There is an established need for a better integration of 3D geological and dynamic models opening the door for a new generation of conceptual geothermal reservoir models (O’Sullivan, 2013). It has been widely recognized in the exploration and exploitation of subsurface resources that multidisciplinary integration of data into a geomodel yielded an important improvement in model consistency. Yet, even when geological models are produced in such a collaborative context, geological models are rarely reconsider after flow simulations.

One of the practical reason for this is that the geological model is discretized into various kind of meshes (regular voxels, eclipse structured grids, Voronoï tessellations…) that hold partial geological information such as petrophysical properties but are not linked to the parameters of the geological models. For example there exists a large variety of toolboxes for one way transformations between geological models and TOUHG2 input formats and many institutions maintain their own set of routines (Audigane et al., 2011; Wellmann et al., 2012). Consequently, if the geological model is changed, even slightly, the description of the physical problem is to be done all over again. Generic multiphysics industrial frameworks easily deal with CAD like descriptions but are much less suited to handle natural objects like 3D geological models.

Moreover, there is no open and shared format describing what should be such a “digital geological model”. Based on that observation, BRGM adopted internally a pragmatic approach for production needs to be able to work with geomodels from different sources. The idea is to build an intermediate representation out of the model that holds all the geological information (Loiselet et al., 2016).

Dernière mise à jour le 13.11.2017