U-pipe Borehole Heat Exchangers (BHE) are widely used today in ground source heating and cooling systems in spite of their less than optimal performance. This thesis provides a better understanding on the function of U-pipe BHEs and Investigates alternative methods to reduce the temperature difference between the circulating fluid and the borehole wall, including one thermosyphon and three different types of coaxial BHEs.

Field tests are performed using distributed temperature measurements along U-pipe and coaxial heat exchangers installed in groundwater filled boreholes. The measurements are carried out during heat injection thermal response tests and during short heat extraction periods using heat pumps. Temperatures are measured inside the secondary fluid path, in the groundwater, and at the borehole wall. These type of temperature measurements were until now missing.

A new method for testing borehole heat exchangers, Distributed Thermal Response Test (DTRT), has been proposed and demonstrated in U-pipe, pipe-in-pipe, and multi-pipe BHE designs. The method allows the quantification of the BHE performance at a local level.

The operation of a U-pipe thermosyphon BHE consisting of an insulated down-comer and a larger riser pipe using CO₂ as a secondary fluid has been demonstrated in a groundwater filled borehole, 70 m deep. It was found that the CO₂ may be sub-cooled at the bottom and that it flows upwards through the riser in liquid state until about 30 m depth, where it starts to evaporate.

Various power levels and different volumetric flow rates have been imposed to the tested BHEs and used to calculate local ground thermal conductivities and thermal resistances. The local ground thermal conductivities, preferably evaluated at thermal recovery conditions during DTRTs, were found to vary with depth. Local and effective borehole thermal resistances in most heat exchangers have been calculated, and their differences have been discussed in an effort to suggest better methods for interpretation of data from field tests.

Large thermal shunt flow between down- and up-going flow channels was identified in all heat exchanger types, particularly at low volumetric flow rates, except in a multi-pipe BHE having an insulated central pipe where the thermal contact between down- and up-coming fluid was almost eliminated.

At relatively high volumetric flow rates, U-pipe BHEs show a nearly even distribution of the heat transfer between the ground and the secondary fluid along the depth. The same applies to all coaxial BHEs as long as the flow travels downwards through the central pipe. In the opposite flow direction, an uneven power distribution was measured in multi-chamber and multi-pipe BHEs.

Pipe-in-pipe and multi-pipe coaxial heat exchangers show significantly lower local borehole resistances than U-pipes, ranging in average between 0.015 and 0.040 Km/W. These heat exchangers can significantly decrease the temperature difference between the secondary fluid and the ground and may allow the use of plain water as secondary fluid, an alternative to typical antifreeze aqueous solutions. The latter was demonstrated in a pipe-in-pipe BHE having an effective resistance of about 0.030 Km/W.

Forced convection in the groundwater achieved by injecting nitrogen bubbles was found to reduce the local thermal resistance in U-pipe BHEs by about 30% during heat injection conditions. The temperatures inside the groundwater are homogenized while injecting the N₂, and no radial temperature gradients are then identified. The fluid to groundwater thermal resistance during forced convection was measured to be 0.036 Km/W. This resistance varied between this value and 0.072 Km/W during natural convection conditions in the groundwater, being highest during heat pump operation at temperatures close to the water density maximum.