Estimation of thermal properties of soil and backfilling material from thermal response tests (TRTs) for exploiting shallow geothermal energy: Sensitivity, identifiability, and uncertainty

Abstract It is critical for designing shallow geothermal systems to accurately estimate thermal properties of soil and backfilling material. This paper reports on a new algorithm for simultaneously estimating thermal conductivities and diffusivities of soil and backfilling material. The algorithm uses a 2D composite-medium line-source solution for borehole ground heat exchangers, enabling early-time ( 10 hr) data of thermal response tests (TRTs) to be used in the parameter estimation. The new algorithm is validated by a reference sandbox experimental data. Several theoretical issues of parameter estimation were explored, including sensitivity, identifiability, and uncertainty. The key findings of this study are: (1) the early-time data of TRTs can greatly reduce the linear dependence of the estimated parameters and thus improve the identifiability of the parameter estimation. (2) The accuracy of the estimates, ranked in descending order are as follows: soil thermal conductivity k s ( ± 2 %), grout thermal diffusivity a b and conductivity k b ( ± 15%), soil thermal diffusivity a s ( ± 60%). (3) ± 10 % uncertainties in borehole radius, heating rate, and half-spacing of U-tube legs lead to the total deviations from the original estimates of k s , a s , k b , or a b equal to 8.4%, 87.5%, 12.8%, and 15.5%, respectively.