Renewable sources of electricity production are strongly increasing in many parts of the world. The production costs are going down quickly, thus accelerating the deployment of new solar and wind electricity generation. In the long-term, these variable sources of electricity could represent a high share of the power system. However, long-term foresight energy models have difficulties describing precisely the integration challenges of Variable Renewable Energy Sources (VRES) such as wind or solar. They just do not represent the short-term technical constraints of the power sector. The objective of this paper is to show a new approach of the representation of the challenges of variability in the long-term foresight energy model POLES (Prospective Outlook on Long-term Energy Systems). We develop a short-term optimization model for the power sector operation, EUCAD (European Unit Commitment And Dispatch) and we couple it to POLES year after year. The direct coupling, with bi-directional exchanges of information, brings technical precision to the long-term coherence of energy scenarios.