Stepwise multi-objective framework for decoupling multi-parameter estimation of ground heat exchangers

Abstract

It is a great challenge to estimate multiple parameters of ground heat exchangers (GHEs) from a thermal response test (TRT) with high reliability. This study proposed a stepwise framework involving the lexicographic multi-objective optimization method to increase the identifiability of multi-parameter estimation of GHEs. Three ideas constitute the foundation of the framework: 1) using a temperature derivative objective-function to complement the temperature objective function, 2) using the lexicographic method for solving the involved multi-objective minimization problem, 3) using an analytical short-time solution for addressing the time-varying features of parameter sensitivity. This framework was illustrated by two variant algorithms for five-parameter estimation (four thermal properties plus borehole thermal resistance). The two variant algorithms were validated by an independent sandbox experiment, and the results show that these algorithms yield estimates for borehole thermal resistance, soil thermal conductivity, grout thermal diffusivity, grout thermal conductivity, and soil thermal diffusivity with relative errors smaller than 2.0 %, 4.0 %, 14 %, 20 %, and 36 %, respectively.