Simulation and real weather data: A comparison for Cyprus case

Abstract

This paper compares three sets of weather data for Cyprus and investigates the relationship between similarity of weather data and similarity of building heating and cooling loads. A building energy simulation weather file containing a year of hourly values of typical weather data for Larnaca was compared with 10 years of daily mean data (dry bulb temperature, global solar radiation, wind speed and relative humidity) for two other locations (Kyrenia and Nicosia). This was done by reducing the hourly data to daily mean values, generating the cumulative distribution functions of the three data sets and calculating the Finkelstein-Schafer statistics as a measure of similarity. It was found that the data for Larnaca were similar to that for Kyrenia and Nicosia, and most similar to that for Nicosia. The relationship between similarity of weather data, as measured by the Finkelstein-Schafer statistic, and the space heating and cooling loads of a simulated simple building was then investigated for some locations in the Mediterranean region. It was found that there was no apparent relationship between the Finkelstein-Schafer statistics and the total heating and cooling loads, but that there was a clear relationship between the Finkelstein-Schafer statistics and the heating and cooling loads considered separately. It is concluded that: the Finkelstein-Schafer statistic is a useful measure of similarity when comparing weather data sets; it is important to consider space heating and cooling loads separately, not just total load; comparing weather data for Palermo, Cairo and Jerusalem with that for Larnaca, a smaller value of the Finkelstein-Schafer statistic is associated with more similar heating and cooling loads; the Finkelstein-Schafer statistics are smaller for Kyrenia and Nicosia than for Palermo, Cairo and Jerusalem, so weather data for Larnaca can be considered for use in building energy simulation studies for Kyrenia and Nicosia. Practical application: There are many simulation weather data files available for different locations that can be used for dynamic thermal simulations of buildings. Yet, it is not practical to generate weather data file for every location. This work suggests that if a building energy simulation weather data set for a location is not available, then similar data for another location can be used, provided that FS statistic is small between the locations. This comparison has been performed for three locations in Cyprus in this work.