Urban mobility is rapidly shifting toward electrification. Yet, limited range and availability of charging infrastructure will remain issues for Electric Vehicles (EVs) for the foreseeable future. Effectively managing these limitations for both individual vehicles as well as entire regions and cities requires accurate predictions of energy demand. Current tool chains such as Simulation of Urban Mobility (SUMO) already make use of simplistic energy consumption models. However, these models are often applicable only to specific vehicles and do not consider individual powertrain components' characteristics. We close this gap and present a physics-based model for calculating energy consumption using component level modeling of the vehicle powertrain. The model is validated for EVs of different car segments against manufacturer data and test bench measurements. Through an open-source integration into SUMO, the model is available for public use. To showcase the possibilities of the proposed model and its flexibility, we perform a case study to predict the future energy demand of an exemplary mid-sized European city. The results show a strongly increasing demand for electric energy and underline the importance of considering heterogeneous vehicle fleets with different individual energy efficiency levels.