistic Model Simulations The results for the probabilistic scenario include two histograms showing the probabilistic distributions for the two input parameters that underwent Monte Carlo simulations. Descriptive statistics of each of those input parameters are presented in Table 3 and Figures 6a – 6b.
Three different locations (500m, 1000m and 2000m) were chosen because of their central location in densely populated areas to the northeast of the site. The results of probabilistic simulations for each location show the likelihood of exceeding certain concentrations, particularly the WHO MCL of 0.004mg/L (C/C 0 = 1x10-6) at specified points in space and time. For all simulations (500m, 1000m and 2000m), there is a 100% probability that the concentration of CCl4 exceeds the WHO MCL. For concentrations higher than this, the probabilities drop (Figures 6 – 9).
Contaminant Plume The EPA and WHO have both set drinking water standards for CCl 4, 0.005mg/L and 0.004mg/L, respectively. The model simulations were run in order to determine at what locations and times this concentration is exceeded, and as a result, pose a health risk for the populations exposed to the groundwater. Deterministic simulations show that
be exposed to concentrations of 0.005mg/L. Deterministic simulations show reduced concentrations for populations living further out than 850 meters.
In the probabilistic model simulations, for all distances and times simulated, there is a 100% probability that the concentration of CCl 4 will exceed the WHO MCL of