Improving computational efficiency of numerical modelling of horizontal ground source heat pump systems for accommodating complex and realistic atmospheric processes

Modelling horizontal ground loops for a horizontal ground source heat pump (HGSHP) system is complex and computationally expensive. The computation precision is highly reliant on the prescription of an undisturbed ground temperature in the unsaturated ground as well as realistic and accurate atmospheric processes at the ground surface boundary. Conventionally, modelling of such a system would include direct application of the atmospheric processes at the soil-atmosphere boundary and solve it in a single-stage approach. However, low efficiency is found for large spatial domain and long-term transient problems as the boundary processes need to be solved and expressed in terms of primary model variables at each simulation time-step. This paper proposes an equivalent two-stage modelling approach, for the first time, based on an advanced coupled thermal-hydraulic (TH) model to improve computation efficiency while maintaining adequate accuracy. In this approach, firstly, the model is solved for an intact ground that is imposed by complex atmospheric processes, e.g., rainfall, solar radiation, humidity, evaporation, etc. at the soil-atmosphere boundary, and the spatial and temporal variations of the primary model variables are recorded. Afterwards, the recorded data are incorporated in the simulator, as model inputs, for the same ground including a HGSHP system. Predicted results from both 2D and 3D simulations show that the ground temperatures calculated by the proposed two-stage approach are in good agreement with that of the traditional single-stage approach. However, the two-stage approach is computationally robust. For the presented 2D and 3D simulations, it required only 32% and 37% of the time of the single-stage approach, respectively, while maintaining great accuracy. This demonstrates the utility of the proposed two-stage approach for modelling complex scenarios of realistic HGSHP systems installed in a large spatial domain and for long-term operation.