Impact of meso-scale weather uncertainty on building performance simulation outcomes

Building performance simulation is a complex experiment that has many inputs that are assumed tobe deterministic but are, in-fact probabilistic. In case of low-risk facile simulations, deterministic inputs may be assumed for the simulation parameters as they do not warrant experiments that cover all the possibilities. However, for high-risk design decisions, uncertainty assessment is a must to avert any failures that result in subpar operational performance. Weather in a city may be localized due to natural or altered urban attributes and any deviation from the statistically derived "average" weather data must be examined for uncertainties in performance outcomes. Case studies on three different buildings in Chicago and Atlanta indicate a large variance in performance outcomes when weather related uncertainties are accounted for in the Uncertainty Analysis as compared to deterministic weather from the nearest airport. The findings of this study suggest that for scenarios with large penalty on not meeting the performance targets, the modeler should be careful about the relative location of the proposed building in the city and should not, defacto accept weather data that is derived from a remote location such as an airport.