Uncertainties in Problem of Ground Improvement Using Prefabricated Vertical Drains (PVD) and Seeking a Cost-Effective Design Applying Monte Carlo Sampling

Abstract

This article studies a procedure of designing the prefabricated vertical drains (PVD) to attain a given total degree of consolidation, simultaneously considering the randomness of variables involved in the problem, investor’s resource (time and budget) and effects of constructor’s skill in construction on the performance of the design. The randomness of all variables and factors involved in the problem was taken into account. The law of distribution for variables and parameters was prescribed according to previous research works. There were a lot of variables including height of preloading, length of PVD installation or thickness of compressible soft soil layers to be improved, time planned by investors, drainage capacity of drains, specification of the drains, coefficients of permeability and compressibility of soil. Normal and lognormal distribution of different variables and parameters were applied suitably. The probability of successful performance for attaining a total degree of consolidation U = 85% or more was found using Monte Carlo Sampling (MCS) method with 105 runs. Some stochastic scenarios were considered to evaluate the overall cost, including item cost of preloading and total length of PVD installation. The effective design is the solution of configuration in which all the possible variations of variables were considered to meet the investor’s time and budget at the least cost.