Reliability analysis based selection procedure for granular filters with respect to soil boiling condition

Abstract

Existing granular filter design guidelines are based on the particle size distribution of the base soil and filter material. Though well established, the approach does not provide a means to assess the performance of the existing filters as well as the levels of associated risk with due consideration of variability in soil properties and representative particle sizes. Previously, authors proposed closed form analytical solutions to obtain factor of safety of the base soil–filter material system with respect to soil retention and permeability criteria. The solutions were developed for soil boiling condition that takes into account relevant soil properties such as unit weight, porosity, permeability, particle sizes, angle of internal friction, etc., as well as, existing hydraulic conditions and seepage discharge. In the present work, the proposed analytical solutions are utilized for the reliability analysis of the base soil–filter media system. The variability in the input soil properties is characterized as log normally distributed continuous random variables and combination of Monte Carlo simulations and first order reliability method are utilized for the reliability analysis. Through example applications, it is demonstrated that the reliability based approach is appropriate for considering variability in mathematical framework and for assessing the performance or choice of requisite filter material that has to satisfy the two critical criteria, i.e. soil retention and permeability.