Investigating wind farm blockage in a neutral boundary layer using large-eddy simulations

Jessica M.I. Strickland, Richard J.A.M. Stevens*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

14 Citations (Scopus)
118 Downloads (Pure)

Abstract

Understanding how blockage influences large wind farms is essential as the first row produces the most energy and is often used as a reference for the subsequent rows. We use large-eddy simulations to investigate wind farm blockage by comparing a stand-alone turbine, an infinite row of turbines, and a wind farm with eight rows. We find that the blockage effect for dense turbine arrays is highly dependent on the non-dimensional turbine spacing. Spanwise neighboring turbines appear to benefit from deflected flow, while close downstream turbines enhance flow over the wind farm, reducing productivity at the front. In agreement with the uniform inflow wind tunnel measurements by Segalini and Dahlberg (2020), we find that the effect of downstream turbines follows a universal trend as a function of the average inter-turbine spacing when the performance of the first wind farm row is normalized with the corresponding isolated row case. However, we also demonstrate that the wind farm layout strongly affects blockage as the results do not follow a universal trend when normalized with the performance of a stand-alone turbine.
Original languageEnglish
Pages (from-to)303-314
Number of pages12
JournalEuropean journal of mechanics. B - Fluids
Volume95
Early online date18 May 2022
DOIs
Publication statusPublished - 1 Sept 2022

Keywords

  • Wind energy
  • Wind farm
  • Renewables
  • Fluid mechanics
  • Fluid dynamics
  • Turbulence
  • Large eddy simulation
  • Blockage
  • Power production
  • Computational fluid mechanics
  • High performance computing
  • UT-Hybrid-D

Fingerprint

Dive into the research topics of 'Investigating wind farm blockage in a neutral boundary layer using large-eddy simulations'. Together they form a unique fingerprint.

Cite this