Uncertainty analysis of groundwater heads around underground storage caverns due to the spatial variability of hydraulic conductivity

Finite element methodology combined with random field theory is developed to overcome limitations of deterministic flow analysis around underground storage caverns. By using this combined model, the uncertainty of heads due to the spatial variability of hydraulic conductivity can be assessed. To determine the probability distribution for field data around underground caverns, various distributions are investigated. The Monte Carlo technique can be effectively applied to obtain an approximate solution to the two-dimensional steady flow of a stochastically defined non-uniform medium. A nearest-neighbour stochastic process model is used to generate a multilateral spatial dependence between hydraulic conductivity values in the block system. The uncertainty in model prediction depends on both the spatial heterogeneity of hydraulic conductivity and the nature of the flow system such as water curtains and boundary conditions. In particular, this uncertainty is related with the well-known gas tightness condition.