Relations that occur between features located in space–like the fact that a street is surrounded by very high buildings, that an airport is close to a city- as well as spatial properties of features–like the height and width of a door- play an important role for many urban applications. Digital models of cities can assist in the evaluation of these relations and properties either through visualisation or through computation, mainly based on geometrical information. Hence, considering the objective of explaining to potential users of these city models what useful information they can derive from these data and how, a possible way to address this objective lies in the usage of a pivot model composed of relevant spatial properties and relations, connected to information meaningful to the user and connected to the possible computation of them on available data. This paper firstly sets the ground for a typology of such relevant relations and properties that are shared by different applications and that can be derived/approximated from existing data. It then proposes a model to describe these properties and relations and connect them to their possible computation based on data (2D or 3D). An important aspect of this model is to distinguish between a conceptual layer where relations occur between “real world” features and an implementation layer where they are calculated based on database features and geometries.