Uncertainty analysis for building performance simulation - a comparison of four tools

Over the last three decades the use of building performance simulation (BPS) tools has increased but its ability to support building design is still limited. State of the art BPS tools have the potential to be used more extensively during the entire design process if their current capabilities are expanded. It is hypothesized that by providing additional features supporting output accuracy assessments (uncertainty analysis) and parameter sensitivity studies BPS tools will provide more valuable design information. Integrating techniques as uncertainty analysis (UA) and sensitivity analysis (SA) however requires not one but a large number of simulations of the same model to be conducted. Prototypes have been built automating the process for four software tools representing the modeling abstraction level from conceptual to detail design. The tools considered are LEA, IES VE, VA114 and HAMbase. Potential sources for uncertainty variation of the results are discussed. The paper concludes with an indication of future work