the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Wind speed time series synthesis using a parametrized power spectral density function
Ram C. Poudel
David Corbus
Ian Baring-Gould
Abstract. We propose a new method to synthesize 1 Hz wind speed and wind power time series data from the industry standard 10-minute wind turbine performance data. The method is based on a parameterized power spectral density (PSD) function decomposed into trend and random components. We illustrate the intra-timestep data synthesis utilizing 1 Hz data from two distributed wind turbines: CART3 (600 kW) and NPS (100 kW).
Ram C. Poudel et al.
Status: open (until 23 Jun 2023)
-
RC1: 'Comment on wes-2023-45', Anonymous Referee #1, 24 May 2023
reply
Overview
This manuscript proposes a statistical method for reconstructing 1 Hz wind speed and power time series using the 10-minute wind statistics. The method is based on a representing power spectral density function via the so-called trend and random components. However, this concept is not clear to the reviewer and is not properly described in this manuscript. Furthermore, the manuscript is not properly formatted (there are too many sections) and the quality of the figures is not satisfactory.
Specific comments
- A number of references are in capital letters. Is there a reason for that?
- Mathematical variables should be italic and written using an equation editor (e.g., Lines 31, 50, 51, 53, 54, 55, etc.).
- Figure 1. This figure is too small and it is very difficult (border impossible) to read labels, properly see time series, etc.
- Figure 1. What instrument was used to collect these 400 Hz data? At such a high acquisition frequency, I would expect more fluctuations in the wind speed. What is written on the y-axis? What is the orange (or red) line in the top row?
- The –5/3 law by Kolmogorov (not 5/3) only holds for the inertial subrange where the average properties of the turbulence are governed by the rate of dissipation. So, his model does not cover the entire energy spectrum, as your L70 might indicate.
- Broken reference.
- 10-5?
- L114–115. What does this stand for: “Apt, What has been learned from frequency domain analysis of wind and solar power, 2019?” Is that a statement or a very long reference?
- Equation 1. Explain all variables in this equation and subsequent in-line equations (e.g., N, n, k, t, i, etc.). While some of them might be obvious, they should still be formally defined.
- There should be a space between the number and the unit.
- Figure 2. There are two black lines in the graph. Which one is PSD(T) and which one is PSD(R)?
- Figure 2. If the authors are not using the evolutionary power spectral density, then how the PSD in Figure 2 (y-axis) can be time-dependent (x-axis)?
- Discussion around Eq. (4) and this whole method are not clear to this reviewer. Perhaps they are sound, but the authors need to do a better job at explaining the concept they are introducing. For example, this discussion does not properly explain the difference between constrained and unconstrained power spectral densities, and Figure 3 further complicates this whole issue.
- Equation (5). What is the source of this equation? If the authors came up with this expression, what data support it? exp is a function and should not be italic.
- ais?
- What is ue?
- If data are not statistically stationary, the proper spectral analysis is the evolutionary power spectral density and not the regular power spectral density. The authors should comment on how this statement relates to their “trend” power spectral density estimate.
- Section 4.3. This section is not clear to me and how does it contribute to, for example, Shiznozuka (1971) method (e.g. Equation 2 in this manuscript)?
- The quality of the figures in the rest of the manuscript is low. Some figures have grey regions outside of the graphs, which is not the standard publishing practice in journals.
- Overall, I find this manuscript very difficult to follow. The authors have too many sections—there are 9 sections and most of them have multiple subsections.
- I do not see how this method contributes to the body of literature compared to TurbSim or other methods that are used to reconstruct fluctuating wind components from mean speed and standard deviation.
Citation: https://doi.org/10.5194/wes-2023-45-RC1 -
AC1: 'Reply on RC1', Ram Poudel, 31 May 2023
reply
Response to Reviewers Comments: wes-2023-45
Dear Editor:
Thank you for all these valuable comments. Obviously, these comments help improve the clarity of idea/concept, organization of write-up, and eventually the quality of the manuscript. Our responses to specific comments are appended below (see attached PDF file).
Please let us know if you have any other comments and suggestions to improve the revised manuscript. Thank you very much!
Ram Poudel
Ram C. Poudel et al.
Ram C. Poudel et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 251 | 53 | 6 | 310 | 0 | 1 |
- HTML: 251
- PDF: 53
- XML: 6
- Total: 310
- BibTeX: 0
- EndNote: 1
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1