In a geothermal reservoir, one may think of two types of connectivitiy : the flow connectivity and the thermal connectivity, both being linked but to what extent ? Flow connectivity depends on the distribution of the sedimentary geobodies and their permeabilities and thermal connectivity, which is governed by both diffusion (isotropic process) and advection in the fluid (connected process), may exist when flow connectivity is poor. Therefore, the connectivity issue in a geothermal reservoir is a rather complex problem as it depends on a coupled transport process in a heterogeneous medium. We propose to explore this by means of statistical physics, namely percolation theory. According to the pioneering work of King [1], this framework allows to characterize the connectivity by means of different statistical variables: the percolation threshold, or the value of net to gross ratio above which there is connectivity, the exponents associated to the effective permeability and the connectivity length. These different parameters depend on the type of reservoir: fluvial, deltaic, tidal sand bars within estuary, ooid tidal inlets, ooid shoals in carbonate platforms and on the topology of the geobodies for each type. The connectivity issue will be investigated by means of (1) theoretical considerations on the renormalization group applied to this type of Boolean topology (i.e. reservoir with two types, or more, of geobodies) and (2) 3D numerical heat, mechanical and flow simulations on very large grids. We will apply the distribution of petrophysical parameters (porosity, permeability, thermal properties) and connectivity indicators to the flow, mechanical and thermal simulation (THM modeling) at a geothermal doublet scale and basin scale. We will compare this simulation with well pumping tests (task 1). The task will also propose recommendations for operators to optimize the well architecture / well positioning to increase the doublet lifetime. For this purpose, we will integrate into the softwares packages (PumaFlow, Eclipse…), the geostatistical realizations produced by Petrel (task 1).

[1] P. R. King, “The connectivity and conductivity of overlapping sand bodies,” in North Sea Oil and Gas Reservoirs—II, Springer, 1990, pp. 353–362.