Influence of the representativeness of reference weather data in multi-objective optimization of building refurbishment

Energy saving measures properly applied to the existing building stock can bring noticeable savings. In particular, optimal cost-effective solutions can be found through multi-objective optimization techniques, such as those based on genetic algorithms (GA), coupled with building energy simulation (BES). Although the robustness of GA multi-objective optimizations to the quality of the inputs is discussed in the literature, the role of the weather data file is not investigated in detail. For this reason, this work analysed the extent to which the method adopted for the development of reference weather data for BES can affect the optimal solutions. Considering a group of simplified building configurations and the location of Trento, Italy, many multi-objective optimizations are performed. The results show changes to both Pareto fronts and optimal retrofit solutions.