In developed countries, the residential and commercial building stock account for a considerable share of final energy demand and greenhouse gas emissions. Building stock modeling is an established tool to assess different development paths of buildings on city, region or country level. Current building stock models (BSM) as well as previous works of the authors, however, lack a holistic approach that take technological, economic and ecological factors into account on an individual building scale. There are, therefore, limitations in the conclusions that can be drawn. In order to increase their significance, current research shows trends towards spatial differentiation, representation of individual building and owners as well as economic decision modeling. However, no model combines all three aspects in a more holistic approach. This paper describes a novel approach which combines spatial differentiation with building specific heat demand modeling and an economic decision simulation. The model developed combines a building specific engineering model with a micro-economic discrete choice approach. Using spatial building data, the engineering model calculates space heat and hot water energy demand on a building level. The alteration of the building refurbishment state is modeled using a discrete choice approach to simulate the decision process of building owners of building envelope refurbish and/or to substitute the heating system. Due to the building specific approach, the decision model is able to take into account building specific information such as size, geometry, room temperature, investment, maintenance and energy costs and achievable energy savings as well as other factors such as local potentials and restrictions on the use of renewable energy. In a case study of the city of Zürich we demonstrate the feasibility and strengths of the new model approach. The results demonstrate that modeling space heating demand on an individual building scale yields specific heat demand distribution across building clusters (and not simply in average values as in other models). The building level approach enables the model to deliver differentiated results of the heat demand development for the whole building stock, building types building periods or spatially distributed as shown in the results.