CHARMS will deliver at the same time operational tools for the exploitation of geothermal reservoirs as well as an unprecedented approach for testing hypothesis and gain scientific understanding in deep geothermal circulations. The innovative nature of the project lies in:

• the tight integration of physical flow simulations with the traditional geological modeling approach that will improve the overall consistency of models and foster collaborative working practice,
• a high performance environment for the accurate hydrothermal modeling of complex geothermal systems implementing new cutting edge numerical algorithms.

An essential condition to achieve a fully integrated model is a two-way interaction with both its static and dynamic components. Considering that we can build a geological model out of a dedicated software (e.g. GeoModeller or Petrel) we need to be able to combine geological information with (petro)physical properties and/or specify boundary conditions for dynamic flow simulations and conversely be able to access and modify the geological model. Complex geological geometries must be discretized accurately, possibly in an anisotropic way, retaining discontinuities such as faults and fractures, and integrating anthropic features such as wells with sophisticated geometries, in order to perform dynamic flow simulations and then visualize and explore their results along with the geology to assess hypothesis and check with observations. Such an interactive framework relies on a convenient and extensible data structure having the representation of a dynamic geological model as a core component. Many features are already available separately through existing libraries or partners’ developments. CHARMS will bring them together in a common framework which will be designed to be as intuitive as possible. We will also improve physics description with the possibility to explicitly specify any equation of states in a high level programing language.

Considering that we want to perform flow simulations on large complex domains we chose to develop a high performance code for the accurate modeling of hybrid dimensional thermal compositional Darcy flows that will:

• handle generic 3D unstructured meshes, possibly containing immersed fractures, without “grid effects”,
• deal with abrupt variations of petrophysical properties and distributions generated from geostatistical techniques, included aperture/permeability distributions along fault/fracture surfaces,
• include the possibility to specify a wide range of boundary conditions with complementary conditions,
• accurately describe transport phenomena (heat and tracer) in an eulerian framework,
• improve convergence behavior when solving the highly nonlinear physics of multiphase hydrothermal circulations,
• target good scalability properties to take advantage of the multi-core and parallel architectures of current computers and benefit from the ever increasing availability of supercomputers.

Core elements have already been implemented in the ComPASS platform and largely validated on several test cases.

Last but not least, operability is one of our main concern as much of the motivations behind CHARMS consortium comes from our experience of the current limitations of existing software for geothermal reservoir simulation both in terms of performance and integration with geological models. CHARMS will include several baseline validation cases taken from the literature as well as benchmarking with already existing case studies representative of different deep geothermal technologies and covering a wide range of geological contexts. Finally, two operational case studies in high temperature conventional fields, will serve as references.

On the geothermal community level, we ambition to rely on these concrete evidences to contribute to a long awaited transition towards efficient simulation codes and an integrated modeling approach. Our open source strategy aims at facilitating cross-fertilizing external contributions and benchmarking with geothermal or other modeling communities.

On the software level we target a modular evolutionary and open framework based on frameworks and high level-API.