Proposal of typical and design weather year for building energy simulation

Abstract

In building design or research processes, building energy simulations (BES) are conducted using weather data. There are two types of weather data for BES: typical weather year is used to estimate annual cooling/heating loads and design weather data to estimate maximum cooling/heating loads. In this study, we propose a new type of weather year data (called the Typical and Design Weather Year: TDWY) that can be used as both typical weather year and design weather data. To create the TDWY, we selected an average year based on Finkelstein-Schafer statistics and applied quantile mapping (QM) to the average year with parent multi-year (MY) weather data. The cumulative distribution functions of the TDWY created by QM consist completely of parent MY weather data for all the weather components used in QM. As the monthly and annual averages of the TDWY based on QM are equal to those of the parent MY weather data, high performance of the TDWY as typical weather year can be expected. In addition, the hourly values of the TDWY include from the minimum value to the maximum value of the parent MY weather data each month, so the TDWY can also be used as design weather data. To validate the performance of the TDWY, we conducted BES. The TDWY showed better than double the performance for estimating average cooling/heating loads compared to the existing typical weather year and could accurately estimate maximum cooling/heating loads.