the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Aug 2023
Brief communication: A simple axial induction modification to WRF’s Fitch wind farm parameterisation
Abstract. We propose a modification to the Fitch wind farm parameterisation implemented in the Weather Research and Forecasting (WRF) model. This modification, derived from the 1D momentum theory, employs a wind speed dependent induction factor to correct the local grid wind speed back to free stream, before computing the turbine’s power and thrust. While the original implementation underestimates power, the modified version shows a better agreement with the power curve. We strongly recommend the modification to be employed for all studies that model not more than one turbine per WRF grid cell.
- Preprint
(654 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
CC1: 'Comment on wes-2023-89', Nicolai Gayle Nygaard, 30 Aug 2023
Very interesting results. Do you think the same induction factor correction is missing on other WRF wind farm parameterizations like EWP?
Disclaimer: this community comment is written by an individual and does not necessarily reflect the opinion of their employer.Citation: https://doi.org/10.5194/wes-2023-89-CC1 -
CC2: 'Reply on CC1', Lukas Vollmer, 01 Sep 2023
Dear Nicolai,
thank you for your question. Indeed we are only discussing proposed modifications to the Fitch parametrization in the introduction section, but not the also popular EWP parametrization. Unless there have been recent changes to the EWP the local wind speed reduction by induction is not considered either. We will add a sentence on this in the revised manuscript.
Disclaimer: this community comment is written by an individual and does not necessarily reflect the opinion of their employer.Citation: https://doi.org/10.5194/wes-2023-89-CC2
-
CC2: 'Reply on CC1', Lukas Vollmer, 01 Sep 2023
-
RC1: 'Comment on wes-2023-89', Patrick Hawbecker, 05 Dec 2023
The authors address a very important shortcoming of the highly-used wind farm parameterization within the WRF model. It has been fairly well known that the Fitch wind farm parameterization underestimates power. While other studies have attempted to overcome this issue by using correction factors, this study presents a generalizable fix by using the free-stream wind speed which should be very beneficial to the modeling and resource assessment community. I have a few comments of which I believe would be beneficial for the paper, but are not critical for publication. That said, I recommend the article for publication.
Specific comments
- Many of the uses for the wind farm parameterization in the wind energy industry involve the impact of wake effects on other wind farms or prospective lease areas. The final paragraph mentions how well the Fitch scheme does with wakes over land but does not mention how the new fix impacts wake resolution. Given that there are no observations for comparison, what does the difference in mean wind speed (over, say, the 12 hour analysis) look like between the Fitch-o, -AIF, and -mAIF cases? Are they drastically different or still quite similar? It will not be a quantitative result, but might shed light on if the wake impact will be large or small.
- The change to the Fitch parameterization now over-predicts power. It is much closer and clearly beneficial, but I am curious as to if the authors can explain why it is over-predicting power.
- This is a very interesting and potentially beneficial update to the Fitch parameterization. A short section of shortcomings and future work would be nice to add.
Technical Corrections
Line 95 “of and” should be “of”
Citation: https://doi.org/10.5194/wes-2023-89-RC1 -
RC2: 'Comment on wes-2023-89', Anonymous Referee #2, 06 Jan 2024
Review of the manuscript: “Brief communication: A simple axial induction modification to
WRF’s Fitch wind farm parameterization” by Lukas Vollmer, Balthazar Arnoldus Maria Sengers, and Martin Dörenkämper submitted for publication in Wind Energy Science.
In the manuscript “Brief communication: A simple axial induction modification to WRF’s Fitch wind farm parameterization” the authors present a modification to the Fitch wind plant parameterization implemented in the Weather Research and Forecasting model to account for the decrease in wind speed in the presence of a wind turbine compared to the free stream wind speed. The authors propose, implement, and demonstrate the effect of induction factor parameterization.
General Remarks
Wind plant parameterization play an important role accurate resource assessment and prediction of power production. The study presented in the manuscript therefore addresses a potentially important topic. The argument for a need to account for wind speed reduction in a presence of a wind turbine compared to the free stream wind speed and therefore for modifying a widely used parameterization are well justified. However, the presented study is lacking in a detailed analysis. Furthermore, it does not address some unexpected results, e.g., that a wind plant parameterization with an induction factor results in higher power output compared to the reference simulation which does not include a wind turbine. Finally, it includes an unrealistic simulation with five 22 MW wind turbines in a 2 km x 2 km grid cell results of which do not provide any useful insight. It is not clear why would one use such a case when there are available data from operating wind farms (e.g., Ali et al. 2023, https://doi-org.cuucar.idm.oclc.org/10.1175/MWR-D-23-0006.1) which can be used to compare wind plant parameterizations. Additional comments are provided blow under Specific Remarks.
Taking all the above into account in addition I do not recommend the manuscript for publication. However, I would encourage the authors to resubmit the manuscript after they redesign the study and include comparison with observed data in addition to providing a more detailed analysis, followed by potential refinement of the axial induction addressing over-prediction of power.
Specific Remarks
Line 5 – This statement can be confusing; it would be better to reword it as follows: “... model only one turbine per WRF gird cell.” However, the question is why is this not recommended if there are more turbines per grid cell?
Line 13 – It would be important to reference the original journal publication Skamarock and Klemp 2008, doi:10.1016/j.jcp.2007.01.037.
Line 24 – Instead of "growing" better would be "increasing."
Line 38 – It would be important to provide information about the surface layer, i.e. surface roughness parameterization used over the ocean?
Line 67 – However, A/(D*dx) is not that ratio - it should be A/(dx*dz). Either the name of the ratio should be modified, or its definition.
Figure 1 – The panels (b, e) are not very clear, it would be better to plot power difference w.r.t. reference vs. wind speed. Also, it seems that Fitch-AIF frequently results in more power than reference, how can that be explained or justified? The panels (c, f) are not very informative, the same information can be conveyed more succinctly with one simple sentence. However, it is not clear why is the power production higher with Fitch-AIF then reference.
Line 94 – It is not clear what is meant by “accounting”, is the same "accounting" applied to Fitch-o?
Line 96 – five 22 MW turbines in 2 km x 2 km, is beyond unrealistic. So why was this case selected if it is "not very realistic?"
Citation: https://doi.org/10.5194/wes-2023-89-RC2 - AC1: 'Author's response to wes-2023-89', Balthazar Sengers, 18 Jan 2024
Status: closed
-
CC1: 'Comment on wes-2023-89', Nicolai Gayle Nygaard, 30 Aug 2023
Very interesting results. Do you think the same induction factor correction is missing on other WRF wind farm parameterizations like EWP?
Disclaimer: this community comment is written by an individual and does not necessarily reflect the opinion of their employer.Citation: https://doi.org/10.5194/wes-2023-89-CC1 -
CC2: 'Reply on CC1', Lukas Vollmer, 01 Sep 2023
Dear Nicolai,
thank you for your question. Indeed we are only discussing proposed modifications to the Fitch parametrization in the introduction section, but not the also popular EWP parametrization. Unless there have been recent changes to the EWP the local wind speed reduction by induction is not considered either. We will add a sentence on this in the revised manuscript.
Disclaimer: this community comment is written by an individual and does not necessarily reflect the opinion of their employer.Citation: https://doi.org/10.5194/wes-2023-89-CC2
-
CC2: 'Reply on CC1', Lukas Vollmer, 01 Sep 2023
-
RC1: 'Comment on wes-2023-89', Patrick Hawbecker, 05 Dec 2023
The authors address a very important shortcoming of the highly-used wind farm parameterization within the WRF model. It has been fairly well known that the Fitch wind farm parameterization underestimates power. While other studies have attempted to overcome this issue by using correction factors, this study presents a generalizable fix by using the free-stream wind speed which should be very beneficial to the modeling and resource assessment community. I have a few comments of which I believe would be beneficial for the paper, but are not critical for publication. That said, I recommend the article for publication.
Specific comments
- Many of the uses for the wind farm parameterization in the wind energy industry involve the impact of wake effects on other wind farms or prospective lease areas. The final paragraph mentions how well the Fitch scheme does with wakes over land but does not mention how the new fix impacts wake resolution. Given that there are no observations for comparison, what does the difference in mean wind speed (over, say, the 12 hour analysis) look like between the Fitch-o, -AIF, and -mAIF cases? Are they drastically different or still quite similar? It will not be a quantitative result, but might shed light on if the wake impact will be large or small.
- The change to the Fitch parameterization now over-predicts power. It is much closer and clearly beneficial, but I am curious as to if the authors can explain why it is over-predicting power.
- This is a very interesting and potentially beneficial update to the Fitch parameterization. A short section of shortcomings and future work would be nice to add.
Technical Corrections
Line 95 “of and” should be “of”
Citation: https://doi.org/10.5194/wes-2023-89-RC1 -
RC2: 'Comment on wes-2023-89', Anonymous Referee #2, 06 Jan 2024
Review of the manuscript: “Brief communication: A simple axial induction modification to
WRF’s Fitch wind farm parameterization” by Lukas Vollmer, Balthazar Arnoldus Maria Sengers, and Martin Dörenkämper submitted for publication in Wind Energy Science.
In the manuscript “Brief communication: A simple axial induction modification to WRF’s Fitch wind farm parameterization” the authors present a modification to the Fitch wind plant parameterization implemented in the Weather Research and Forecasting model to account for the decrease in wind speed in the presence of a wind turbine compared to the free stream wind speed. The authors propose, implement, and demonstrate the effect of induction factor parameterization.
General Remarks
Wind plant parameterization play an important role accurate resource assessment and prediction of power production. The study presented in the manuscript therefore addresses a potentially important topic. The argument for a need to account for wind speed reduction in a presence of a wind turbine compared to the free stream wind speed and therefore for modifying a widely used parameterization are well justified. However, the presented study is lacking in a detailed analysis. Furthermore, it does not address some unexpected results, e.g., that a wind plant parameterization with an induction factor results in higher power output compared to the reference simulation which does not include a wind turbine. Finally, it includes an unrealistic simulation with five 22 MW wind turbines in a 2 km x 2 km grid cell results of which do not provide any useful insight. It is not clear why would one use such a case when there are available data from operating wind farms (e.g., Ali et al. 2023, https://doi-org.cuucar.idm.oclc.org/10.1175/MWR-D-23-0006.1) which can be used to compare wind plant parameterizations. Additional comments are provided blow under Specific Remarks.
Taking all the above into account in addition I do not recommend the manuscript for publication. However, I would encourage the authors to resubmit the manuscript after they redesign the study and include comparison with observed data in addition to providing a more detailed analysis, followed by potential refinement of the axial induction addressing over-prediction of power.
Specific Remarks
Line 5 – This statement can be confusing; it would be better to reword it as follows: “... model only one turbine per WRF gird cell.” However, the question is why is this not recommended if there are more turbines per grid cell?
Line 13 – It would be important to reference the original journal publication Skamarock and Klemp 2008, doi:10.1016/j.jcp.2007.01.037.
Line 24 – Instead of "growing" better would be "increasing."
Line 38 – It would be important to provide information about the surface layer, i.e. surface roughness parameterization used over the ocean?
Line 67 – However, A/(D*dx) is not that ratio - it should be A/(dx*dz). Either the name of the ratio should be modified, or its definition.
Figure 1 – The panels (b, e) are not very clear, it would be better to plot power difference w.r.t. reference vs. wind speed. Also, it seems that Fitch-AIF frequently results in more power than reference, how can that be explained or justified? The panels (c, f) are not very informative, the same information can be conveyed more succinctly with one simple sentence. However, it is not clear why is the power production higher with Fitch-AIF then reference.
Line 94 – It is not clear what is meant by “accounting”, is the same "accounting" applied to Fitch-o?
Line 96 – five 22 MW turbines in 2 km x 2 km, is beyond unrealistic. So why was this case selected if it is "not very realistic?"
Citation: https://doi.org/10.5194/wes-2023-89-RC2 - AC1: 'Author's response to wes-2023-89', Balthazar Sengers, 18 Jan 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 687 | 225 | 37 | 949 | 28 | 21 |
- HTML: 687
- PDF: 225
- XML: 37
- Total: 949
- BibTeX: 28
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
