Mastering complexity through sobriety in numerical simulations of energy production

Abstract

Energy production uses equipment and structures with specific features: their physical dimension (a hydroelectric dam, a nuclear power plant or an 8MW offshore wind turbine are very large structures), their lifespan (over 100 years for some dams) and great physical complexity (multi-physical, multi-scale phenomena), with variable and partially known input data. Numerical simulation provides a major source of information for building, operating and deconstructing them, as this industrial sector is subject to the same efficiency constraints (economic and technical) as any other. Whether for numerical reasons (problem size too large for current techniques, curse of dimensionality) or sobriety, it is necessary to control and reduce the complexity of numerical simulations. In this presentation, we illustrate these issues of complexity and sobriety through the lens of two particular physics: solid mechanics and neutronics. We show the obstacles still present and the research and development directions envisaged in terms of complexity reduction for numerical simulation

Mickael Abbas

Mickaël Abbas graduated in mechanical engineering from Compiègne University of Technology in 1998. He went on to complete a DEA (post-graduate diploma) in numerical modeling, followed by a doctoral thesis under the supervision of Jean-Louis Batoz on the numerical simulation of the crimping process. In 2002, he joined the code_aster finite element code development team at EDF R&D as a research engineer. He then became project manager for innovative numerical methods in mechanics in 2010, then expert engineer in numerical simulation in 2015, and finally, since 2023, senior expert engineer in numerical simulation at EDF R&D. As scientific head of code_aster development, he is involved in a wide range of research activities in numerical mechanics and applied mathematics, with a particular focus on advanced discretization methods and model reduction. He has co-supervised a dozen theses on these topics, always in relation to EDF's industrial applications.

Jean-Philippe Argaud

Jean-Philippe Argaud (Dr) works at Électricité de France in the EDF Lab Paris Saclay Research Center, where he contributes as a research engineer with expertise in mathematical applications. He has led industrial projects, particularly in risk management, data assimilation and neutronics. His recent work focuses on data assimilation and model reduction, applied to physical simulations in many areas of the physics of industrial energy production and distribution systems. He also contributes to the development of open-source scientific methods and tools, to promote the adoption of these practices in engineering study tools.