The role of ground hydrothermal spatial variability on energy pile group thermal performance

Shallow geothermal energy pile systems are a novel, low-carbon technology to heat and cool buildings via heat exchange with the ground. Hydrothermal ground characteristics (e.g., thermal conductivity and permeability) are critical parameters in system optimisation. However, the effect of their inherent spatial variability is mostly unexplored, and system design is commonly performed assuming uniform parameter values. This study investigates the influence of spatial hydrothermal heterogeneity on the thermal performance of an energy pile group. A detailed three-dimensional hydrothermal finite element model is developed and validated to solve for transient diffusion-convection heat transfer for an energy pile group under seepage. Lognormal autocorrelated random fields of the hydrothermal properties are generated and implemented in the model as a series of Monte Carlo Simulations. Results show that the homogeneous hydrothermal assumption is appropriate when the heterogeneity coefficient of variation (COV) is lower than 1.0. However, the uncertainty in energy provision becomes considerable, exceeding 18 %, for ground variability values of COV > 2.0. Moreover, neglecting spatial variability in ground permeability is more detrimental to thermal performance prediction than in ground thermal conductivity, particularly in the vertical direction. This is exacerbated at low values of seepage velocity, with changes in the mean energy provided peaking 33 % at a value of 0.3 m/day, indicating the need for greater accuracy in hydraulic site investigations. Further, the uncertainty in cyclic concrete-pile temperature over a full thermal cycle increases by up to 5( degrees)C compared to that from utilising the homogeneous assumption.