Estimation method for layered ground thermal conductivity using genetic algorithm based on a 2-D heat transfer model

Abstract

Estimation of ground thermal conductivity plays an important role in design of borehole heat exchanger (BHE) for ground source heat pump (GSHP) systems, and the ground is usually heterogeneous with layered thermal properties. Therefore, this study presented a novel estimation method for layered ground thermal conductivity by using distributed thermal response test (DTRT) and genetic algorithm. Firstly, a comprehensive new 2-D heat transfer model for coaxial BHE was built to simulate borehole fluid temperature profiles, it revealed that the distribution of ground thermal conductivity showed a significant influence on outlet temperature and temperature profiles. Then, based on the built model, the vertical distribution of ground thermal conductivity was estimated through DTRT data by using the genetic algorithm. It was found that the DTRT data at 2 h was effective to estimate the layered ground thermal conductivity by using the new 2-D heat transfer model. Furthermore, the estimated distribution of ground thermal conductivity was almost independent of numbers or thickness of sub-layer, and the largest difference in distributions of ground thermal conductivity among different layer thickness was less than 0.3 W/(m·K). Also, the estimation convergence showed independence with layer thickness, converged results could be obtained after 10 iterations for different layer conditions.