Towards large-scale simulation of indoor thermal conditions in buildings supplied with district heating: Model evaluation based on experimental data

Utility companies are studying ways to improve the energy efficiency of district heating networks and promote flexible solutions. However, the impact of any modification must be verified, especially when dealing with existing buildings whose heating systems were originally sized based on different design conditions. Predictive simulations are carried out for this scope. Several building energy simulation options are available, computational effort and availability of input data being the strictest constraints for their adoption. In this preliminary work, real temperature measurements from a multi-family house in a demo site supplied by an experimental district heating branch have been used to assess the performances of dynamic energy models of different complexity. Images, local registries and literature studies have been used as sources of input information. Despite the high degree of approximation of the available data, encouraging results are found, with a root mean squared error below 1.5 °C over the heating season for a single dwelling, while inertial effects need to be better modelled when the whole building is simulated. The opportunity of improving the consistency and coordination between public databases is also discussed as a way to increase energy mapping possibilities.