A correlation-based model for building ground-coupled heat loss calculation using Artificial Neural Network techniques

Heat loss from a building through the ground becomes relatively important as the above-grade components of the building fabric are built more energy efficient. For annual space heating and cooling load calculations that require short computation time, with reasonable accuracy, for modelling heat losses to ground through the foundations, this study presents a new correlation-based model. The correlation algorithms were derived from the results of a commercial 3D transient heat transfer program using Artificial Neural Network. We demonstrate that for steady-state heat loss through slab-on-ground, the proposed correlation model agrees with the CSIRO analytical solution within 6.8%, and for transient heat loss through slab-on-ground and basement the new model agrees with a well validated commercial 3D heat transfer program. The new simplified approach presented covers most foundation configurations of interest, with low processing requirements, and is ideal for use in building energy performance simulations due to the short computation time.