PhD Thesis (2022-2025)

Title: Discrete karst network simulations: application to the Barrois limestones.

Supervisors: Pauline Collon (RING Team), Vincent Bailly-Comte (BRGM)

 Karstic networks, which are often present in limestones, are characterized by the presence of a complex ensemble of interconnected underground drains which strongly control the fluid flow in the massif. Correctly understanding and modelling their hydrogeological behavior requires to integrate the karstic network in the flow simulation. The Barrois Limestones are characterized by a strong karstification and are subject to a characterization study co-supervised by BRGM and ANDRA. The abundance of already acquired data, as well as those which will be soon available, represent a real opportunity to identify and develop a stochastic simulation approach for karstic networks which includes a maximum amount of geological and hydrogeological constraints. The intrinsic coherence of those constraints will be guaranteed by the construction of a conceptual and integrative model following the KARSYS methodology (Jeannin & al., 2013). The goal of the work is to generate a set of possible geometries for the Barrois karstic network which honors the geological information and field data. Even if there is a lot of available data, the system cannot be exactly known. One of the scientific challenges of this work will be to create a new approach allowing to better take into account the variety of possible equiprobable geometries, while fully exploiting the field data to reduce this variety (and remove incoherent configurations with respect to data). The coupling with the flow simulation models will be done by using multiscale numerical models which will be executed in the Barrois Limestones project. Those different couplings will allow to include the karstic network generation approach in a global inversion cycle.