Reliability analysis of hydrologic containment of underground storage of liquefied petroleum gas

The objectives of this paper are (1) to introduce a numerical first order method to calculate the gas containment reliability within a heterogeneous, two-dimensional, unlined rock caverns (URCs), and (2) to suggest a strategy for improving the gas containment reliability. In order to achieve these goals, we first analyzed the spatial variability of saturated hydraulic conductivity (K_s) at a field site. We then conducted deterministic simulations to demonstrate the importance of heterogeneity of K_s in the analysis of gas tightness performance of URCs. Considering the uncertainty of the heterogeneity in the real world situations, we subsequently developed a numerical first order method (NFOM) to determine the gas tightness reliability at crucial locations of URCs. Using the NFOM, the effect of spatial variability of K_s on gas tightness reliability was investigated. Results show that as variance or spatial structure anisotropy of K_s increases, most of the gas tightness reliability at crucial locations reduces. Meanwhile, we compare the results of NFOM with those of Monte Carlo simulation, and we find the accuracy of NFOM is mainly affected by the magnitude of the variance of K_s. At last, for improving gas containment reliability at crucial locations at this study site, we suggest that vertical water-curtain holes should be installed in the pillar rather than increasing density of horizontal water-curtain boreholes.