A case study for sensitivity-based building energy optimization

Abstract

Building design optimization process is associated with uncertainties due to climate change, unpredictable occupant behavior, and physical degradation of building material over time. The inherent uncertainties in the design process reduce the reliability and robustness of the optimal design solution(s) and affect design decision-making results. This research studies the capabilities of parametric design tools in adopting probabilistic methods to handle uncertainties in building performance optimization. Variance-based methods, e.g., Monte Carlo sensitivity analyses are implemented to identify the most critical parameters in design optimization problems and improve the efficiency of design optimization. The optimal solutions achieved with variance-based methods are satisfying the design objectives more efficiently, also remain robust to changes and uncertainties.