Modelling the passive thermal response of a building using sparse BMS data

This paper describes the application of system identification to the thermal response of a conventional office space using data collected from an existing building management system. Autoregressive models are estimated and it is found that the output-error model provides the best fit between simulated and measured internal-air temperature data. The inclusion of electricity consumption data improves the fit for all model structures used, though this is evidently not modelled as providing a simple internal heat-gain. The model error is comparable with that reported for models validated in laboratory test rooms, but without the use of full instrumentation.