Data driven economic scenarios for retrofitting residential buildings in a northern Italian region

European directives and strategies, such as the 'European Green Deal' and the 'Ren- ovation Wave', point out the importance of the building sector in achieving the climate goals set by the European Union for 2050. However, a higher renovation rate for the existing buildings is required to achieve these goals. Many barriers prevent the renovation rate from growing. Regarding financial barriers, the long payback times of renovation interventions and the high risk perceived by the potential investors make the renovation rate remain low. Based on data from energy performance certificates, this research proposes a data-driven method to create economic retrofit scenarios for residential buildings using Artificial Intelligence techniques and Monte Carlo simulations. Namely, energy savings have been predicted using an Artificial Neural Network on clusters of residential buildings and the Life Cycle Costs forecasted by Monte Carlo simulations taking into account the uncertainty in many of the inputs. Results obtained by applying the method to a region in northern Italy illustrate two scenarios for the energy retrofit of the built environment, one assuming a payback time of fifteen years and the other of twenty- five years. In both cases, the maximum allowable investment, which varies according to the specific characteristics of the buildings, is much lower than the retrofit costs recorded in the same area in recent years.