Obtaining the hydraulic and thermal properties of the main hydrostratigraphic units surrounding a standing column well using a thermal response test

In general, classical interpretation of pumping tests assume homogeneous hydraulic properties and cannot characterize aquifer heterogeneity in the form of a profile. In this work, a thermal response test, which is routinely performed prior to the design of ground source heat pump systems, is conducted on an open-walled borehole to identify the hydraulic conductivity profile and thermal properties of the main geological units. Parameter identification and uncertainty quantification are performed in a Bayesian framework and rely on a neural network and a spectral method to speed up the solution of the direct problem. The analysis of a data set from field testing showed that the use of temperature measurements alone provided a hydraulic conductivity profile, with an average value close to that provided by a conventional pumping test. For the specific conditions observed on the field, it was also possible to precisely identify the thermal properties, as was the case for the main geologic unit, but not for the thin geologic units whose distributions were similar to the priors used for inference. This indicates that in some geologic settings, a thermal response test could fail to provide the thermal properties of some units. The approach, however, paves the way for new characterization techniques.