LAPSE:2023.20781
Published Article

LAPSE:2023.20781
Wind Farm Power Prediction Considering Layout and Wake Effect: Case Study of Saudi Arabia
March 20, 2023
Abstract
The world’s technological and economic advancements have led to a sharp increase in the demand for electrical energy. Saudi Arabia is experiencing rapid economic and demographic growth, which is resulting in higher energy needs. The limits of fossil fuel reserves and their disruption to the environment have motivated the pursuit of alternative energy options such as wind energy. In order to regulate the power system to maintain safe and dependable operation, projections of current and daily power generation are crucial. Thus, this work focuses on wind power prediction and the statistical analysis of wind characteristics using wind data from a meteorological station in Makkah, Saudi Arabia. The data were collected over four years from January 2015 to July 2018. More than twelve thousand data points were collected and analyzed. Layout and wake effect studies were carried out. Furthermore, the near wake length downstream from the rotor disc between 1 and 5 rotor diameters (1D to 5D) was taken into account. Five robust machine learning algorithms were implemented to estimate the potential wind power production from a wind farm in Makkah, Saudi Arabia. The relationship between the wind speed and power produced for each season was carefully studied. Due to the variability in the wind speeds, the power production fluctuated much more in the winter. The higher the wind speed, the more significant the difference in energy production between the five farm layouts, and vice versa, whereas at a low wind speed, there was no significant difference in the power production in all of the near wake lengths of the 1D to 5D rotor diameters downstream from the rotor disc. Among the utilized prediction models, the decision tree regression was found to have the best accuracy values in all four utilized evaluation metrics, with 0.994 in R-squared, 0.025 in MAE, 0.273 in MSE, and 0.522 in RMSE. The obtained results were satisfactory and provide support for the construction of several wind farms, producing hundreds of megawatts, in Saudi Arabia, particularly in the Makkah Region.
The world’s technological and economic advancements have led to a sharp increase in the demand for electrical energy. Saudi Arabia is experiencing rapid economic and demographic growth, which is resulting in higher energy needs. The limits of fossil fuel reserves and their disruption to the environment have motivated the pursuit of alternative energy options such as wind energy. In order to regulate the power system to maintain safe and dependable operation, projections of current and daily power generation are crucial. Thus, this work focuses on wind power prediction and the statistical analysis of wind characteristics using wind data from a meteorological station in Makkah, Saudi Arabia. The data were collected over four years from January 2015 to July 2018. More than twelve thousand data points were collected and analyzed. Layout and wake effect studies were carried out. Furthermore, the near wake length downstream from the rotor disc between 1 and 5 rotor diameters (1D to 5D) was taken into account. Five robust machine learning algorithms were implemented to estimate the potential wind power production from a wind farm in Makkah, Saudi Arabia. The relationship between the wind speed and power produced for each season was carefully studied. Due to the variability in the wind speeds, the power production fluctuated much more in the winter. The higher the wind speed, the more significant the difference in energy production between the five farm layouts, and vice versa, whereas at a low wind speed, there was no significant difference in the power production in all of the near wake lengths of the 1D to 5D rotor diameters downstream from the rotor disc. Among the utilized prediction models, the decision tree regression was found to have the best accuracy values in all four utilized evaluation metrics, with 0.994 in R-squared, 0.025 in MAE, 0.273 in MSE, and 0.522 in RMSE. The obtained results were satisfactory and provide support for the construction of several wind farms, producing hundreds of megawatts, in Saudi Arabia, particularly in the Makkah Region.
Record ID
Keywords
Machine Learning, regression, Renewable and Sustainable Energy, wake effect, wind energy, wind power
Subject
Suggested Citation
Barashid K, Munshi A, Alhindi A. Wind Farm Power Prediction Considering Layout and Wake Effect: Case Study of Saudi Arabia. (2023). LAPSE:2023.20781
Author Affiliations
Barashid K: Computer Science Department, Umm Al-Qura University, Makkah 24211, Saudi Arabia
Munshi A: Computer Engineering Department, Umm Al-Qura University, Makkah 24382, Saudi Arabia; Smart Lab, Umm Al-Qura University, Makkah 24382, Saudi Arabia [ORCID]
Alhindi A: Computer Science Department, Umm Al-Qura University, Makkah 24211, Saudi Arabia [ORCID]
Munshi A: Computer Engineering Department, Umm Al-Qura University, Makkah 24382, Saudi Arabia; Smart Lab, Umm Al-Qura University, Makkah 24382, Saudi Arabia [ORCID]
Alhindi A: Computer Science Department, Umm Al-Qura University, Makkah 24211, Saudi Arabia [ORCID]
Journal Name
Energies
Volume
16
Issue
2
First Page
938
Year
2023
Publication Date
2023-01-13
ISSN
1996-1073
Version Comments
Original Submission
Other Meta
PII: en16020938, Publication Type: Journal Article
Record Map
Published Article

LAPSE:2023.20781
This Record
External Link

https://doi.org/10.3390/en16020938
Publisher Version
Download
Meta
Record Statistics
Record Views
551
Version History
[v1] (Original Submission)
Mar 20, 2023
Verified by curator on
Mar 20, 2023
This Version Number
v1
Citations
Most Recent
This Version
URL Here
http://psecommunity.org/LAPSE:2023.20781
Record Owner
Auto Uploader for LAPSE
Links to Related Works
(0.09 seconds)
[0.09 s]
