An integrated fuzzy-stochastic modeling approach for risk assessment of groundwater contamination

Stochastic Risk Assessment of Groundwater Contamination under Uncertainty: A Canadian Case Study

“…The related procedures can be summarized as follows (Li et al, 2007): (a) generate random numbers between zero and one for each stochastic input parameter; (b) transform the random numbers to the corresponding random variates based on the specified probabilistic distribution for each parameter; (c) store the generated random variates into an array for each parameter; (d) obtain a value from the array for each parameter and set it as the deterministic input parameter in the multiphase multi-component model; (e) compute the contaminant concentrations with the multiphase compositional numerical model for each Monte Carlo run; (f) store the resulting outputs of the contaminant concentrations for each Monte Carlo run for further statistical analysis; (g) repeat steps (a) to (f) for a number of times (specified number of Monte Carlo runs); (h) stop computation of contaminant concentrations when the specified number of Monte Carlo runs has been done, and exit to step (i), and (i) analyze the contaminant concentrations and compute the statistic moments (i.e., peak concentration at the domain for each run, mean and standard deviation of the concentrations over the specified number of times of Monte Carlo runs).…”

“…However, 0.3 mg/L of xylene concentration in groundwater is corresponding to a HI of 0.041 (when using parameter values of IR = 2 L/day, EF = 350 days/ year, ED = 30 years, BW = 70 kg) by applying the RfD value of 0.2 mg/kg·d as recommended by USEPA (2003), which is considered to be an acceptable level of risk since the HI is lower than 1.0. In order to combine the concepts of both approaches, the risk is characterized into the following three levels in this study: (a) slightly risky with HI in the range of 0.003 to 0.041, where HI of 0.003 is corresponding to a xylene concentration of 0.02 mg/L which is the strictest guideline (Li et al, 2007); (b) risky with HI in the range of 0.041 to 1.0, and (c) highly risky with HI greater than 1.0. The contaminant concentration denoted as CW in Equation (1) will be a stochastic variable due to the uncertainties in the input parameters of the multiphase compositional simulator, leading to uncertainties in the calculated HI as described by Equation (2).…”

“…Lee et al (2002) applied the USEPA guidelines and the empirically measured contaminant levels and exposure parameters to estimate health risk in a groundwatercontaminated community after on-site remediation, by using Monte Carlo simulations to account for uncertainties. Li et al (2006Li et al ( , 2007 proposed a fuzzy-stochastic risk assessment approach for examining uncertainties associated with both source/media conditions and evaluation criteria in the contaminated site management system. However, many of the previous risk assessment studies were either based on simple numerical models or for short-term only.…”

“…Such simplification cannot sufficiently characterize the complexities especially existing at petroleum-contaminated sites. In addition, there are only few studies on integrating advanced numerical subsurface models, such as multi-phase multi-component contaminant transport and fate models, within an uncertainty-based risk assessment framework (Li et al, 2007). This kind of integration, however, can effectively facilitate the assessment of long-term risks at hydrocarbon-contaminated sites.…”

Applying Fuzzy Logic to Assess Human Perception in Relation to Conservation Plan Efficiency Measures Within a Biosphere Reserve

“…Fuzzy logic (FL) is a mathematical expert system that has been used in mixed, imprecise, and uncertain environments, dealing with uncertainties expressed as interval values and fuzzy sets and its approach is based upon linguistic expressions involving input and output variables rather than numerical probabilistic, statistical, or perturbation variables (Li and Huang 2010). It has commonly been used in informatics and robotic systems and has lately become a popular tool in environmental impact studies (Bojorquez-Tapia et al 2002;Li et al 2007;Gagliardi et al 2007) and ecological and risk assessments (Adriaenssens et al 2004;Salihoglu and Karaer 2004;Gevrey et al 2006), as well as for the evaluation of tropical forest conditions (Ochoa-Gaona et al 2010) and water resources management (Guo et al 2010) having the advantage of rendering subjective and implicit decisionmaking more objective and analytical, with its ability to accommodate both quantitative and qualitative data (Ekmekçioglu et al 2010).…”