Residential buildings are seen as major energy consumers among other industrial sectors worldwide. As we move towards low-consuming and nearly zero energy buildings, higher performance requirements on sustainability and robustness are manifested. Consequently, a better comprehension and integration of building performance determinants in the very early phases of design has become essential for building constructors. Buildings experts rely on energy modeling and simulation tools for estimating actual and future energy consumptions. However, such tools still do not account precisely for occupant behaviors and their variability, yielding thus to unrealistic and inaccurate results. Therefore, it is assumed that a better comprehension of occupant behaviors and their consumption trends could improve the identification of technical solutions and energy saving potentials, resulting consequently in more robust building designs. In our research work, we aim to develop a model for forecasting the spectrum of occupant-related energy consumption in residential buildings. In this scope, we establish a comprehensive model-driven approach which gives a probabilistic mapping between household profiles and their corresponding domestic energy consumptions. A bottom-up model using an activity-based approach is adopted. In this paper, we present the structure of the model, reveal its different objects together with their interactions, and introduce its ontology. Later on, we demonstrate the modeling and simulation flows leading to calculate the energy consumption for example household types.