Sensitivity analysis and validation of a numerical thermal response test of soil using in-situ data

This article presents the development, validation, and sensitivity analysis of a three-dimensional numerical model of a thermal response test (TRT) for a ground heat exchanger, implemented in ANSYS Fluent and based on in-situ measurement data. Numerical TRT simulations were carried out to investigate the influence of the turbulence model, time step, and number of solver iterations on simulation results. The model's response was evaluated for varying thermal conductivity values of bentonite and the U-pipe material. It was shown that these parameters significantly affect the glycol temperature during the initial phase of the test. The impact of the soil thermal conductivity, defined as a boundary condition, was also analyzed. In the final stage of the study, this value was set to 2.15 W/(m & sdot;K) in the numerical solver, resulting in strong agreement between simulation results and field measurements, with an average relative error of only 0.67% for the mean fluid temperature. The effective thermal conductivity obtained from the numerical model (2.32 W/(m & sdot;K)) differed by only 0.4% from the experimental value (2.33 W/(m & sdot;K)). It was found that analysis of TRT data using the Infinite Line Source (ILS) model led to a different estimation of soil thermal conductivity than the value defined in the numerical solver, overestimating it by up to 9.3% in this case. The findings confirm the applicability of numerical modeling in supporting TRT development and the design of geothermal systems. The novelty of this study lies in the detailed presentation of the model quality assessment procedure, its implementation using real test data, and the discussion on the accuracy of the ILS model when used for TRT data interpretation.