Probabilistic analysis of pollutant migration from a landfill using stochastic finite element method

Here a probabilistic study is attempted to analyze the pollutant migration in the soil liner while assuming the parameters of the soil liner viz hydraulic conductivity, dispersion coefficient, porosity, and sorption coefficient as random fields. For this purpose a stochastic finite element method based on the perturbation theory is developed and applied to a one-dimensional solute transport problem. For the stochastic case the peak of mean concentration is higher than that of the deterministic case and arrives early. The effects of parameter correlation and correlation length on the probabilistic behavior of concentration at the bottom of the soil liner are presented. For the case wherein sorption coefficient is negatively correlated with the hydraulic conductivity, dispersion coefficient, and porosity, the probability of concentration exceeding a threshold value is found to be the highest and the corresponding peak value of this probability occurs at an early time. The correlation length of the random field is found to be an important parameter and for a large correlation length the probability is higher and approaches to an asymptotic value. The parametric study shows that the probability of exceeding the threshold concentration is found to be sensitive to the variations in the key parameters.