How reliable are standard thermal response tests? An assessment based on long-term thermal response tests under different operational conditions

Abstract

In this contribution, we analyze the results of a number of thermal response test (TRT) experiments performed during several years at the same location at our university campus in Valencia (Spain), a permeable saturated soil area with possible groundwater flow conditions. A combination of different heat injection rates, TRT operation times of up to 32 days, and various methods for parameter estimation of ground thermal properties have been applied to study their influence on the result and accuracy of TRTs. Our main objective has been to experimentally quantify the influence of groundwater flow heat advection using moving infinite and finite line-source theories, as well as to analyze the influence of factors such as test duration, sensor accuracy, and external thermal influences. We have shown that the traditionally used infinite and finite line-source models, as well as the moving line-source models, can accurately represent experimental temperature evolution, but that there are many caveats regarding the significance parameters extracted and its reproducibility and stability. These features can be improved if data from the first test days are disregarded for the analysis, obtaining a much faster convergence to the definitive soil parameter estimates, including the effective Péclet number that represents groundwater flow in our particular case.