Influence of Groundwater on Measurements of Thermal Properties in Fractured Aquifers

Abstract

Shallow boreholes equipped with borehole heat exchangers (BHE) connected to ground-coupled heat pumps are used for the heating and cooling of buildings. Accurate estimates of ground properties are essential to properly dimension a ground-coupled heat pump system. For small projects, the ground properties are assumed for different rock types. In large projects, thermal response tests (TRT) are carried out to estimate the actual properties.

Convection and groundwater flow affect the TRT results and the operation of a ground-coupled heat pump system. This thesis investigates the influence of groundwater on measurements of thermal properties such as effective (in situ) thermal conductivity (λeff) and borehole resistance (Rb) in fractured aquifers.

A statistical analysis for selected rock types shows that λeff is in general higher than rock thermal conductivities measured from rock cores (λrock). Databases of λeff and λrock, hydraulic yield of wells, and driller’s well protocols from the Oslo region are used to test if  λeff can be predicted. It is shown that λeff cannot be predicted accurately: Heterogeneities in rock mineral content, rock types along a borehole, regional groundwater flow and convection due to different heat input rates during TRTs are too large. It is documented that a high thermal conductivity is not necessarily linked to a high quartz content, but rather to the orientation of insulating layers of low conductive materials. Thermal conductivity estimates based on λrock from mapped bedrock types can give only a vague indication about the thermal conductivity one may find at a site.

The influence of groundwater on λeff was investigated in a field experiment. Two TRTs were carried out in the same borehole: A first standard TRT and a second TRT with artificially induced groundwater flow. Temperature profiles after both TRTs showed groundwater flow in a few fractures only. The measured λeff was higher for the case of groundwater flow. The influence of groundwater flow cannot be discovered from the measured  λeff itself if the flow is restricted to limited areas of the borehole. But temperature profiles taken a few hours after a finished TRT allow a proper interpretation of the TRT results.

A multi-injection-rate TRT (MIR-TRT) showed that the measured thermal properties changed with increasing heat input rate. The required borehole length for a ground-coupled heat pump system would be reduced as buoyancy-driven convection increases in the borehole. A second MIR-TRT was carried out with a groundwater pump installed at the base of the same borehole. Groundwater was pumped up to the surface and re-infiltrated into the borehole. The estimate for required borehole lengths was reduced by 9 to 25 % in comparison to the preceding MIR-TRT. Consequently, artificial convection may be used to reduce the required borehole length of a ground-coupled heat pump system.