Building energy retrofits using ant colony optimisation

Building retrofits offer an effective solution to enhance energy efficiency in buildings. However, identifying the most effective retrofit remains a challenge due to numerous possible retrofit measures which may have different impacts on building energy consumption. Simulation-based optimization methods (coupling building simulation software with an optimisation algorithm) are powerful tools to identify the best retrofit for a building. In this research, nine variables of a typical Australian office building are optimised under three different sets of building simulation assumptions using Ant Colony Optimisation for Mixed Variables (ACOMV). Results show that up to 19.1% energy savings can be achieved after applying optimal retrofit measures. Results also indicate that incorrect building simulation assumptions (e.g. internal loads) can considerably reduce energy savings obtained by optimisation method. This research provides new design insight into the importance of simulation assumption in building energy optimisation.