A Space Reduction Heuristic for Thermal Unit Commitment Considering Ramp Constraints and Large-Scale Generation Systems

Round 1

Reviewer 1 Report

 

Unfortunately, I can't identify which SI Energies the article is classified under. According to this, it would be possibly better to judge the suitability of the article for Energies journal.

The manuscript is at a very good professional level.

I have only a few comments:

-Line 135: what does "[? ]" mean?

-I believe that the hardware application mentioned in Chapter 4.3 is currently outdated and the above calculations could be done faster. For example, a more powerful Workstation of a newer generation could be used instead of conventional hardware.

-It is unclear what TUC applications are involved and the general reader will not get a complete picture of the problem.

- According to the above methodology, the solution appears to be a cost path, but it is unclear whether these are the same TUC or similar systems.

Please fill in the missing information on TUCs to make it clear where such units occur.

- Have the results been verified on specific systems or is this just theoretical?

 

Author Response

Dear reviewer, we thank you for your suggestions and for contributing to the improvement of our paper.  We also thank you for complimenting our work. Our comments are found below. Regards, the authors.

 

-Line 135: what does "[? ]" mean?

This was related to a missing reference. The corrected citation is in line 146 of the revised paper.

-I believe that the hardware application mentioned in Chapter 4.3 is currently outdated and the above calculations could be done faster. For example, a more powerful Workstation of a newer generation could be used instead of conventional hardware.

You are correct. However, the contribution of the work is in the reduction of the search space while keeping good quality solutions, regardless of the hardware. We added a short text (line 184) mentioning that all simulations utilized the same hardware, which makes all comparisons fair and pertinent.

-It is unclear what TUC applications are involved and the general reader will not get a complete picture of the problem.

The application regards the short-term operation planning of the generating units to attend hourly power and reserve demands. We have complemented the first paragraph of the introduction to help make it more clear.

- According to the above methodology, the solution appears to be a cost path, but it is unclear whether these are the same TUC or similar systems.

We included a citation in line 235 of the revised paper as an attempt to clarify this matter. In summary, the systems are the same from Kazarlis. The ramp rates came from the work of Simopoulos, as already mentioned in the same paragraph of the new citation.

-Please fill in the missing information on TUCs to make it clear where such units occur.

We believe the clarifications added to the introduction clarify this matter.

- Have the results been verified on specific systems or is this just theoretical?

Results have not been verified on specific system. However, we emphasize that the system utilized (from Kazarlis) is vastly used in the literature as a validation system for new methods for the TUC problem.

Reviewer 2 Report

Summary:

The authors proposed a heuristic to narrow the search space region for thermal unit
commitment problems. The significance of such techniques stems from the combinatorial
explosion of problem variables when a large number of generating units are present in the system.
 The method employs sensitivity indices to gather information about the system and
fix many of the binary decision variables over the planning horizon.
The authors investigate the method by demonstrating its effectiveness
in large-scale systems subject to ramp constraints. Despite the significantly increased complexity,
 the results of the paper show that the method can achieve high-quality solutions
significantly faster than other approaches in the literature.

Comments:

** The paper deals with a classical optimization problem in the field of energy. Several versions of the problem exist in the literature and several resolution approaches have been proposed. The organization of this paper does not simplify the understanding and enhancement of the contribution. To solve this problem, I propose to the authors the following organization: after the introduction section, the authors should reserve the second section for the literature review (instead of the current section 3) in which two summary tables must appear: the first is a summary organization of the main works of literature on the different versions of TUC (different constraints, objectives, type of modeling, different resolution methods, etc.). The second table should summarize the work specific to the thermal unit commitment with ramp constraints (used methods, results, size of instances solved, etc). Afterward, a paragraph should contain the gap filled by the paper. And finally, enumerate as bullet points the main contributions of the paper.

** In this sentence "In this work, the problem is given by a mixed integer quadratic programming formulation" it's not clear whether the quadratic formulation is an original contribution of the authors or a formulation from literature. The authors should clarify their contributions by using "we propose". Or clearly cite the reference.

** In tables providing results, the authors should give the formula used for the "optimality gap"


** A question for authors about table 7: did you solve the problem for more than 100 units (by augmenting the gap) to study the "asymptotic" or the limit of the solving approach?
 


** some missing reference as in line 135 "The spinning reserve is
taken as 10% of the load, as in [? ]."

Author Response

Dear reviewer, we thank you for your suggestions and for contributing to the improvement of our paper. Our comments are found below. Regards, the authors.

 

- The paper deals with a classical optimization problem in the field of energy. Several versions of the problem exist in the literature and several resolution approaches have been proposed. The organization of this paper does not simplify the understanding and enhancement of the contribution. To solve this problem, I propose to the authors the following organization: after the introduction section, the authors should reserve the second section for the literature review (instead of the current section 3) in which two summary tables must appear: the first is a summary organization of the main works of literature on the different versions of TUC (different constraints, objectives, type of modeling, different resolution methods, etc.). The second table should summarize the work specific to the thermal unit commitment with ramp constraints (used methods, results, size of instances solved, etc). Afterward, a paragraph should contain the gap filled by the paper. And finally, enumerate as bullet points the main contributions of the paper.

We have written the contributions as bullet points so that readers can easily identify them (lines 52 to 54). Regarding the suggestion for the review section, we fear that such an addition would require much more time on our side to work on revising the paper. Furthermore, providing a deep review is not within our scope. To help address the issue risen by the reviewer, we have enriched the Introduction. In addition, we included a citation to a recent review (line 24), which we believe does a great job in describing the UC state of the art.

- In this sentence "In this work, the problem is given by a mixed integer quadratic programming formulation" it's not clear whether the quadratic formulation is an original contribution of the authors or a formulation from literature. The authors should clarify their contributions by using "we propose". Or clearly cite the reference.

We included a citation (line 64) to make it clear that the formulation came from the literature.

- In tables providing results, the authors should give the formula used for the "optimality gap".

Although not presented as a formula, we have included a description (line 176) of the optimality gap according to what Mosek applies during the optimization. We hope such inclusion addresses the suggestion.

- A question for authors about table 7: did you solve the problem for more than 100 units (by augmenting the gap) to study the "asymptotic" or the limit of the solving approach?

We have limited our simulations to the large systems usually benchmarked in the literature. Thus, we have not studied systems with more than 100 units.

- Some missing reference as in line 135 "The spinning reserve is taken as 10% of the load, as in [? ]."

This issue was corrected (line 150).

Round 2

Reviewer 2 Report

First, from the revised version, one can read the main contributions of this paper:

1. An adaptation of the method from [19] to significantly reduce the search space for the TUC problem considering ramp constraints.
2. An evaluation of the method feasibility regarding large-scale systems not considered in [19].

For me, this two contibutions are not sufficient to publish this paper. The orginality content (structural contribution and elaborate results) of the paper with respect to reference [19] is poor and not sufficient.

Second, in response to my previous report, the authors stated that "Regarding the suggestion for the review section, we fear that such an addition would require much more time on our side to work on revising the paper."

This answer is not convincing. The authors should revise the paper considering "relevant" comments from reviewers. They don't have to be concerned about how long it will take.

Author Response

We thank the reviewer for the comments and constructive criticism. Our replies are below.

Comment 1:

For me, these two contributions are not sufficient to publish this paper. The originality content (structural contribution and elaborate results) of the paper with respect to reference [19] is poor and not sufficient.

Authors: We consider the contributions to be relevant since they build upon the recent space reduction method from [19] by addressing large systems and ramp constraints. As stated in Montero (2022), a few publications address these characteristics simultaneously. We have expanded the description of the contributions (lines 47 to 56) to better emphasize the importance of the contributions.

Comment 2:

Second, in response to my previous report, the authors stated that "Regarding the suggestion for the review section, we fear that such an addition would require much more time on our side to work on revising the paper."

This answer is not convincing. The authors should review the paper considering "relevant" comments from reviewers. They don't have to be concerned about how long it will take.

Authors: the barrier to address this comment comes from the fact that Energies/MDPI itself asks authors to submit the revised version in 5 days. In other words, the concern about how long it would take comes from the journal's instructions. We would not be able to address the comment in a good quality manner in the period of time given by the journal.

Regards,

The authors.

Round 3

Reviewer 2 Report

Dear Authors,

I understand that you need more time to correct the proposed major revisions. Actually, 5 days are insufficient. It's not your fault!

While I am not completely convinced by your responses (which I will keep for myself), I should acknowledge that your paper is now better than several previously published papers.
Good luck!

Sincerely,