A reverse data-driven energy estimation approach for evaluating renewable energy systems implementation in New York City

Abstract: Urban areas are responsible for a significant part of greenhouse gas emissions. Megacities, especially, have cumulative impacts due to different anthropogenic heat sources and urban heat island phenomena. It is therefore vital to replace fossil fuel burning heating sources with renewable sources in the current building stock. Renewable energy sources can be applied through building energy retrofits, which need to be technically and economically feasible. It is, however, always a complex matter to have a fast and reasonably accurate energy use prediction for the retrofit feasibility study. New York City was chosen for this research due to the wide variety of publicly accessible data, including high-quality metered energy and high-resolution geomatics data. This study proposes a reverse method to predict hourly building energy performance according to the total metered energy use. Buildings were divided into different categories based on type and vintage. A robust relation between building categories and hourly energy use was found. Hourly energy trend was defined to describe the energy use of each category. The output was separated into cooling- and heating loads to be used in a geothermal system design program. The geothermal system was considered as an appropriate system to be established in the existing buildings. The results showed that 50% to 100% of the heating loads of the buildings could be covered by geothermal as an alternative to fossil fuels. The introduced reverse method can significantly reduce the time of energy estimation compared to the simulation-based methods. Hence, providing a retrofit proposal to implement renewable energy systems such as geothermal systems would be faster, easier, and affordable for practitioners.