Building performance simulation (BPS) applied at earlier stages of design has the potential to assist design parameter decisions that significantly impact a building's life cycle. Since at such early stages, numerous design variables are still undetermined (e.g. window type, insulations, etc.) the scenarios to be simulated through BPS will be vast and will require extensive time and computational power. Previous studies have tested the sensitivity of a building's energy consumption during its operation, to design parameters. Most of those studies, however, have used a single case study in their analysis. Thus, the objective of this paper is to evaluate the dependency of results of such sensitivity analyses, on the case study being used. To accomplish that, a hybrid method combining one-parameter-at-a-time (OAT) and global samplings was used. Within the cold climate scope of Quebec, Canada, multiple buildings were used to investigate the sensitivity of energy and economy performance to design parameters (architectural, electrical, and mechanical systems); as well as the sensitivity of parameters' impact on building models. Results indicate that architectural and electrical parameters are sensitive to the model. To expand on the understanding of the root cause of this sensitive behaviour, hypotheses were developed and evaluated through global sampling.
