Gaussian FLOWERS: Wind-rose-based analytical integration of Gaussian wake model for extremely fast AEP estimation

Whittaker, Caidan; LoCascio, Michael J.; Martínez-Tossas, Luis A.; Bay, Christopher J.; Bastankhah, Majid

Abstract

A major cost in the study of wind farm layout optimization is the repeated evaluation of the annual energy production (AEP). The current approach to estimating AEP requires a large set of flow simulations to be performed that cover each discrete wind speed and direction combination contained within the wind rose, followed by a probability-weighted sum of the power production resulting from each simulation. Even with inexpensive engineering wake models, this numerical integration scheme can lead to high computational costs. In this paper, we derive an analytical formulation for estimating farm AEP across every wind direction, based on a Gaussian wake velocity model, which reduces the number of wind farm simulations to a single function evaluation. As a result, we find that the Gaussian-FLOWERS approach reduces the time for AEP calculations by more than two orders of magnitude with a small trade-off in accuracy when compared to a conventional approach. This massive reduction in computation cost is useful to reduce overall costs in wind farm layout optimization studies.

Citation

Whittaker, C., LoCascio, M. J., Martínez-Tossas, L. A., Bay, C. J., & Bastankhah, M. (2025). Gaussian FLOWERS: Wind-rose-based analytical integration of Gaussian wake model for extremely fast AEP estimation. Journal of Renewable and Sustainable Energy, 17(1), Article 013306. https://doi.org/10.1063/5.0245886

• Journal Article Type: Article
• Acceptance Date: Jan 10, 2025
• Online Publication Date: Feb 10, 2025
• Publication Date: 2025-01
• Deposit Date: May 22, 2025
• Publicly Available Date: May 22, 2025
• Journal: Journal of Renewable and Sustainable Energy
• Electronic ISSN: 1941-7012
• Publisher: American Institute of Physics
• Peer Reviewed: Peer Reviewed
• Volume: 17
• Issue: 1
• Article Number: 013306
• DOI: https://doi.org/10.1063/5.0245886
• Public URL: https://durham-repository.worktribe.com/output/3958707