Probabilistic method for the size design of energy piles considering the uncertainty in soil parameters

Energy piles have attracted increasing attention as profitable solutions for the utilization of shallow geothermal energy. Although various complex geotechnical design models of energy piles have been proposed, a simplified sizing method that considers the uncertainty propagation in the soil is required. In this study, a Monte Carlo Simulation-based method was proposed for the size design of energy piles considering the uncertainty of parameters in the soil, such as thermal conductivity and friction angle. The small-sample analysis method of Markov chain Monte Carlo simulation combined with the Bayesian theoretical framework was developed to generate equiv-alent samples. Subsequently, the thermal response function and thermomechanical load transfer methods were employed to address the failure probability of the energy pile in the serviceability limit state and ultimate limit state. In addition, a case study was presented to illustrate the implementation of the proposed probabilistic sizing method. The analysis results of the case study confirm the necessity of modeling the soil uncertainty in the energy pile size design.