Design of Adapted Thermal Response Tests for Retaining-Wall Energy Geo-Structures
Thermal Response Tests (TRTs) are widely used for conventional borehole heat exchangers to estimate ground thermal properties and support geothermal system design. However, their direct application to some specific energy geo-structure elements is not straightforward, such as retaining walls, foundation slabs, diaphragm walls, and basement walls have finite geometries, shallow installation depths, non-radial heat transfer, and boundary conditions that differ from those of vertical boreholes. For the retaining-wall energy geo-structures, the embedded pipe length, wall dimensions, wall surface area, concrete properties, surrounding ground, and interaction with adjacent spaces can all affect the measured thermal response. Therefore, an adapted TRT interpretation is required to derive practical design parameters rather than relying only on conventional borehole-based thermal conductivity estimates. This work proposes a COMSOL Multiphysics framework for the design and interpretation of adapted TRTs for retaining-wall energy geo-structures. The objective is to estimate effective heat-exchange parameters, including heat exchange capacity per embedded pipe length, heat exchange capacity per active wall area, equivalent fluid–pipe–concrete–ground thermal resistance, and effective thermal interaction with the surrounding ground. These parameters are intended to support practical design and comparison of different energy geo-structure layouts. The numerical model consists of a retaining wall with an embedded HDPE pipe network, the surrounding ground, reinforced concrete domains, and the circulating fluid. Heat transfer in the ground and concrete is modelled using the Heat Transfer in Solids interface, where transient conduction governs the thermal response of the solid domains. The embedded pipe circuit is represented using the Heat Transfer in Pipes interface as a simplified one-dimensional formulation, which accounts for axial advection, fluid heat capacity, internal convection, pipe-wall thermal resistance, and heat exchange with the surrounding solid domains. This 1D pipe representation allows the full embedded pipe network to be included without the high computational cost of resolving the pipe geometry in three dimensions. The adapted TRT is simulated by applying a controlled heat injection rate to the circulating fluid for a prescribed test duration, followed by a thermal recovery period. The response is evaluated through inlet and outlet fluid temperatures, average wall temperature, local wall/ground temperature probes, and stored or dissipated heat. The influence of heat input level, test duration, embedded pipe length, pipe spacing, wall dimensions, and boundary conditions is investigated. Externally exposed ground surfaces are represented using convective and solar-radiative boundary conditions, while embedded interfaces are coupled by heat-flux continuity. Additional insulated-boundary cases are considered to quantify the sensitivity of the TRT interpretation to building-side or parking-side thermal interaction. The results clarify how retaining-wall geometry and pipe layout influence the apparent TRT response and the derived effective parameters. The COMSOL-generated simulations will also form a structured synthetic dataset for future Deep Neural Network-assisted interpretation, enabling rapid estimation of effective thermal parameters and improved selection of test duration, heat input, and monitoring locations for different energy geo-structure configurations.
