An ARMA type weather model for air-conditioning, heating and cooling load calculation

Abstract

In order to design the capacities of a heating, ventilating and air-conditioning (HVAC) system's elements, configuration of the building and the HVAC system for minimizing energy consumption, it is very important to know the air-conditioning, heating and cooling load of a building. To compute the load, weather data are very important; however, what kind of weather data should be used is a difficult problem. Conventional load calculation methods are divided into two classes, i.e., peak-load estimation and annual-load simulation. Diurnally periodic weather data are used for the peak-load estimation, but the correlation of weather elements, i.e., temperature, solar radiation, moisture contents, etc., can hardly be taken into account. Reference year weather data are used for annual-load simulation, but the results can only give the seasonal summed-up load, no information being obtained for the detailed load variations owing to the shortness of the data period. To overcome the problem, the authors constructed an ARMA-type weather model by applying a system identification technique to the original weather data. The merits of the modeling are: (1) the statistical properties of weather data are kept in the model dynamically; (2) long-term data are reduced to a small number of parameters; (3) the characteristics of weather data can be analyzed systematically; (4) even the climate of a certain location, where a precise and/or long-term data record is not available, could be modeled if the above investigations can be made at a close location; (5) the model can be used for the stochastic heating and cooling load calculation method which was developed by the authors. The methodology to build the model, examples of the weather model and the stochastic heating and cooling load results using the model are given, reasonable consistency being obtained with a simulated load.