Comparison of multi-year and reference year building simulations

Abstract

Buildings are generally modelled for compliance using reference weather years. In the UK these are the test reference year (TRY) used for energy analysis and the design summer year (DSY) used for assessing overheating in the summer. These reference years currently exist for 14 locations around the UK and consist of either a composite year compiled of the most average months from 23 years worth of observed weather data (TRY) or a single contiguous year representing a hot but non-extreme summer (DSY). In this paper, we compare simulations run using the reference years and the results obtained from simulations using the base data sets from which these reference years were chosen. We compare the posterior statistic to the reference year for several buildings examining energy use, internal temperatures, overheating and thermal comfort. We find that while the reference years allow rapid thermal modelling of building designs they are not always representative of the average energy use (TRY) exposed by modelling with many weather years. Also they do not always give an accurate indication of the internal conditions within a building and as such can give a misleading representation of the risk of overheating (DSY). Practical applications: An understanding of the limitations of the current reference years is required to allow creation of updated reference years for building simulation of future buildings. By comparing the reference years to the base data sets of historical data from which they were compiled an understanding of the benefit of multiple simulations in determining risk can be obtained.