Recent improvement in Alya for Wind Energy Large Eddy Simulations towards Exascale
Thursday, July 1, 2021 12:00 PM to 12:15 PM · 15 min. (Africa/Abidjan)
Stream#1
Information
Contributors:
Abstract:
The rapid adoption of solar and wind power generation by several countries has demonstrated that renewable energy can competitively supply significant fractions of local energy needs without producing carbon dioxide emissions. Considerable progress in this direction is urgently needed to reduce Global warming. This talk focuses on recent progress in the HPC code Alya to perform Computational Fluid Dynamics - Large Eddy Simulations (LES) that are of interest to the wind energy community. Wind power is the renewable source with the most successful deployment over the past decade (2007-2016), growing from 93 GW to 490 GW of world-installed capacity during this time. However, energy losses are estimated to be 20% and up to 30% in complex terrain wind farms. A 2% improvement in efficiency would save over $1 billion annually. A better understanding of the complex Multiscale and Multiphysics flow is needed to reduce wind plant losses. Thanks to ever-growing computational power and improved algorithms, numerical simulations are an excellent tool to understand the wind flow better. This talk will briefly describe improvements to the physical modeling of the flow over complex terrain. Those improvements have allowed us to simulate the Alaiz hill's flow using LES for four days, including thermal coupling, Coriolis forces, and tendencies coming from the mesoscale code WRF. Moreover, HPC improvements that bring Alya closer to the exascale such as porting to GPUs and interfacing with external linear algebra libraries, will be described.
Abstract:
The rapid adoption of solar and wind power generation by several countries has demonstrated that renewable energy can competitively supply significant fractions of local energy needs without producing carbon dioxide emissions. Considerable progress in this direction is urgently needed to reduce Global warming. This talk focuses on recent progress in the HPC code Alya to perform Computational Fluid Dynamics - Large Eddy Simulations (LES) that are of interest to the wind energy community. Wind power is the renewable source with the most successful deployment over the past decade (2007-2016), growing from 93 GW to 490 GW of world-installed capacity during this time. However, energy losses are estimated to be 20% and up to 30% in complex terrain wind farms. A 2% improvement in efficiency would save over $1 billion annually. A better understanding of the complex Multiscale and Multiphysics flow is needed to reduce wind plant losses. Thanks to ever-growing computational power and improved algorithms, numerical simulations are an excellent tool to understand the wind flow better. This talk will briefly describe improvements to the physical modeling of the flow over complex terrain. Those improvements have allowed us to simulate the Alaiz hill's flow using LES for four days, including thermal coupling, Coriolis forces, and tendencies coming from the mesoscale code WRF. Moreover, HPC improvements that bring Alya closer to the exascale such as porting to GPUs and interfacing with external linear algebra libraries, will be described.
