In building performance simulation, understanding the potential for parameter variations to cause a disproportionately large change in a performance metric is an important aspect of the modelling and design process. This is especially true if the proposed building is expected to meet a performance target such as net-zero energy consumption. In the context of this paper, variations refer to design modifications which lead to large changes in a performance metric. This paper proposes a methodology to identify influential variations around a performance criterion. This methodology aids in the understanding of possible discrepancies between predicted and realized building performance. A net-zero energy house case-study demonstrates the methodology. A variability analysis of the case-study indicated that combinations of variations caused energy consumption to be larger than on-site generation in 20% of variational scenarios. A back-tracking search identified that 8 of 26 variables were responsible for significant changes in net-energy consumption. In particular, energy-related occupant behaviour, solar orientation, and variables related to the sizing of a roof-based photovoltaic system can significantly influence net-energy consumption. The case-study helped quantify two optimal approaches for passive solar design - one relying on high insulation levels and lower window areas, and the other relying on good insulation levels and large window areas.

Additional Metadata
Keywords back-tracking search, Monte Carlo, net-zero energy, optimization, variability analysis
Persistent URL
Journal Journal of Building Performance Simulation
Bucking, S, Zmeureanu, R. (Radu), & Athienitis, A. (Andreas). (2014). A methodology for identifying the influence of design variations on building energy performance. Journal of Building Performance Simulation, 7(6), 411–426. doi:10.1080/19401493.2013.863383