Geothermal energy, namely the mobilization of the subsurface heat at very low, low or high temperatures, is one of the methods to achieve the energy transition. There is a real risk that an installation may not reach a geothermal resource with sufficient flow and temperature characteristics to ensure the cost-effectiveness of the project during its life time. These risks consist of t low water flow / thin thickness of reservoir (meter-scale), the risk of interference between geothermal systems in high density of well infrastructures or the risk of early thermal breakthrough. These risks constitute real obstacles for the future development of geothermal energy in Ile-de-France. This projet consists to work on innovation by proposing solutions that optimize and explore the development of new reservoirs. The optimization of the use of deep geothermal energy is a major challenge for the Région Ile-de-France, which has a population of nearly 12 million inhabitants and still growing. This optimization of geothermal production of aquifers requires (1) precise knowledge of the reservoir heterogeneity in terms of sedimentary geometries, porosity/permeability, reservoir connectivity and (2) reliable numerical simulations of flows and temperature evolution in the underground 30 years or even 100 years after production starts. The main objectives are to successfully perform the upscaling pore- scale laboratory measurements and kilometre-scale sedimentary connectivity of reservoir bodies in order to better predict the resource. The homogenization will give the equations valid at every point of the domain, for both fluid and solid constituents, and in the same time the effective coefficients such as porosities, permeabilities, mechanical deformation (Gassman’tensor and Biot’s coefficient) and effective heat dispersivity. Their determination from the geological data will be the main challenge of this project coupling geological and mathematical concepts.