FCCN HPC

Medium and High-fidelity Aerodynamic Models for Wind Farm Power Optimization

Funded by:

Topic:

Comparison of multi-fidelity CFD models for the simulation of turbine interaction.

Reference:

2022.15647.CPCA.A2

Access typology:

A2 Regular Access

Cluster Facility:

LCA, Coimbra, Portugal

CPU time:

2 500 000 core.hours

GPU time:

0 gpu.hours

Platform:

Navigator – X86 cluster

Duration:

12 months

Start date:

01.05.2023

Summary:

blueOASIS was awarded 2.25M core-hour by the RNCA/FCCN in the Navigator cluster, of University of Coimbra, to investigate multi-fidelity CFD methods (Actuator Disk/body-resolved Sliding Grids; RANS/SRS) to model the aerodynamics of the rotor in turbine-turbine interactions.

The insights of this investigation are critical for wind farm optimization, where the power yield of the farm is attempted to be maximized. This requires accurate methods to capture the turbine-turbine interaction, which are at the same time computationally efficient – therefore it is important to know how the low-fidelity models compare to the higher-fidelity ones.