Speaker(s): Paul Cupillard

Date: Thursday 28th May 2020 - 2:00 pm

Speaker(s): Antoine Mazuyer

Date: Thursday 13th May 2020 - 11:00 am

Abstract:

Geological sequestration of CO2 in large quantity is one of the most promising methods to curb global warming. To achieve this goal, large volumes of CO2 have to be permanently stored in geological reservoirs. Modeling and simulating CO2 sequestration is a multi-disciplinary effort involving a large spectrum of physical and chemical phenomena. We focus this work on the generation of 3D structural models and meshes to support coupled numerical simulations with flow and poromechanical equations. Building such models and meshes is challenging as the poromechanical changes to stresses and strains are not limited to reservoir layers where CO2 is stored, but also in the surrounding formations (overburden, underburden, sides). Modeling surrounding rock formations is necessary to capture fault reactivation, uplifting and caprock failure issues during the injection. In this work, we propose guidelines to build such models using implicit and explicit modeling strategies and meshes. This work is supported by real field data from a depleted gas field in the Gulf of Mexico state water. This field is considered as a faithful analog to CO2 sequestration reservoirs in the regional Miocene sands.

Speaker(s): Bruno Levy

Date: Wednesday 06th May 2020 - 2:00 pm

Abstract:

There is a lot of scientific activity around Optimal Transport (OT for short), that has became a "buzzword". OT is at the connection between math, physics and computer science. It has also some interesting applications in AI, which attracted even more attention to it. In this presentation, I will take a completely different path: I will introduce OT with a physicist's hat on my head: My goal is to give an intuitive idea of OT, and its connections with physics. In particular, OT can be used as a component of a new class of numerical simulation methods, with interesting properties. In particular, it is possible to design a king of "moving mesh method" while exactly controlling the volume of the mesh elements. This can be used to simulate fluids, or to solve some inverse problems in astrophysics.

Speaker(s): Francois Bonneau

Date: Thursday 09th April 2020 - 4:00 pm

Abstract:

In this presentation, I will present the current state of my thoughts and developments related to stochastic Discrete Fracture Network simulations.

 

ring

Speaker(s): Guillaume Caumon

Date: Thursday 2nd April 2020 - 4:00 pm

Abstract:

Loop is a collaborative project dedicated to the development of an open source platform for modeling of highly deformed geological terrains and inversion. In this seminar, I will make a brief summary of the main advances shared by the 2020 Loop meeting participants in Western Australia. Themes include the overall project architecture, map and model validation, modeling, joint geophysical inversion.

Speaker(s): Corentin Gouache

Date: Thursday 26th March 2020 - 2:00 pm

Abstract:

Earthquakes occur everywhere and everytime. Moreover, large earthquakes that really impact our society are sparse events. That's why statistical analysis of seismicity attended to describe the large amount of small earthquakes in order to extrapolate to the large earthquakes through scaling laws. Thus, it has been found that independent seismicity uses to follow a Poissonian law whereas the dependant events follow a time-reversal law. However, these laws have been described in California or Japan, where seismicity is strong. Regarding the low amount of data in low seismicity areas like French mainland, the adequacy of these laws can be questioned. This seminar highlights the drawbacks brought by a classical model of seismic hazard estimation when applied on low seismic activity areas. A scheme is proposed to overcome these issues.

Speaker(s): Luc Scholtes 

Date: Thursday 12th March 2020 - 1:00 pm

Location: Meeting room, ENSG Bat E, Nancy.

Abstract:

Predicting the strength of shale rocks is a basic but nonetheless critical requirement when designing CO2 or nuclear wastes storages. Even though geologic structures need to be considered above the meter scale when evaluating the strength of such systems, the intact rock behavior has to be thoroughly characterized before any further analysis. A particular feature of shale rocks is that they present highly nonlinear failure envelopes. Interestingly, despite this unanimously observed behavior, the linear Mohr-Coulomb criterion is still widely used in engineering practice (see for instance the petroleum or reservoir geomechanics taught in university courses). Of course, non-linear strength criteria exist and should thus be considered more systematically when dealing with shale rocks (see for instance the criterion proposed by Singh). Nonetheless, as practical as they are for assessing the integrity of geologic materials, these formulations (empirical for the most of them) do not explain the underlying mechanisms involved in the failure processes. As a matter of fact, shale rocks present anisotropic structures, characterized by laminations, parallel layering or bedding features. These fabric properties induce a strong anisotropic behavior which directly controls the strength of shale rocks. It is thus crucial to consider such characteristics to better understand the associated failure mechanisms. In this work, we describe the dependency of Tournemire shale strength on the orientation of the loading relative to the bedding over a large range of confining pressures (from 2 to 80 MPa) based on a comprehensive series of experimental tests. Then, we assess the relevance of two practical empirical failure criteria to describe Tournemire shale strength. Finally, we simulate the triaxial tests performed on Tournemire shale using a specifically developed anisotropic discrete element model to get further insights into the mechanisms controlling its strength.

Speaker(s): Zoé Renat

Date: Thursday 27th February 2020 - 1:00 pm

Location: Room G201, ENSG Bat G, Nancy.

Abstract:

Microseismic events are earthquakes of weak intensity and are difficult to locate with traditional methods. In the presentation, I will present some methods based on backward propagation of seismic recordings: time reversal method & reverse time modeling.

Speaker(s): Khalifa Eldursi

Date: Wendnesday 19th February 2020 - 1:00 pm

Location: Meeting room, ENSG Bat E, Nancy.

Abstract:

Canada’s uranium production, currently ranked second worldwide and accounting for 17% of the world’s total, is entirely from the Athabasca Basin in northern Saskatchewan. The unconformity-related uranium deposits associated with this basin are among the highest-grade and largest uranium deposits in the world. Many of these deposits were discovered based on an empirical model that uranium mineralization took place near faults crosscutting the sub-Athabasca unconformity, where uranium-bearing oxidizing fluids met and reacted with reducing agents (i.e., graphite and hydrocarbons derived from, as well as ferrous iron-rich lithologies) and precipitated uranium ores. However, after more than 40 years of intensive exploration, this model is facing more and more challenges, and a better understanding of the factors controlling mineralization is needed.