​Multi-energy grids that include both production and storage will be one of the hallmarks of the energy transition. The integration of renewable energy will depend on these new energy systems. One challenge that must be overcome is how to manage these systems for better production planning. For this to happen, the systems must be dimensioned optimally during the design phase. This is currently done using mathematical models that integrate sometimes difficult trade-offs between conflicting priorities like how fast calculations can be completed, how accurate the indicators are, and how complex the model is.

CEA-Liten came up with a new energy system dimensioning and management method that could help. The approach was tested on the electricity, heating, and domestic hot water networks serving 23 buildings in Grenoble's Cambridge neighborhood. The study integrated a gas boiler, the use of domestic heating oil, solar thermal panels, a heat pump, and heat storage. The electricity component of the study was limited to the grid. CEA-Liten's PERSEE software, which includes several different methods, was used to represent the neighborhood's energy system in operation hour by hour, year-round.

The next step was to determine to what extent the approach allows models more representative of short-term system operation imperatives to be used while optimizing year-round operation to leverage seasonal impacts on heating—all without jeopardizing the feasibility of the simulation with overly complex models. When conventional methods based on simplified models are used, production and storage systems tend to appear under-dimensioned due to suboptimal operation. This is characterized by an excessive use of more polluting fossil-based energy sources (like domestic heating oil in this study). CEA-Liten's new method allows the system to be managed almost as efficiently as if demand for the entire year were known in advance, for a difference of 1% to 2%. It also proved to be highly adaptable if actual demand does not align with the forecasts. The method works so well because operating constraints were maintained, but not simplified, and the energy production means in the system were optimized to reduce the use of fossil fuels.