the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Oct 2023
Measurement and analysis of high altitude wind profiles over the sea in a coastal zone using a scanning wind LiDAR – application to wind energy
Abstract. The lack of observations at heights relevant to the wind energy industry is a major challenge for the development of the next generation offshore wind turbines expected to operate within the first tens kilometres from the coast with turbine tip reaching more than 250 m. Observations in the coastal zone, complex by its very nature as being the site of sea breezes, low-level jets, land/sea transition, are key for both understanding the marine atmospheric boundary layer processes interacting with the turbine and to the parametrization of the wind profile well above the surface layer. These needs face the difficulties of measuring in the region 150–500 m and above the sea surface. In this paper, we present an original methodology to measure the 10-min averaged wind profile at 1.5 km offshore using a scanning doppler LiDAR (Light Detection And Ranging) installed inland. The validated methodology provides a well resolved vertical profile of the horizontal wind speed and direction up to 500 m above the sea. The methodology is implemented in a 7 month test campaign in the northeastern Atlantic coast. The analysis of the wind conditions shows a proportion of low-level jets, whose origin is discussed, of 15.5 % mainly coming from land and at night with a core well inside the rotor area of a 10 MW wind turbine. Wind shear events above the design values are observed 30 % of the time and provide a third of the total power production. High shear events are shown to be more probable during low-level jet (68 % of the time) compared to non low-level jet events (22 %). The description of low-level jets and high shear events is key as they are situations where the wind profile differs from the standard values used for wind turbine design and may affect the load and fatigue predictions.
- Preprint
(6146 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on wes-2023-141', Anonymous Referee #1, 15 Nov 2023
General comments
This paper analyses 7 months of scanning lidar measurements of winds at multiple heights at a coastal site in northwest France, with the aim of characterising the vertical profile, assessing vertical wind shear and the frequency of low-level jets (LLJs) to understand their impact on wind energy. The manuscript presents relevant and interesting analysis, however the novelty of the study is unclear and the aims and key points could be more clearly defined and emphasised throughout the text to strengthen the narrative. One major concern is that the lidar measurements have not been independently verified. It is recommended that this and several further points below should be considered to improve the manuscript before publication.
Specific comments
- Measurement reliability, validation/verification: a key issue is that the scanning lidar measurements have not been independently verified against an independent reference source and therefore it is unknown how reliable the measurements are (and therefore that the measurement setup is correct). If there is an absence of any other in-situ reference sources near the site, the study could at least compare the measurements with reanalysis (e.g. ERA5) to show that they agree in terms of the overall wind conditions.
- Measurement methodology: the text says that this methodology is original but it is not clear from the text (until the conclusion) how the method presented is actually new and differs from a standard wind field reconstruction of PPI scans observed at different elevation angles.
- Wind profiles: A major aim is to analyse high altitude wind profiles, however no wind profiles (ie. speed vs height) are presented. The manuscript could benefit from showing the mean profiles over the period, and greater insight would be provided by showing these relative to night, day, during LLJ, without LLJ etc, as is partitioned in the other analysis. Analysis of wind veer would also add further insight of relevance to wind energy.
- The application section (S3.5) is weak. Wind energy is advancing towards larger turbines and therefore the section should additionally assess a larger turbine(s) (e.g. IEA 15MW or 18MW) given that the measurements presented specifically go to higher heights which are directly relevant for such turbines. The study also has the advantage of the detailed wind profile observations which can be used to derive rotor equivalent wind speed (e.g. https://wes.copernicus.org/articles/5/1169/2020/) which is more appropriate for the analysis, and would provide a more accurate understanding on the impact of shear on power production across the rotor area(s).
- Seasonal biasing: The measurements are limited to 7-month period centred on the Summer season and therefore are seasonally biased. Moreover, this part of the year is the period of lowest wind speeds and therefore lowest wind power production. The implications of this should therefore be discussed and caveated.
- Meteorological conditions: the results would benefit from incorporating additional analysis of the meteorological conditions during measurement period (e.g. temp, pressure, precipitation, humidity, atmospheric stability). This would provide deeper insight in physically interpreting the wind profiles, vertical shear and LLJ events and help backup the physical reasoning made to explain the results observed. If no in-situ measurements are available then reanalysis could be used to examine this.
- Introduction wrt low level jet importance: A key aim of the paper is to characterise LLJ frequency but the introduction does not explain clearly how LLJ specifically impact wind power and wind turbine performance and exactly what other studies have found. For example, the term ‘quite frequent’ describing LLJ frequency from literature (line 48) is extremely vague.
- Section 2:
- Horizontal homogeneity: the scanning lidar wind field reconstruction method relies upon the assumption of horizontal flow homogeneity. In Section 2.2.2 it should be more clearly emphasised that the comparison made shows that the wind fields measured by the scanning lidar are relatively homogeneous. By definition, it appears that the NRMSE threshold filter essentially removes heterogenous flow cases which is why the correlations in Section 2.2.2 are so high? This should be commented on. Also, it would be possible to use the data to obtain an estimate of how many horizontally heterogenous vs homogeneous flow events are present in the measurements. This would be a very interesting result to report, especially given the land-sea coastal transition zone, and help further back up the homogeneity assumption.
- Figure 4: standard practice for validation is to include the results from a linear regression.
- The backlash effect on the scan head may impact the pointing accuracy especially when scanning at multiple elevations. Has this been examined and mitigated?
- Section 2.1: more details should be given about the hard target procedure as different methods exist and Shimada et al., (2020) used a specific approach called ‘soft target calibration’. This is important as the viability of the measurement heights relies on this process.
- Line 106 – Please clarify in the text that -29dB is the minimum CNR threshold applied. Normally an upper CNR threshold is also applied to ensure that any high returns from hard objects are removed.
- Does the obtaining u and v from the linear regression of Eqn 3. yield the same results as the Shimada et al., (2020) method? If not, the statements of reported accuracy (119-121) may not hold.
- Section 2.2.3: Please provide a physical explanation of why the USA exhibits a high sensitivity to wind speed and the scanning lidar a low sensitivity.
- Section 3:
- This section would benefit from explaining to the reader more clearly why specific results/statistics are being presented and physically what the results mean.
- Section 3.1: It is not clear what the percentage frequencies in Table 2 are calculated relative to. Are % reliable profiles defined as number of reliable profiles relative to the full time period? Is the LLJ occurrence a percentage relative to the number of profiles observed or relative to the full time period (i.e. total number of 10-min periods in a month). These definitions need to be made clearer in order to interpret the statistics properly.
- The low data availability during July and August is a concern as it brings into question how representative the measurements are of this period and therefore of the LLJ frequency. Please discuss and caveat this.
- Frequencies in histograms might be easier to interpret if y-axis (counts) were expressed as a percentage relative to the observed period.
- Section 3.4: would be useful to comment on why LLJ are linked to high shear (stability?).
Technical corrections
Please re-check the grammar throughout, listed below are only some examples:
Line 3 – ‘tip’ should be ‘tips’
Line 21 – ‘form’ should be ‘from’
Line 22 – ‘to’ – reads better is it’s ‘from’
Line 24 – Please define the approximate distance range from the coast within which the meteorological land/sea (coastal) transition region occupies and add reference(s) to back this up.
Line 38 – ‘observation’ should be ‘observations’
Line 45 – ‘Older and more recent studies’ change to ‘Previous studies’
Lines 45-48 – for readability the list of references would be better moved to a bracketed list at the end of the sentence.
Line 51 – references stated wrongly – should be ‘(Soares et al., 2014; Svenson et al, 2019)
Line 55 – what ‘consequences’ are expected of LLJs on wind energy in terms of wind power generation and mechanical loads? Need to be more specific so that the motivation of studying them is clearly identified.
Line 64 – please define what is ‘low data availability’ from the sLidar in the Wagner study – the statement ‘data availability was rather low’ is too vague. Also include the frequency of LLJs that this study found
Line 101-102 – rotation direction – this is more clear if changed from ‘direct and indirect’ to ‘clockwise and anticlockwise’ (whatever way around is correct).
Line 178 – what is meant by the ‘cord’ of a scan?
Section 2.2.3: lots of equations are given within the text body which makes the difficult to track, and therefore it might be clearer to define equations on separate lines. Define z_alpha, k and z.
Citation: https://doi.org/10.5194/wes-2023-141-RC1 -
RC2: 'Comment on wes-2023-141', Anonymous Referee #2, 18 Nov 2023
This paper presents an interesting study of the use of a scanning lidar for profiling the wind speed close to the shore. It also presents a climatology of the low level jets and an analysis of the resulting energy yield from a wind turbine. There are some points which should be addressed before publication:
- The novelty of the study should be emphasised. From the background in the paper, it would seem that this is not the first time a scanning lidar has been used to determine profiles. Also, has there been any previous work to look at the climatology of low level jets in this part of the Atlantic cost?
- As mentioned in this paper, there have been studies of LLJs in neighbouring areas such as the North Sea (e.g. https://wes.copernicus.org/articles/4/193/2019/). A quantification of such studies would be beneficial for comparison with this Atlantic site.
- The construction of a profile ought to be validated with a profiling lidar, e.g. on a floating platform. Has this been done previously? Is there any scope for validation with the set-up here?
- The work suggests that there is not much horizontal variation in the wind speed within 500m of the intended measurement point. Whilst the paper acknowledges that turbulent fluctuations will play a role in the difference between the validation pairs, there is still quite some scatter from the regression lines. At such a relatively short distance from the coast, a developing marine internal boundary layer is likely to affect the correlation between measurements depending on the prevailing stability conditions and wind direction (land/sea, sea/land). The work would benefit from an analysis of how the correlation varies with these parameters. In the case of stability, if actual measurements are not available, this could be inferred from a reanalysis.
- As mentioned by another reviewer, inclusion of some actual profiles (instantaneous and/or mean) would be beneficial, e.g. by time of day, season, direction, LLJ, non-LLJ.
- Some of the terminology is not consistent, e.g. ‘shore wind’ in figure 6c rather than ‘land wind’ (presumably?).
- It is stated that the day/night and sea/land wind distributions are very similar. Some explanation is given, but it is acknowledged that no directional analysis was done (line 258). It would seem highly likely that the winds are coming from the land at night due to thermal effects and thus most of the LLJs also will be coming from the land at night. This ought to be a simple analysis to do and should be included.
- Line 239: I assume this should refer to ‘sea wind’ and land wind’?
- In figure 9, the blue boxes, whiskers and circles are not defined or if they are, I didn’t fully understand the explanation.
- A more recent turbine (larger) would be better to use, e.g. IEA 15MW or even 22MW as these are more representative of the next generation turbines being sited offshore.
- The analysis of the impact of jets would be more enlightening if the profile shape was taken into consideration (e.g. rotor equivalent wind speed). The link to the mean speed is interesting but is quite a limited analysis and not a direct consequent of the jet shape, only its magnitude at hub height which is quite localised (particular to this site) and also dependent on the nine months of data.
- The English could be improved in places, e.g. the verb ‘to allow’ does not catenate, i.e. cannot be followed directly by ‘to’ + infinitive.
Citation: https://doi.org/10.5194/wes-2023-141-RC2 - AC1: 'Comment on wes-2023-141', Boris Conan, 21 Jan 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 377 | 134 | 10 | 521 | 10 | 11 |
- HTML: 377
- PDF: 134
- XML: 10
- Total: 521
- BibTeX: 10
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1