the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Feb 2025
Source-load matching and energy storage optimization strategies for regional wind-solar energy systems
Abstract. In response to the issue of limited new energy output leading to poor smoothing effects on grid-connected load fluctuations, this paper proposes a load power smoothing method based on "one source multiple loads." The method comprehensively considers the proximity of the source and load, as well as the correlation between their power fluctuations, using this as a tracking evaluation standard for source-side and different load-side matching in regional power grids. Initially, loads are clustered and divided based on power frequency division. The EEMD algorithm is then applied to obtain wind and solar energy outputs with greater complementarity and smoother fluctuations, leveraging their low-frequency correlation. Subsequently, a load tracking coefficient is used to compare the matching degree between wind-solar power output and different loads, selecting the most compatible load and output for source-load matching and smoothing. Concurrently, a gray wolf optimization algorithm based on Tent-chaotic mapping is employed to optimize edge energy storage at different load sides, minimizing overall grid-connected load power fluctuations. Numerical results demonstrate that the proposed method can fully utilize the stable output from the low-frequency correlation of wind and solar energy, combined with energy storage, to significantly reduce the fluctuation rate of regional grid-connected loads. This effectively promotes local absorption of source loads, thereby alleviating the pressure on the grid side caused by the randomness and volatility on both sides of the source load.
- Preprint
(738 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on wes-2025-11', Anonymous Referee #1, 12 Mar 2025
In the paper “Source-load matching and energy storage optimization strategies for regional wind-solar energy system”, the authors present a collection of methods to better match wind and solar generation to demand. The objective is important, and some parts of the presented analyses are interesting. However, the overall flow of the analyses and data is not clearly presented, the use of different methods is not justified, and the presented case study is very limited.
- The authors state: “We prioritize the more stable low-frequency output of wind-solar to match load power fluctuations according to load tracking criteria” (page 2). Why is the focus on low frequencies? Please explain and justify.
- The authors mention: “…use Euclidean distance to judge the scenario set scenario and the original output of the corresponding day, and select the closest day as the replacement of the output of that day” (page 3). Why is a day replaced? Why is such replacement needed? Please explain and justify.
- As a continuation from the previous comment, the authors also say: “The load group with the highest source-load matching degree is selected as the output satisfaction object for that day”. Why is load clustered and such matching done? Why is this step needed in your analyses? Please explain and justify.
- The authors use many methods, such as complementary ensemble empirical mode decomposition, K-means clustering, firefly optimization, gray wolf optimization and Tent-chaotic mapping. However, the reasons for using these methods are not clear and they are not compared to other potential methods. Please explain what is the reason for using each method, and why you chose these specific methods over some other potential methods.
- The authors state: “Based on the fact that Spearman's coefficient and Euclidean distance present complementary advantages and disadvantages in measuring correlation” (page 6). What are these advantages and disadvantages and why are they complementary?
- The authors present many methods and approaches in Sections 2 and 3. However, to me it’s not clear what is the role of the different methods, why the different analysis parts are needed and what is the overall flow of the analyses. Please better justify why the different parts of the analysis are needed, how they relate to each other and how they contribute to the overall objective of the paper. I suggest adding a flow diagram linking all the analysis parts, including data flows, to the paper.
- The case study presented in Section 4 seems very limited, with only 1 wind power plant and 1 solar plant analyzed for 1 summer. The experiments should be broadened to cover at minimum a full year and even better many years. And at least a few wind and solar power plant cases should be studied. Currently, the case study does not consider seasonality in wind and solar (only summer is considered), and I consider analyzing only 1 weather year and 1 wind and solar plant too limited to draw meaningful conclusions.
- 3 is not clear. Please better describe what is shown in the figure. E.g., why there are 6 scenarios and why the time interval is between 90 and 100 min? What can we conclude from the shown scenarios?
- 4 is not clear to me. Why there seem to always be a minimum generation of around 19 MW? Please also better describe what we see in the figure.
- 5 is unclear. Add y-axis labels. What are the different sub-plots?
Citation: https://doi.org/10.5194/wes-2025-11-RC1 -
AC1: 'Reply on RC1', Yongqing Zhu, 03 Apr 2025
1. The authors state: “We prioritize the more stable low-frequency output of wind-solar to match load power fluctuations according to load tracking criteria” (page 2). Why is the focus on low frequencies? Please explain and justify.
Answer: The focus on low frequencies is justified by the fact that wind-solar outputs exhibit stronger complementarity in the low-frequency band (e.g., daily/seasonal variations), which are more predictable and stable than high-frequency fluctuations (e.g., hourly gusts/clouds). This aligns with load demand patterns, which also have significant low-frequency components (e.g., diurnal/weekly cycles). By leveraging low-frequency correlation, we reduce volatility and enhance grid stability. Details are expanded in Section 2.1 and the Introduction.
2. The authors mention: “…use Euclidean distance to judge the scenario set scenario and the original output of the corresponding day, and select the closest day as the replacement of the output of that day” (page 3). Why is a day replaced? Why is such replacement needed? Please explain and justify.
Answer: The replacement aims to generate representative scenarios of wind-solar output for source-load matching. By selecting the closest historical day via Euclidean distance, we ensure the generated scenario captures realistic variability while minimizing errors. This is critical for optimizing energy storage and load smoothing, as described in Section 2.1.
3. As a continuation from the previous comment, the authors also say: “The load group with the highest source-load matching degree is selected as the output satisfaction object for that day”. Why is load clustered and such matching done? Why is this step needed in your analyses? Please explain and justify.
Answer: Load clustering groups similar consumption patterns to improve source-load matching efficiency. By categorizing loads into clusters, we identify the most compatible load group for daily wind-solar output, reducing overall grid fluctuations. This approach is validated in Section 4, where clustered loads show higher matching degrees (e.g., Table 1).
4. The authors use many methods, such as complementary ensemble empirical mode decomposition, K-means clustering, firefly optimization, gray wolf optimization and Tent-chaotic mapping. However, the reasons for using these methods are not clear and they are not compared to other potential methods. Please explain what is the reason for using each method, and why you chose these specific methods over some other potential methods.
Answer: CEEMD: Used for noise-free frequency decomposition, addressing modal mixing in EMD (Section 2.1).
Firefly-K-means: Improves clustering robustness and avoids local optima (Section 2.2).
Gray Wolf Optimization with Tent-chaotic Mapping: Enhances global search for energy storage optimization, outperforming traditional GWO (Section 3.1).
Justifications for each method are explicitly added in their respective sections.
5. The authors state: “Based on the fact that Spearman's coefficient and Euclidean distance present complementary advantages and disadvantages in measuring correlation” (page 6). What are these advantages and disadvantages and why are they complementary?
Answer: Advantages: Spearman measures monotonic correlation (rank-based), while Euclidean distance quantifies magnitude proximity.
Disadvantages: Spearman ignores magnitude, and Euclidean distance 忽略 trend direction.
Complementarity: Combining them balances correlation strength and magnitude alignment (Section 2.3).
6. The authors present many methods and approaches in Sections 2 and 3. However, to me it’s not clear what is the role of the different methods, why the different analysis parts are needed and what is the overall flow of the analyses. Please better justify why the different parts of the analysis are needed, how they relate to each other and how they contribute to the overall objective of the paper. I suggest adding a flow diagram linking all the analysis parts, including data flows, to the paper.
Answer: A flow diagram (Figure 2) is added to visualize the workflow, including data decomposition, clustering, matching, and optimization steps. The interconnections between methods are explicitly described in Section 2.
7. The case study presented in Section 4 seems very limited, with only 1 wind power plant and 1 solar plant analyzed for 1 summer. The experiments should be broadened to cover at minimum a full year and even better many years. And at least a few wind and solar power plant cases should be studied. Currently, the case study does not consider seasonality in wind and solar (only summer is considered), and I consider analyzing only 1 weather year and 1 wind and solar plant too limited to draw meaningful conclusions.
Answer: There is relatively little source load matching between two similar wind and solar power plants. Currently, this article proposes a feasible case, and we will continue to explore more matching cases in the future. While the current study uses summer data for simplicity, we acknowledge the need for broader validation. Future work will include multi-year datasets and multiple plants to account for seasonality and diverse conditions. This is noted in the Conclusion.
8. 3 is not clear. Please better describe what is shown in the figure. E.g., why there are 6 scenarios and why the time interval is between 90 and 100 min? What can we conclude from the shown scenarios?
4 is not clear to me. Why there seem to always be a minimum generation of around 19 MW? Please also better describe what we see in the figure.
5 is unclear. Add y-axis labels. What are the different sub-plots?
Answer: Figure 3: x-axis labels added (time in hours). Scenarios are generated via CEEMD-based clustering.
Figure 4: Explained as source-load matching results, showing the selected load group with the highest degree (Section 4).
Figure 5: y-axis labels added (SOC [%]). Subplots show energy storage states post-optimization (Section 4).
Citation: https://doi.org/10.5194/wes-2025-11-AC1
-
RC2: 'Comment on wes-2025-11', Anonymous Referee #2, 21 Mar 2025
Overview:
This paper proposes a regional grid management by taking into account renewable energy production and energy storage management to reduce the fluctuations in the power load and dispatch pressure on the electrical grid. The paper utilizes a “one source multiple load” framework that matches up the loads with the low pass frequency of the renewable assets to utilize the renewable assets most efficiently. The paper is interesting and presents a valuable method for dispatching multiple energy storage systems to meet the load while also taking into account the health of the battery. I would recommend publishing the paper after the major comments below have been addressed to increase the clarity and impact of the paper.
Major Comments:
Line 98: What does “Repeat the above steps n times.” mean? Could the authors elaborate on how they choose n, and what it’s based on?
Line 153-155: There seems to be a sentence or two missing on what Equation 14 represents and the sentence afterwards is unclear.
Fig 3: Could the authors explain this figure more? What does the x-axis mean here? How did they obtain these figures?
Lines 262-268: This section is unclear. Are the authors referencing a previous paper (please cite this here), or are they talking about their own results? Figure 4 is never reference in the text.
Line 282: Here, the text references Fig. 5 and talks about the overall load power, however Fig. 5 shows the SOC of the energy storage systems. Were the authors referring to Fig. 6 here?
Fig. 7 and Lines 292-296: Could you give more statistics from this plot? Are we trying to reduce the peaks of this signal, or the average?
Fig. 8: Can you give the load fluctuation rate for Scenario 2 and Scenario 3?
Line 315-316: Is it obvious that the overall fluctuation from Scenario 3 is less than Scenario 2? How is this calculated?
Fig. 10 & 11: Can these figures be made larger? They are very hard to read.
Minor Comments:
- Figures 1 & 2 are small. Could the size of these be increased to aid readability?
- Line 67: “the scenario set scenario and the original output of the corresponding day” – Typo? Should scenario be in there twice? This is a little hard to understand.
- Line 90: Fix formatting of Gaussian white noise symbols in text
- Line 131: “is thematch between the source-side…” Typo
- Line 146: “with the other distributionsand the probability…” typo
- Line 179: Fix symbol formatting – this is recurring, check symbol formatting throughout the paper
- Line 229: “To better protect the energy storage and prolong the life of the storage.” Sentence fragment. What is this referring to?
Citation: https://doi.org/10.5194/wes-2025-11-RC2 -
AC2: 'Reply on RC2', Yongqing Zhu, 03 Apr 2025
1. Line 98: What does “Repeat the above steps n times.” mean? Could the authors elaborate on how they choose n, and what it’s based on?
Answer: n is chosen based on convergence criteria (e.g., 50 iterations to ensure noise cancellation). This is specified in Section 2.1.
2. Line 153-155: There seems to be a sentence or two missing on what Equation 14 represents and the sentence afterwards is unclear.
Answer: Equation 14 normalizes Spearman coefficients to [0,1], ensuring compatibility with Euclidean distance. Details are added in Section 2.3.
3. Fig 3: Could the authors explain this figure more? What does the x-axis mean here? How did they obtain these figures?
Answer: Figure 3 illustrates wind-solar scenario generation via CEEMD. The x-axis represents time (hours), and scenarios are derived from decomposed low-frequency components (Section 4).
4. Lines 262-268: This section is unclear. Are the authors referencing a previous paper (please cite this here), or are they talking about their own results? Figure 4 is never reference in the text.
Answer: There are no citations to the previous article, referring to the methods described earlier in this article.
Figure 4 is now referenced in Section 4, explaining source-load matching results.
5. Line 282: Here, the text references Fig. 5 and talks about the overall load power, however Fig. 5 shows the SOC of the energy storage systems. Were the authors referring to Fig. 6 here?
Answer: The text now correctly references Figure 6 for load power changes and Figure 5 for SOC states (Section 4).
6. Fig. 7 and Lines 292-296: Could you give more statistics from this plot? Are we trying to reduce the peaks of this signal, or the average?
Answer: Fluctuation rate metrics (e.g., standard deviation) are added in Section 4, showing a 50% reduction in volatility post-optimization.
7. Fig. 8: Can you give the load fluctuation rate for Scenario 2 and Scenario 3?
Answer: Fluctuation rates for Scenario 2 (0.35) and Scenario 3 (0.18) are explicitly stated in Section 4.
8. Fig. 10 & 11: Can these figures be made larger? They are very hard to read.
Answer: Figures 10 and 11 are enlarged, and labels are enhanced for clarity.
9. Line 315-316: Is it obvious that the overall fluctuation from Scenario 3 is less than Scenario 2? How is this calculated?
Answer: The results demonstrate that integrating source-load matching with optimized energy storage (Scenario 3) reduces load volatility more effectively than direct wind-solar utilization (Scenario 2). The calculation method (Equation 20) mathematically validates these improvements by quantifying deviations from the mean load profile.
Minor Comments:
- Figures 1 & 2 are small. Could the size of these be increased to aid readability?
- Line 67: “the scenario set scenario and the original output of the corresponding day” – Typo? Should scenario be in there twice? This is a little hard to understand.
- Line 90: Fix formatting of Gaussian white noise symbols in text
- Line 131: “is the match between the source-side…” Typo
- Line 146: “with the other distributions and the probability…” typo
- Line 179: Fix symbol formatting – this is recurring, check symbol formatting throughout the paper
- Line 229: “To better protect the energy storage and prolong the life of the storage.” Sentence fragment. What is this referring to?
Answer:
- Typos and formatting issues (e.g., Equations 1, 2, 15) are corrected.
- Symbol formatting (e.g., superscripts/subscripts) is standardized.
- Sentence fragments (e.g., Line 229) are revised for clarity.
Citation: https://doi.org/10.5194/wes-2025-11-AC2
Status: closed
-
RC1: 'Comment on wes-2025-11', Anonymous Referee #1, 12 Mar 2025
In the paper “Source-load matching and energy storage optimization strategies for regional wind-solar energy system”, the authors present a collection of methods to better match wind and solar generation to demand. The objective is important, and some parts of the presented analyses are interesting. However, the overall flow of the analyses and data is not clearly presented, the use of different methods is not justified, and the presented case study is very limited.
- The authors state: “We prioritize the more stable low-frequency output of wind-solar to match load power fluctuations according to load tracking criteria” (page 2). Why is the focus on low frequencies? Please explain and justify.
- The authors mention: “…use Euclidean distance to judge the scenario set scenario and the original output of the corresponding day, and select the closest day as the replacement of the output of that day” (page 3). Why is a day replaced? Why is such replacement needed? Please explain and justify.
- As a continuation from the previous comment, the authors also say: “The load group with the highest source-load matching degree is selected as the output satisfaction object for that day”. Why is load clustered and such matching done? Why is this step needed in your analyses? Please explain and justify.
- The authors use many methods, such as complementary ensemble empirical mode decomposition, K-means clustering, firefly optimization, gray wolf optimization and Tent-chaotic mapping. However, the reasons for using these methods are not clear and they are not compared to other potential methods. Please explain what is the reason for using each method, and why you chose these specific methods over some other potential methods.
- The authors state: “Based on the fact that Spearman's coefficient and Euclidean distance present complementary advantages and disadvantages in measuring correlation” (page 6). What are these advantages and disadvantages and why are they complementary?
- The authors present many methods and approaches in Sections 2 and 3. However, to me it’s not clear what is the role of the different methods, why the different analysis parts are needed and what is the overall flow of the analyses. Please better justify why the different parts of the analysis are needed, how they relate to each other and how they contribute to the overall objective of the paper. I suggest adding a flow diagram linking all the analysis parts, including data flows, to the paper.
- The case study presented in Section 4 seems very limited, with only 1 wind power plant and 1 solar plant analyzed for 1 summer. The experiments should be broadened to cover at minimum a full year and even better many years. And at least a few wind and solar power plant cases should be studied. Currently, the case study does not consider seasonality in wind and solar (only summer is considered), and I consider analyzing only 1 weather year and 1 wind and solar plant too limited to draw meaningful conclusions.
- 3 is not clear. Please better describe what is shown in the figure. E.g., why there are 6 scenarios and why the time interval is between 90 and 100 min? What can we conclude from the shown scenarios?
- 4 is not clear to me. Why there seem to always be a minimum generation of around 19 MW? Please also better describe what we see in the figure.
- 5 is unclear. Add y-axis labels. What are the different sub-plots?
Citation: https://doi.org/10.5194/wes-2025-11-RC1 -
AC1: 'Reply on RC1', Yongqing Zhu, 03 Apr 2025
1. The authors state: “We prioritize the more stable low-frequency output of wind-solar to match load power fluctuations according to load tracking criteria” (page 2). Why is the focus on low frequencies? Please explain and justify.
Answer: The focus on low frequencies is justified by the fact that wind-solar outputs exhibit stronger complementarity in the low-frequency band (e.g., daily/seasonal variations), which are more predictable and stable than high-frequency fluctuations (e.g., hourly gusts/clouds). This aligns with load demand patterns, which also have significant low-frequency components (e.g., diurnal/weekly cycles). By leveraging low-frequency correlation, we reduce volatility and enhance grid stability. Details are expanded in Section 2.1 and the Introduction.
2. The authors mention: “…use Euclidean distance to judge the scenario set scenario and the original output of the corresponding day, and select the closest day as the replacement of the output of that day” (page 3). Why is a day replaced? Why is such replacement needed? Please explain and justify.
Answer: The replacement aims to generate representative scenarios of wind-solar output for source-load matching. By selecting the closest historical day via Euclidean distance, we ensure the generated scenario captures realistic variability while minimizing errors. This is critical for optimizing energy storage and load smoothing, as described in Section 2.1.
3. As a continuation from the previous comment, the authors also say: “The load group with the highest source-load matching degree is selected as the output satisfaction object for that day”. Why is load clustered and such matching done? Why is this step needed in your analyses? Please explain and justify.
Answer: Load clustering groups similar consumption patterns to improve source-load matching efficiency. By categorizing loads into clusters, we identify the most compatible load group for daily wind-solar output, reducing overall grid fluctuations. This approach is validated in Section 4, where clustered loads show higher matching degrees (e.g., Table 1).
4. The authors use many methods, such as complementary ensemble empirical mode decomposition, K-means clustering, firefly optimization, gray wolf optimization and Tent-chaotic mapping. However, the reasons for using these methods are not clear and they are not compared to other potential methods. Please explain what is the reason for using each method, and why you chose these specific methods over some other potential methods.
Answer: CEEMD: Used for noise-free frequency decomposition, addressing modal mixing in EMD (Section 2.1).
Firefly-K-means: Improves clustering robustness and avoids local optima (Section 2.2).
Gray Wolf Optimization with Tent-chaotic Mapping: Enhances global search for energy storage optimization, outperforming traditional GWO (Section 3.1).
Justifications for each method are explicitly added in their respective sections.
5. The authors state: “Based on the fact that Spearman's coefficient and Euclidean distance present complementary advantages and disadvantages in measuring correlation” (page 6). What are these advantages and disadvantages and why are they complementary?
Answer: Advantages: Spearman measures monotonic correlation (rank-based), while Euclidean distance quantifies magnitude proximity.
Disadvantages: Spearman ignores magnitude, and Euclidean distance 忽略 trend direction.
Complementarity: Combining them balances correlation strength and magnitude alignment (Section 2.3).
6. The authors present many methods and approaches in Sections 2 and 3. However, to me it’s not clear what is the role of the different methods, why the different analysis parts are needed and what is the overall flow of the analyses. Please better justify why the different parts of the analysis are needed, how they relate to each other and how they contribute to the overall objective of the paper. I suggest adding a flow diagram linking all the analysis parts, including data flows, to the paper.
Answer: A flow diagram (Figure 2) is added to visualize the workflow, including data decomposition, clustering, matching, and optimization steps. The interconnections between methods are explicitly described in Section 2.
7. The case study presented in Section 4 seems very limited, with only 1 wind power plant and 1 solar plant analyzed for 1 summer. The experiments should be broadened to cover at minimum a full year and even better many years. And at least a few wind and solar power plant cases should be studied. Currently, the case study does not consider seasonality in wind and solar (only summer is considered), and I consider analyzing only 1 weather year and 1 wind and solar plant too limited to draw meaningful conclusions.
Answer: There is relatively little source load matching between two similar wind and solar power plants. Currently, this article proposes a feasible case, and we will continue to explore more matching cases in the future. While the current study uses summer data for simplicity, we acknowledge the need for broader validation. Future work will include multi-year datasets and multiple plants to account for seasonality and diverse conditions. This is noted in the Conclusion.
8. 3 is not clear. Please better describe what is shown in the figure. E.g., why there are 6 scenarios and why the time interval is between 90 and 100 min? What can we conclude from the shown scenarios?
4 is not clear to me. Why there seem to always be a minimum generation of around 19 MW? Please also better describe what we see in the figure.
5 is unclear. Add y-axis labels. What are the different sub-plots?
Answer: Figure 3: x-axis labels added (time in hours). Scenarios are generated via CEEMD-based clustering.
Figure 4: Explained as source-load matching results, showing the selected load group with the highest degree (Section 4).
Figure 5: y-axis labels added (SOC [%]). Subplots show energy storage states post-optimization (Section 4).
Citation: https://doi.org/10.5194/wes-2025-11-AC1
-
RC2: 'Comment on wes-2025-11', Anonymous Referee #2, 21 Mar 2025
Overview:
This paper proposes a regional grid management by taking into account renewable energy production and energy storage management to reduce the fluctuations in the power load and dispatch pressure on the electrical grid. The paper utilizes a “one source multiple load” framework that matches up the loads with the low pass frequency of the renewable assets to utilize the renewable assets most efficiently. The paper is interesting and presents a valuable method for dispatching multiple energy storage systems to meet the load while also taking into account the health of the battery. I would recommend publishing the paper after the major comments below have been addressed to increase the clarity and impact of the paper.
Major Comments:
Line 98: What does “Repeat the above steps n times.” mean? Could the authors elaborate on how they choose n, and what it’s based on?
Line 153-155: There seems to be a sentence or two missing on what Equation 14 represents and the sentence afterwards is unclear.
Fig 3: Could the authors explain this figure more? What does the x-axis mean here? How did they obtain these figures?
Lines 262-268: This section is unclear. Are the authors referencing a previous paper (please cite this here), or are they talking about their own results? Figure 4 is never reference in the text.
Line 282: Here, the text references Fig. 5 and talks about the overall load power, however Fig. 5 shows the SOC of the energy storage systems. Were the authors referring to Fig. 6 here?
Fig. 7 and Lines 292-296: Could you give more statistics from this plot? Are we trying to reduce the peaks of this signal, or the average?
Fig. 8: Can you give the load fluctuation rate for Scenario 2 and Scenario 3?
Line 315-316: Is it obvious that the overall fluctuation from Scenario 3 is less than Scenario 2? How is this calculated?
Fig. 10 & 11: Can these figures be made larger? They are very hard to read.
Minor Comments:
- Figures 1 & 2 are small. Could the size of these be increased to aid readability?
- Line 67: “the scenario set scenario and the original output of the corresponding day” – Typo? Should scenario be in there twice? This is a little hard to understand.
- Line 90: Fix formatting of Gaussian white noise symbols in text
- Line 131: “is thematch between the source-side…” Typo
- Line 146: “with the other distributionsand the probability…” typo
- Line 179: Fix symbol formatting – this is recurring, check symbol formatting throughout the paper
- Line 229: “To better protect the energy storage and prolong the life of the storage.” Sentence fragment. What is this referring to?
Citation: https://doi.org/10.5194/wes-2025-11-RC2 -
AC2: 'Reply on RC2', Yongqing Zhu, 03 Apr 2025
1. Line 98: What does “Repeat the above steps n times.” mean? Could the authors elaborate on how they choose n, and what it’s based on?
Answer: n is chosen based on convergence criteria (e.g., 50 iterations to ensure noise cancellation). This is specified in Section 2.1.
2. Line 153-155: There seems to be a sentence or two missing on what Equation 14 represents and the sentence afterwards is unclear.
Answer: Equation 14 normalizes Spearman coefficients to [0,1], ensuring compatibility with Euclidean distance. Details are added in Section 2.3.
3. Fig 3: Could the authors explain this figure more? What does the x-axis mean here? How did they obtain these figures?
Answer: Figure 3 illustrates wind-solar scenario generation via CEEMD. The x-axis represents time (hours), and scenarios are derived from decomposed low-frequency components (Section 4).
4. Lines 262-268: This section is unclear. Are the authors referencing a previous paper (please cite this here), or are they talking about their own results? Figure 4 is never reference in the text.
Answer: There are no citations to the previous article, referring to the methods described earlier in this article.
Figure 4 is now referenced in Section 4, explaining source-load matching results.
5. Line 282: Here, the text references Fig. 5 and talks about the overall load power, however Fig. 5 shows the SOC of the energy storage systems. Were the authors referring to Fig. 6 here?
Answer: The text now correctly references Figure 6 for load power changes and Figure 5 for SOC states (Section 4).
6. Fig. 7 and Lines 292-296: Could you give more statistics from this plot? Are we trying to reduce the peaks of this signal, or the average?
Answer: Fluctuation rate metrics (e.g., standard deviation) are added in Section 4, showing a 50% reduction in volatility post-optimization.
7. Fig. 8: Can you give the load fluctuation rate for Scenario 2 and Scenario 3?
Answer: Fluctuation rates for Scenario 2 (0.35) and Scenario 3 (0.18) are explicitly stated in Section 4.
8. Fig. 10 & 11: Can these figures be made larger? They are very hard to read.
Answer: Figures 10 and 11 are enlarged, and labels are enhanced for clarity.
9. Line 315-316: Is it obvious that the overall fluctuation from Scenario 3 is less than Scenario 2? How is this calculated?
Answer: The results demonstrate that integrating source-load matching with optimized energy storage (Scenario 3) reduces load volatility more effectively than direct wind-solar utilization (Scenario 2). The calculation method (Equation 20) mathematically validates these improvements by quantifying deviations from the mean load profile.
Minor Comments:
- Figures 1 & 2 are small. Could the size of these be increased to aid readability?
- Line 67: “the scenario set scenario and the original output of the corresponding day” – Typo? Should scenario be in there twice? This is a little hard to understand.
- Line 90: Fix formatting of Gaussian white noise symbols in text
- Line 131: “is the match between the source-side…” Typo
- Line 146: “with the other distributions and the probability…” typo
- Line 179: Fix symbol formatting – this is recurring, check symbol formatting throughout the paper
- Line 229: “To better protect the energy storage and prolong the life of the storage.” Sentence fragment. What is this referring to?
Answer:
- Typos and formatting issues (e.g., Equations 1, 2, 15) are corrected.
- Symbol formatting (e.g., superscripts/subscripts) is standardized.
- Sentence fragments (e.g., Line 229) are revised for clarity.
Citation: https://doi.org/10.5194/wes-2025-11-AC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 96 | 34 | 7 | 137 | 10 | 6 |
- HTML: 96
- PDF: 34
- XML: 7
- Total: 137
- BibTeX: 10
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1