Application of Monte Carlo Simulation and Optimization to Multi-Objective Analysis of Sustainable Building Designs

During the design phase of a building, there are often multiple options for selecting building materials and components that make up the building. This variety of options often results in multiple possible designs, each having different construction time, cost, and environmental impact. In order to determine optimal designs, optimization procedures, such as genetic algorithms, have been typically applied. However, current optimization procedures do not consider data uncertainties in productivity, environmental impact, and unit cost; thus, it is not known how data uncertainties may impact on the determination of optimal solutions. Simulation of Environmental Impacts of Construction, or SimulEICon, is a multi-objective analytical tool for observing relationships between time, cost and environmental impact during the early design stage of a building. In this paper, the extension of existing SimulEICon is discussed to include Monte Carlo simulation to account for data uncertainties and availability of data, and analyze their impact together with genetic algorithm-based, multi-objective optimization. Monte Carlo sampling is used to address the inherent uncertainty in the data. All data are behaviorally modeled using probability distributions based on various parameters and used to simulate the environmental impact, construction duration and cost of a building. The analytic tool is implemented by using MATLAB. The results are used to explain trade-off relationships between multi-objectives and to validate the impact of uncertainties. By observing results of different simulations it becomes evident that the effect of uncertainty is inherent to each solution set. This emphasizes how important the impact of uncertainty and availability of data is to a project.