An efficient grouping strategy for modeling thermal interactions in large borefields
Accurately modeling thermal interactions between ground heat exchangers (GHEs) is essential for the design and simulation of geothermal heat pump (GHP) systems, particularly in large and irregular borefields. However, the computational cost of evaluating these interactions increases rapidly with the number of boreholes, often becoming a limiting factor. This work introduces a novel, efficient, method to alleviate this burden through a similarity-based grouping approach for g-function construction. The proposed method identifies groups of boreholes exhibiting similar thermal behavior—characterized by comparable heat extraction rates—by solving a single linear system derived from the superposition principle. Each group is then represented in the model by a single representative borehole, significantly reducing the problem complexity while maintaining high accuracy. In contrast to previously proposed aggregation or clustering methods that depend on spatial proximity or geometric symmetry, this approach directly leverages thermal behavior and avoids iterative refinement or convergence steps, resulting in a computationally robust and scalable solution. In a case study with 1,958 randomly positioned GHEs, the grouped model produces a g-function with negligible deviation from the full-resolution solution, while reducing computation time to just 0.8 seconds on a standard workstation laptop.