The Semi-Scheduling Mode of Multi-Energy System Considering Risk–Utility in Day-Ahead Market

Round 1

Reviewer 1 Report

This paper develops a semi-scheduling model for pumped-hydro-storage units in day-ahead markets. In my opinion, this paper needs major clarifications and modifications to be considered for publications. Please consider my comments below:

1 – In general, visibility of figures could be enhanced (e.g. Fig. 1).

2 – The whole text should be deeply revised to avoid grammar mistakes (e.g. line 111).

3 – As far as I understand, the proposed method is strongly founded on heuristic criteria. In this regard, I recommend comparing it with analytic techniques, which ensure the global optimum reachability.

4 – The literature about energy management in microgrids is rich, even particularizing to the case of uncertainties modelling in PHS-based systems. In this sense, the authors did not revise/compare some recent papers (e.g. 10.1002/er.8058).

5 – The novel semi-scheduling methodology is difficult to understand, especially its foundations. The authors should answer the question, why should the new proposal be considered instead of other existing approaches? Or, why is the new proposal necessary?

6 – One of the main advantages claimed for the semi-scheduling model is its computational cost, which is, a priori, competitive with full-scheduling methods. Despite this affirmation, I did not see any discussion/comment/result to support such idea.

Thanks to the authors for your effort and time.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper proposes a semi-scheduling method with heat-pumped storage considering risk utility. However, the procedures are very ambiguous, the applicability is limited, and the performance improvement compared to the other scheduling methods is not clear.

(1) The method is named a WSHTPC model by picking up a character from the names of power sources. This means the proposed method is limited to the very special condition where all the sources are included.

(2) The authors state that “we can not completely copy the foreign semi-scheduling mode.” This means that the semi-scheduling mode is not a new concept. Additionally, it is not clear why the foreign model cannot be applied to China and why the proposed model is suitable for China because the conditions in China are not described. Thus, the authors should verify the novelty and research contribution of the study by specifying the different points between the foreign and this semi-scheduling method and which conditions in China make this method useful.

(3) In Fig. 2, there is no branch condition between “Calculate generating power” and “Calculate pumping power”.

(4) It is not presented the reason why the functions in Eqs (6) and (7) are selected among many concave and convex functions.

(5) The risk utility function is described with respect to the increase or decrease in uncertainty of renewables. However, in Eqs. (6) and (7), the independent variable is the “output” of renewables, not the “risk (that is, uncertainty)” of renewables. Why are they risk functions?

(6) The performance is compared between the cases with and without pumped storage. However, it is obvious that the performance is better when storage devices are used. Thus, the authors should verify the performance advantage of the proposed method by comparing it with the other scheduling methods.

(7) There are a lot of previous studies on scheduling with energy storage systems(ESS), particularly with battery ESS. Except that the specific type of storage is the pumped storage in this study setup, the reviewer cannot find any difference from the other scheduling methods with storage systems. Thus, the authors should make it clear which conditions are different between the case with pumped storage and the case with other types of storage systems.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors,

the paper has been notably improved. However, the literature review is still too scarce (only 20 refs.!). As I commented in my previous revision, the literature about energy management in microgrids is very rich and the authors should provide a deeper review to better understand the contributions of this paper.

All the best.

Author Response

Point: The paper has been notably improved. However, the literature review is still too scarce (only 20 refs.!). As I commented in my previous revision, the literature about energy management in microgrids is very rich and the authors should provide a deeper review to better understand the contributions of this paper.

Response: Thank you very much for your good comment. I feel sorry for the inadequate literature review. I have deleted some irrelevant references, and updated some relevant papers in the revised paper. Finally, there are 31 references in the revised paper.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have appropriately addressed most of the comments of this reviewer. However, some points are still unclear as follows.

 

Point 4: It is not presented the reason why the functions in Eqs (6) and (7) are selected among many concave and convex functions.

==> The authors state that the functions in Eqs. (6) and (7) are typical concave and convex functions. However, they are not selected for case studies but fixed in the formulation. Thus, the authors should prove and/or underpin that any risk-utility characteristics can be curve-fit by the function form in Eqs. (6) and (7).

 

Point 7: There are a lot of previous studies on scheduling with energy storage systems(ESS), particularly with battery ESS. Except that the specific type of storage is the pumped storage in this study setup, the reviewer cannot find any difference from the other scheduling methods with storage systems. Thus, the authors should make it clear which conditions are different between the case with pumped storage and the case with other types of storage systems.

==> I understand that the pumped storage and other types of ESS are different. However, the differences are in the view of investment. Their energy buffering advantages are the same as each other, particularly for the battery ESS responding fast. Therefore, the differences are very few in the view of system operation including the scheduling of sources, and the proposed method is for scheduling. Consequently, the authors need to clarify that the proposed method has novelty and contributions compared to other methods with other types of ESS (especially battery ESS) in terms of the operational strategy that the ESS fills the buffer when the price is low or the supply is enough, and the ESS releases the stored energy when the price is high or the demand is high.

 

Author Response

The authors have appropriately addressed most of the comments of this reviewer. However, some points are still unclear as follows.

Point 4: It is not presented the reason why the functions in Eqs (6) and (7) are selected among many concave and convex functions.

==> The authors state that the functions in Eqs. (6) and (7) are typical concave and convex functions. However, they are not selected for case studies but fixed in the formulation. Thus, the authors should prove and/or underpin that any risk-utility characteristics can be curve-fit by the function form in Eqs. (6) and (7).

Response 4: Thank you very much for your good comment. I feel sorry that my inappropriate expression to make you misunderstand. To simplify the model, the risk attitudes of pumped storage units are only divided into risk-aversion and risk-preference in this paper. There are two reasons that the functions in Eqs (6) and (7) are selected as risk utility functions. One reason is that the logarithmic function and exponential function are typical concave and convex functions, which could indicate the sensitivity of y to the increase of x. And the other more important reason is that it is necessary to construct the common analytic expression, such as linear function, logarithmic function, exponential function and so on [31], to call curve_fit function on Python platform to fit the utility function. Furthermore, it should be noted that the fitting maybe more difficult for some other forms of concave and convex functions, even no fitted results when calling curve_fit function on Python platform. Thus, the general forms of logarithmic function and exponential function, being introduced from the reference [31], are selected as the risk utility functions, as shown in Eqs (6) and (7) in this paper.

[31] S Qian, Y Gan, F Tian, et al. Operational Research[M]. Beijing: Tsinghua University Press, 2014: 494.

Point 7: There are a lot of previous studies on scheduling with energy storage systems(ESS), particularly with battery ESS. Except that the specific type of storage is the pumped storage in this study setup, the reviewer cannot find any difference from the other scheduling methods with storage systems. Thus, the authors should make it clear which conditions are different between the case with pumped storage and the case with other types of storage systems.

==> I understand that the pumped storage and other types of ESS are different. However, the differences are in the view of investment. Their energy buffering advantages are the same as each other, particularly for the battery ESS responding fast. Therefore, the differences are very few in the view of system operation including the scheduling of sources, and the proposed method is for scheduling. Consequently, the authors need to clarify that the proposed method has novelty and contributions compared to other methods with other types of ESS (especially battery ESS) in terms of the operational strategy that the ESS fills the buffer when the price is low or the supply is enough, and the ESS releases the stored energy when the price is high or the demand is high.

Response 7: Thank you very much for your good comment. I feel sorry that my inappropriate expression to make you misunderstand. Indeed, in the application scenario of peaking shaving, valley filling and frequency regulation, due to the fast responding ability, the battery ESS has more advantages than pumped storage. However, with larger storage capacity and more low-carbon recycling mode, pumped storage realizes the storage and release of electrical energy through the mutual conversion between the electrical energy and the water potential energy, which is more suitable to promote the consumption of the large-scaled renewable energy. In this paper, based on the renewable energy output, the risk utility is introduced to provide reserve capacity of pumped storage to promote the consumption of renewable energy. And through the case study in Section 6.3, we can see from the Fig.8 (c) that the wind power can be consumed more in the semi-scheduling mode. Generally, similar with other ESS, the pumped storage pumps when the supply is enough, and the pumped storage generates when the demand is high.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

The authors have properly answered to the comments of this reviewer.