Credible building modelling is essential to the building energy certification and building renovation towards better energy performance and indoor climate. The primary objective of this work is to assess the consequences related to the level of effort necessary to simulate the building geometry and its facilities and to guide practitioners in building modelling by providing insights about the model simplification consequences both for the energy performance and comfort. This work focuses on the sensitivity of model geometry simplification and heating system and the influence of these on the energy and thermal comfort KPIs in standard simulation conditions. Moreover, the study presents models verification towards operational performance, investigates the complexity of adapted simulation conditions, e.g., heating setpoint, actual people load on the model credibility comparing to monitored data. The sensitivity study main conclusion is that there are relatively small differences in heating demand among models with different zoning methods of geometries, while the implementation of detailed heating systems in the simulation has a more noticeable effect on the results of all output KPIs. The Model verification activity main conclusion is that adapted people's load can improve the model accuracy. Models with detailed geometry, lead to more accurate results when the heating set-point in the model is defined as monitored data per apartment. For dwellings with a limited number of IAQ measured points, use of the standard set-point is advised instead of monitored. For the apartments with sufficient IAQ sensors, adapted heating setpoint and people load can significantly improve the model predictions.