A holistic simulation-based optimization approach for dimensioning cost optimal and nearly-zero-energy buildings

Building design is an inherently multi-objective process, which entails a trade-off being made between two or more conflicting design objectives, like minimizing both primary energy consumption (PEC) and life cycle cost (LCC). This paper presents an implementation for a holistic automatic simulation-based optimization approach with search techniques for multi-objective designs of combined building elements and energy systems. In line with the new recast of the Energy Performance of Buildings Directives (EPBD, 2010/31/EU), the PEC and the LCC of a single family house in cold climate of Finland are minimized considering various design variables: types of the building (a standard house, a low energy house and a passive house), HVAC system (equipment sizes and operating temperatures) and different types of district and on-site energy generation including micro-generation. Optimal solutions (cost-optimal and nearly-zero-energy integrated building designs) are found among more than 900,000 possible ones (possible combinations of 88 options of 9 design-variables) by simulating only 3000 solutions. The optimization is performed by a modified genetic optimization algorithm (PR_GA, variant of NSGA-II). The quality of the optimization results is verified by comparing the obtained optimal solutions with true optimal ones found by applying exhaustive search technique. The comparison showed that the holistic simulation-based optimization approach can reduce the computational effort significantly achieving an acceptable level of close-to-optimal solutions.