A Statiscally-Based Hourly Weather Data Generator for Driving Energy Simulation and Equipment Design Software for Buildings

This paper describes an operating hourly weather simulation model which is utilized to drive building energy simulation and equipment design software. The model contains both deterministic and probabilistic portions that perform sun angle calculations, sky opacity, dry-bulb temperatures, dewpoint temperatures, wind speed, and barometric pressures. Sample summary statistics have been shown to give good agreement with means, extremes, and distributions of real weather data records. A sample energy analysis is also included to illustrate the model's flexibility for different applications. The model is also useful for creating synthetic hourly weather data files in TRY or TMY formats for use in a variety of energy calculation software or other applications. A special data compression feature of the program permits it to simulate less than full months and still maintain its statistical integrity for energy calculations. This feature enables the energy simulations to run in one-quarter the time required by other simulation approaches. In the compressed mode, the model has been especially useful for material selection and daylighting evaluations during the early architectural design stage when quick turn around is desired. The current database contains weather statistics for 194 U.S. cities, but this can be expanded to include additional cities at the will of the user. Input of new cities can be in either Inch-Pound or SI units. The program runs on most MS-DOS microcomputers.