Abstract

For new building and infrastructure developments, it is possible to incorporate the mechanism for heat transfer between the building and the ground through the foundation elements (e.g. piles and diaphragm walls). This geothermal underground infrastructure approach is considered as potentially cost-effective due to small additional installation cost and less occupied land/underground space. For its implementation, it is of significance to discuss the impact to the ground source heat pump (GSHP) performance from the foundation itself as well as other underground structures in the planning period. This study describes a case of a combined thermal pile and wall system installed at a recently constructed underground station box. The following three potential impacts emerged while investigating the case study; (1) a nearby underground tunnel which has been in operation for more than 100 years, and (2) the station box and new tunnel itself when the commercial service begins. Finite element simulations were conducted to investigate the magnitude of the issues. Results show that, the heat generated from both nearby underground tunnel and the station box itself can influence the performance of GSHP . In this case study, the amount of heat extracted increased several times of the original design when the heat supply from the existing underground infrastructure is considered. At the same time the overall changes in the ground temperature after many years of GSHP operation is discussed. In the other part, it was doubted that the conductivity value of the soil estimated by classic thermal response test (TRT) methodology is not proper in the pile/wall. A detailed computational fluid dynamics (CFD) model was conducted to illustrate this inaccuracy. Results shows: (1) the classic TRT results should be over-evaluated due to the heat transfer co-efficient across the pipe walls. (2) the different heat transfer co-efficient due to the friction to the liquid at the wall interface, and (3) water flow rate is a key in the difference of TRT results. An acceptable compensatory method is to apply velocity-related heat transfer co-efficient calculation in linear-sourced model to simulate the heat transfer process.