Regression-Based, Long-Term Prediction of Building Performance Under Climate Change : Building Energy Demand and Bipv Energy Generation Analysis Across Turkey

Acquiring a clear understanding of the long-term consequences of climate change (CC) on the buildings requires systematic analyses. This research presents a methodological framework for lifetime energy demand and PV energy generation predictions of a given building considering the CC impacts through multivariate regression models. As a case study, a hypothetical office building in Turkey was selected. An existing linear morphing methodology was utilized to generate future weather files (2050 and 2080) for all 81 cities in Turkey. For each year and city, corresponding weather metrics were calculated, and heating/cooling demand and PV energy generation values were computed through building energy simulations. Obtained data were used to develop two sets of multivariate regression models: (i) models to predict future weather metrics and (ii) models to predict future energy demand and generation. These models allowed lifetime energy demand and generation analysis (including associated carbon footprint and cost of energy) of the building considering CC impacts using only the current weather metrics of its location. For the lifetime of a building, considering CC impacts yielded significantly higher cooling (up to +0.6 MWh/m 2 in warmest cities) and lower heating loads (up to -0.7 MWh/m 2 in coldest cities), while the impact of CC on PV energy generation was limited. Our regression-based approach can be further expanded to include not only various building parameters and types, but also supply-demand matching potentials