This project attempts to tackle the problem of analysing the heat demand of build- ings at a large (district or city) scale. Specifically, building data from Geneva is used as a case study while building an environment with suitable tools for a continued analysis. Detailed modelling of these systems is limited greatly by not only com- putational resources, but also by data availability. Meta-modelling is a potential technique to overcome these issues, and the first steps towards this are analysis of available data and parameter interactions, as well as identification of key influence parameters, through the use of stepwise regression techniques. A set of tools to accomplish these goals is introduced, and then used to perform an initial analysis.