Is Building Energy Simulation based on TMY Representative: A Comparative Simulation Study on DOE Reference Buildings in Toronto with Typical Year and Historical Year Type Weather Files

The goal of this study is to determine whether the use of Typical Year (TY) type weather file is adequate for estimating heating and cooling load in Building Energy Simulation (BES). Two Typical Year (TY) Type weather files – Canadian Weather Year for Energy Calculation(CWEC) 1990s and CWEC 2016 are compared with Historical Year (HY) weather data Canadian Weather Energy and Engineering Datasets(CWEEDs) 1998 to 2014. Three stages of analysis are completed – 1) comparison of the cumulative distribution of Dry Bulb Temperature and Global Horizontal Irradiance; 2) analysis of extreme weather event occurrences through comparing temperature-based weather indices; 3) comparison of BES results from simulating DOE prototype buildings. Results show a hotter summer and warmer winter in CWEC 2016 indicated by a shift of median temperature in stage 1, increased hot weather characterizing indices and decreased cold weather characterizing indices in stage 2 when compared with CWEC 1990s. This translates to a lower heating energy use and higher cooling energy use in stage 3 with simulation using CWEC 2016. The total 17-year heating and cooling energy sum estimation using CWEC 2016 shows lower percentage difference for all buildings, ranging from 0 to -8% difference, where most buildings are -1% or -2% from the estimation using historical weather data. However, neither simulation using CWEC 1990s nor CWEC 2016 yield consistent accuracy when compared to the yearly Historical simulation results, since the occurrence of years with hotter/colder temperature shows a periodical cycle, some years have more hot weather events, whereas other years have more cold weather event. This study concludes that Typical Year simulation alone is not sufficient for Building Energy Simulation because 1) past weather data might not be a good indication of recent or future weather events due to changes in climate; 2) the update cycle of TY type weather files are infrequent and therefore most recent climate may not be included in the analysis. Future work should focus on the development of a commonly accepted practice in using multi-year simulation, as well as a commonly accepted format for future or stochastic weather file format.