Physical modeling of coupled water and heat flow within a borehole heat exchanger array in the vadose zone

This thesis is focused on characterization of the heat transfer and water flow processes in a physical model of a borehole array in an unsaturated soil layer. The overall goal is to develop a dataset for validation of coupled thermo-hydraulic flow models used for simulating the efficiency of heat injection or extraction from soil-borehole thermal energy storage systems. The physical model consists of a layer of unsaturated silt compacted atop a layer of sand within an insulated, 0.53 m-tall, 0.6 m-diameter cylindrical tank. A water table was imposed at the top of the sand layer. Three steel "U"-tube pipes were inserted through the silt layer into the top of the sand layer to represent a triangular array of geothermal borehole heat exchangers, and several tests were performed with different heat exchanger spacings. Heated fluid was circulated through the steel pipes to inject heat into the unsaturated silt layer at a constant rate, during which time changes in volumetric water content and temperature were measured at different depths along the center of the silt layer using dielectric sensors. The thermal conductivity and specific heat of the silt were also monitored using a thermal probe at the center of the soil layer at mid-height. The temperature of the silt at different distances from the heat exchangers, the inlet and outlet temperatures of the fluid, as well as the temperature and relative humidity of the air at the soil surface were monitored. Regardless of the heat exchanger spacing, the temperature of the unsaturated silt layer was observed to increase to a relatively steady-state value after a short period of time. Dielectric sensor measurements initially show an increase in water content at all depths in the soil, indicating that that water is moving away from the heat exchangers, albeit at a slower rate than the heat flow process. Further, water was observed to condense at the soil surface, indicating that water vapor moved upward though the soil layer due to buoyancy. In the test with the smallest radial heat exchanger spacing of 80 mm, after the initial increase in water content, a sharp decrease in water content was observed. This indicates that water was driven from the center of the array into the surrounding soil, and that a convective cycle of water phase change did not occur for this small heat exchanger spacing. In the test with the greatest radial heat exchanger spacing of 300 mm, the soil in the center of the array did not experience a decrease in water content after the initial increase, which may indicate that a convective cycle was formed. In the case of the test with an intermediate heat exchanger spacing of 160 mm, an intermediate behavior was observed. In the tests, downward liquid water flow due to gravity could not be confirmed through evaluation of the water content data. This phenomenon, which is expected if a convective cycle of water phase change is observed in the array, may have been observed for longer testing times. The thermal conductivity measured using the thermal probe was observed to increase significantly during the heating process, and was observed to be a function of both the degree of saturation of the soil and the temperature. In the case of the smallest radial borehole heat exchanger spacing, the thermal conductivity of the soil inside of the array was observed to decrease as the water content decreased, while the thermal conductivity outside of the heat exchanger array was observed to increase with increasing radial location. This confirms that the borehole spacing can have an important effect on the long-term heat storage in the vadose zone, and that small spacings may lead to an increase in thermal energy transfer to the soil outside of the array. In the case of the largest radial borehole heat exchanger spacing, a stable increase in thermal conductivity of the soil within the array was observed. As expected, the heat storage in the unsaturated soil within the array increased with the spacing of the heat exchangers. However, the increase in water content in the array with the widest spacing also likely contributed to the increase in heat storage. A recommendation from this study is to use a geotechnical centrifuge to better simulate the role of capillary rise and downward liquid water flow on the coupled heat transfer and water flow process within a small-scale physical model. These features may be critical in establishing a convective cycle, which may enhance the efficiency of heat injection into a soil-borehole thermal energy storage system.