Application of a Global Uncertainty and Sensitivity Analysis for Identifying Influential Sustainable Building Design and Operation Parameters

The careful identification and treatment of sensitive building design and operation parameters can aid sustainable building modelling and optimization efforts. Using Design Builder software, this paper applies Latin hypercube sampling (LHS) and standard regression coefficient (SRC) for uncertainty and sensitivity analysis (UA/SA) to ascertain influential building design features influencing annual energy demands and thermal comfort levels. A commercial office building in Gandhinagar, India (Koppen-Geiger: Bsh climate), is chosen for analysis. The building’s simulation model shows 90.01% agreement with the utility bills. Specifically, the influence of four unique wall constructions, five roof insulations, four window-to-wall ratios (WWR), five window types, four overhang depths, five air conditioner’s (AC) coefficients of performance (COP) and five cooling set points (CSP) are assessed on building annual energy demand (AED) and annual thermal discomfort hours (ATDH) as per ASHRAE 55 2004. LHS permutes these eight parameters to produce 300 unique building design and operation possibilities, simulated sequentially to generate the model outputs. Notably, the lowest and highest AED and ATDH vary by a factor of four (5423–22923 kWh) and three (742–2242 h), respectively. The standard regression coefficient global SA identifies AC COP and CSP as the two critical influential parameters swaying AED, followed by WWR, Wall Construction and Roof Insulation. ATDH chiefly depends on AC CSP, followed by WWR, wall construction and roof insulation. Thus, substantial AED and ATDH cutbacks can be achieved by selecting optimal AC COPs and CSP, WWR, wall construction and roof insulation. Other less influential parameters like building orientation and overhangs do not produce comparable benefits and do not require further optimization.