the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Apr 2024
Challenges in Detecting Wind Turbine Power Loss: The Effects of Blade Erosion, Turbulence and Time Averaging
Abstract. Blade leading-edge erosion (LEE), atmospheric turbulence intensity (TI) and shear can significantly impact wind turbine performance and annual energy production (AEP). This study employs aeroelastic simulations to investigate their combined effects. An offshore original equipment manufacturer (OEM) provided aeroelastic model was used to simulate various scenarios. Turbulence intensity was varied for a range of wind speeds and the blade polars were modified to simulate different degrees of erosion, represented by varying levels of roughness. Also, simulations with and without the inclusion of wind shear were investigated. Findings reveal that even mild simulated erosion can reduce AEP by 0.82 %, while more severe erosion leads to a 2.83 % decrease. Increasing TI exacerbates these losses, with a 25 % TI causing up to a 3.5 % AEP reduction for eroded blades. These effects were most pronounced at lower wind speeds. Furthermore, standard time-averaging practices in power curve analyses can obscure the true magnitude of TI and LEE's impact on short-term power fluctuations. This work emphasises the critical importance of considering both blade condition and TI for accurate AEP assessments, optimal maintenance scheduling and improved wind turbine design in the context of site-specific atmospheric conditions.
- Preprint
(643 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2024-35', Anonymous Referee #1, 23 Apr 2024
Dear authors, the work presented is of good quality, and the data and results are interesting for both research and industry focused on AEP loss assessment due to blade erosion in wind turbines. However, I believe that some improvements are required to achieve the high quality necessary for publication in this journal. These improvements should primarily address the novelty and impact of the research and provide further clarification on the scientific reliability of certain aspects of the methodology.
1. Please take care of some typos and incorrect sentences:
- pg1. "data from Ding et al. (2022)."
- pg2. "strategies by Badihi et"
- pg2. "Furthermore, as wind turbines' rotors increase in length and tip speeds increase, see Hansen (2008)"
- pg20 "lower turbulence intensities are comparable to turbulence intensity."2. In the introduction and conclusions, the authors state the novelty of verifying the effect of TI and erosion on the prediction of AEP and performance losses using an aero-servo-elastic simulation tool. This is not new; similar results and comparisons have been presented, for example, in the following literature:
- Cappugi et al. "Machine learning-enabled prediction of wind turbine energy yield losses due to general blade leading edge erosion." Energy Conversion and Management 245 (2021): 114567.
- Castorrini et al. (2023) Assessing the progression of wind turbine energy yield losses due to blade erosion by resolving damage geometries from lab tests and field observations. Renewable Energy, 218, 119256.
- Campobasso et al. (2023) Probabilistic analysis of wind turbine performance degradation due to blade erosion accounting for uncertainty of damage geometry. Renewable and Sustainable Energy Reviews, 178, 113254.
- Papi et al. (2021). Uncertainty quantification on the effects of rain-induced erosion on annual energy production and performance of a Multi-MW wind turbine. Renewable Energy, 165, 701-715.
- Please assess the novelty of the approach or results with respect to what has already been presented and published in similar works.3. I understand the issue of working with proprietary data and models; however, the authors should indicate more clearly how the presented study and results, which are not replicable as the data from the case study are not disclosed and not generalizable as the results apply to a particular erosion type and distribution and to a specific WT model, can be beneficial for industry or research in this field.
4. At pg3: "Although the tested airfoil is not an identical match to that in the HAWC2 model, this approach is deemed a suitable approximation for representing the outboard region of eroded turbine blades." This sentence is not supported by data in the paper; it could be helpful to show that the power curve obtained with the original airfoil matches the power curve obtained with the substituted clean airfoil.
5. pg.3 "To represent the effects seen in the wind tunnel experiments, derived factors were used to approximate the results. For simplicity, the lift polar representing the clean airfoil was scaled by a factor of 0.9. Additionally, two artificial drag polars were created by scaling the drag polar representing the clean airfoil by factors of 1.5 and 2.0, respectively." This aspect of the methodology is not very clear and requires clearer explanation in the text. Furthermore, a proper justification should be reported in terms of the scientific soundness of the approach, by reference to existing literature proving the reliability of the approach, or by some verification test or evidence.
6. Figs 7, 8, 10, 12, 13 show many overlapping curves and are very difficult to read. I suggest reducing the number of entries to a subset of the most significant ones.
7. Please improve the clarity of captions for Tables 1 to 5. (From the captions of Table 1 and 2, it is not possible to understand the difference between the data presented in the two tables).
8. I understand it is part of the title and the research, however, the discussion about time averaging in the power curve is somewhat off-topic with respect to the other two. Indeed, TI and erosion are features related to the model of the case study, while the time averaging is a data processing parameter. Furthermore, I needed more than one reading to understand the point from section 3.4. I would suggest reducing this part, possibly moving it to an appendix, recalling the main outcome not in the results but in the presentation of the case study and methodology. This will improve the readability of the paper and also simplify the outcomes and conclusions.
9. The conclusion section needs some revision. It sometimes presents repeated information ("This research contributes valuable insights into the multifaceted effects of turbulence intensity, blade roughness, and time averaging on wind turbine performance") or conclusions that seem not really related to the data presented ("it highlights how data analysis techniques can either mask or reveal the subtle effects of erosion and turbulence"). Furthermore, a clearer conclusion should be drafted on the discussion about extracting information on erosion-related performance damage from in-field measurements and SCADA data.
Citation: https://doi.org/10.5194/wes-2024-35-RC1 -
AC1: 'Reply on RC1', Tahir Malik, 14 Jun 2024
Dear Peer Reviewer,
Thank you for your valuable and insightful comments on our manuscript. Your feedback has been instrumental in improving the quality and clarity of the paper.
Attached, you shall find a detailed response to each of your comments, along with an updated version of the manuscript reflecting these improvements.
Thank you once again for your thorough review and constructive suggestions.
Best regards,
Tahir Malik
-
AC1: 'Reply on RC1', Tahir Malik, 14 Jun 2024
-
RC2: 'Comment on wes-2024-35', Anonymous Referee #2, 23 Apr 2024
Paper review: “Challenges in Detecting Wind Turbine Power Loss: The effects of Blade Erosion, Turbulence and Time Averaging”.
The work makes an analysis of the effect over power production of erosion and turbulence intensity also including the impact of time averaging in the results. The work gives more insight into the complexity of power losses estimation, specially when turbulence intensity appears. The idea of isolating the effect of the different causes is well addressed. In this review, some improvements are suggested regarding the way the results are presented.
- Line 1: typo (TI)can -> (TI) can.
- Line 5: Erosion and roughness are closely related and sometimes their meanings are mixed up. In the abstract, please specify clearly why erosion is represented in this study with different roughens values.
- Line 20: energy losses of 7% (if it is AEP, please specify it).
- Line 30: strategiesBadihià strategies Badihi
- Line 30: (2022)Gonzalez -> (2022) Gonzalez
- Line 30: The sentence “ the lack of insight limits the potential benefits of quantifying APE loss” is difficult to understand, please rewrite it.
- Line 35: “ protection (LEP) and aerofoils more robust towards the effects of erosion” please give previous examples and references for LEP and robust airfoils.
- Section 2.2. Please in this section it is necessary to specify the % of blade in which roughness is considered. It is 15% but can be found in subsequent sections.
- Line 79: it is mentioned that the NACA airfoil used in Krog Kruse experiments it is not the same as the one in the HAWC2 model. Please specify if in the computations presented in the paper the airfoil used in the outer part of the blade is the NACA one or the one in the confidential wind turbine model.
- Line 87: please improve the explanation of the derived factors. If they are used in the airfoil coming from the HAWC2 model it is not very clear specified.
- Line 94: please specify what is the ‘plate behaviour’
- In Section 2.2 it is not clear whether the control system will be activated during all the simulations performed in this work. That is, describe if the control system detects that the power production is not achieved due to the affected airfoils will make any actions and mitigations.
- Line 169: ‘with imposed wind shear conditions’ : please explain which are these conditions.
- Figure 8: it is difficult to compare since the lines are in top of each other in 2 groups. Maybe a table or separating in different figures for each TI could help. The question that should be answered in this figure is: for all the TI studied the % of power loss is similar or depends on the TI value?
- Line 213: the sentence ‘ averaging effect of time averaging’ is a bit confusing for the reader. Please rewrite it.
- Figure 9 caption: change -> Change
- Figure 9: all the cases for P40 are computed for different TI, but the reference is always clean and TI 6%: is this consistent?
- Figure 12: In the legend ‘Clean 6% TI’ appears twice, please specify the difference between them in the text and in the legend.
- Line 272: It is suggested that the unexpected behavior detected for certain turbulence conditions that is presented in Table 1 is removed for the study. Once it is clarified it could be presented in future works.
- Table 1 caption: ‘ the same profile’ -> it is not clear which profile
- Table 1 and Table 2 captions: please include the velocity in this case to be consistent with Tables 3-4-5
- Line 398: these factors and these aspects appear several times in the paragraph.
- Conclusions section: In line 409 it is said that the impact of blade erosion was less significant and right after, ‘Blade roughness can significantly affect power-production’ Even thought it is ok, do not write them so close because is a bit confusing.
- Please review the whole conclusions section, it is very schematic.
- Competing interest sentence: the word ‘by’ appears twice and is a typo.
Citation: https://doi.org/10.5194/wes-2024-35-RC2 -
AC2: 'Reply on RC2', Tahir Malik, 14 Jun 2024
Dear Peer Reviewer,
Thank you for your valuable and insightful comments on our manuscript. Your feedback has been instrumental in improving the quality and clarity of the paper.
Attached, you shall find a detailed response to each of your comments, along with an updated version of the manuscript reflecting these improvements.
Thank you once again for your thorough review and constructive suggestions.
Best regards,
Tahir Malik
-
RC3: 'Comment on wes-2024-35', Anonymous Referee #3, 04 May 2024
WES 2024-35
The article explores the effects of erosion, turbulence intensity and averaging on wind turbine power curve. A realistic test case is used as wind data from an actual wind farm and an aeroelastic model of a real wind turbine are used. The methods used also appear reliable, as a state-of-the-art aeroelastic software tool is used. That said I have several concerns with the study.
First and most importantly the scientific archival value of the study is not apparent, and perhaps not strong enough for journal publication. Despite the large quantity of data presented in the study the main take away that I was able to see from the study is the fact that differences in power curve and AEP caused by erosion can be similar or even smaller in magnitude than the differences caused by turbulence intensity. This can make it hard to diagnose blade erosion in the real world, as the decrease in power output may be masked by variations in turbulence intensity. This message is interesting, but the large amount of data that is presented in the paper is redundant for this relatively straightforward message. In addition, authors do not suggest ways to work around this issue but the discussion in the paper is limited to presenting the data and little else. Finally, as the authors suggested, the large effect of turbulence on wind turbine power curve can already be found in the existing scientific literature, making the specific contribution of this work somewhat more unclear, and diminishing the scientific value.
To add to this, the data appears to be presented without a clear goal in mind. This makes the manuscript, despite it being divided in many subsections, very difficult to read. It is unclear what is the “glue” between the sections and how they contribute to the final take aways. In addition, many graphs are hard to understand (such as Figures 7, 13). Some other Figures are hard to tell apart – for instance it took me quite a while to understand the difference between figures 16-1 and 18-19.
Figure 11 and especially 10 are misleading, and if I have understood how they are computed, incorrect in my opinion. In fact, the Cp seems to be computed by dividing mean power by mean wind speed. This is incorrect, as Cp is an instantaneous value and should be computed based on instantaneous power and wind speed, and then averaged. Please explain how these values are computed. Authors attempt to warn readers about the high values of Cp in figure 10 at lines 237-241 but the explanation could be improved. The main reason for the large Cp values is the fact that wind turbine power near cut-in as a function of wind speed is cubic, thus increases in wind speed increase power more than decreases in wind speed do.
Detailed comments:
Please check the text carefully when more than one reference is inserted, as there is often no space between one and another (example at L119)
Figure 8 is very hare do understand. Figure 7 is too but perhaps to a lesser extent. I would suggest to divide figure 8 into more graphs, one per TI. However, I would also be open to other solutions.
L193: Differences are analysed at 11m/s. This wind speed is very close to rated. Some differences can be marked by the fact that some power regulation is already taking place at this wind speed. I would choose a lower wind speed (9-10 m/s) to highlight the differences in aerodynamic performance.
L201-203 I don’t understand where this is going… please rephrase
Section 3.2.2 Again, I don’t understand where this section is going. Why isn’t the effect of erosion and turbulence analysed separately? As it is the main take away for me is, once again, that turbulence makes it hard do detect power decreases, and this was already effectively explained in the previous sections.
L245: The main message is repeated again here
L273-274: A counter-intuitive finding is mentioned here, but this is not followed up with an explanation from what I can tell
L321: again the main take away of the paper is reinforced once again here. I am really struggling to see past this message, and thus don’t quite understand why all this data is needed to present this message.
L333-337: It would be interesting to explain this non-monotone trend
Figure 13: Very hard to understand, same comment as for Figure 8
Figure 16: I find It very hard to tell the difference between this and figure 14-15. Maybe change y label to something more explicit such as (P-P_clean^TI0)/P_clean^TI0. Also, what is the point of all these figures? They seem quite similar to one another and a bit redundant
Citation: https://doi.org/10.5194/wes-2024-35-RC3 -
AC3: 'Reply on RC3', Tahir Malik, 14 Jun 2024
Dear Peer Reviewer,
Thank you for your valuable and insightful comments on our manuscript. Your feedback has been instrumental in improving the quality and clarity of the paper.
Attached, you shall find a detailed response to each of your comments, along with an updated version of the manuscript reflecting these improvements.
Thank you once again for your thorough review and constructive suggestions.
Best regards,
Tahir Malik
-
AC3: 'Reply on RC3', Tahir Malik, 14 Jun 2024
-
EC1: 'Comment on wes-2024-35', Alessandro Bianchini, 04 May 2024
Dear Authors,
the reviews of your paper are now in. Please answer to the Reviewers' comments and then proceed with the preparation of a revised version of the study. In particular, pay attention to the comments of Reviewer 3, who raised some concerns on the analysis.
Citation: https://doi.org/10.5194/wes-2024-35-EC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 389 | 74 | 28 | 491 | 17 | 14 |
- HTML: 389
- PDF: 74
- XML: 28
- Total: 491
- BibTeX: 17
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1