Demand Response Coupled with Dynamic Thermal Rating for Increased Transformer Reserve and Lifetime

Round 1

Reviewer 1 Report

This paper presents a least-cost-technology to assess the available reserve for an oil-immersed distribution transformer using Demand Response and Dynamic Thermal Rating. The result demonstrates significantly increased reserve margins as compared to traditional approaches and verifies the validity and flexibility of the proposed methods. A few comments for the authors are listed below:

1. It is said that the problem is written using the YALMIP library in MATLAB and solved with CPLEX 12.0. Any reason that CPLEX 12.0 is selected to solve the problem? Is it fast enough to converge?
2. The validation runs for DR integrated management and design are provided in the results section. It would be better to include a table to compare the computational times before and after the conventional piecewise linearization is implemented since in section 3.3 it is claimed that using heuristic and meta-heuristics methods in MATLAB Optimization Toolbox was suffered from expensive computational times and non-systematic convergence due to the problem size, especially with 1-min resolution for the reference temporal set, while linearization solves the issue.
3. In Figure 14, why there’re flat DR power and power shedding share between Reserve 30%-45%? Can it be addressed?
4. There are a few grammar mistakes/typos need to be corrected, e.g., (Ln 109): “A specific attention is attached the problem formulation for integrated DR design and management.”, (Ln 162): “the Table 2”, (Ln 210): “allows”, etc.

Author Response

This paper presents a least-cost-technology to assess the available reserve for an oil-immersed distribution transformer using Demand Response and Dynamic Thermal Rating. The result demonstrates significantly increased reserve margins as compared to traditional approaches and verifies the validity and flexibility of the proposed methods. A few comments for the authors are listed below:

Dear Reviewer, we want to express all our gratitude for all your comments and advices that help to improve the quality and clarity of the paper.

2. It is said that the problem is written using the YALMIP library in MATLAB and solved with CPLEX 12.0. Any reason that CPLEX 12.0 is selected to solve the problem? Is it fast enough to converge?

Dear Reviewer, indeed CPLEX is selected in our manuscript as the YALMIP Toolbox provides a seamless integration of the solver in the MATLAB environment. In addition, CPLEX is deemed more scalable than the linear programming solver embedded in MATLAB,  Due to 1-min time resolution, required by IEC 60076-7 standard, the number of variables in the optimization problem becomes an issue for long horizons/reserve margins (up to 3 million variables for the longest horizons).

The corresponding modifications are presented on the page 11 and highlighted by yellow color.

2. The validation runs for DR integrated management and design are provided in the results section. It would be better to include a table to compare the computational times before and after the conventional piecewise linearization is implemented since in section 3.3 it is claimed that using heuristic and meta-heuristics methods in MATLAB Optimization Toolbox was suffered from expensive computational times and non-systematic convergence due to the problem size, especially with 1-min resolution for the reference temporal set, while linearization solves the issue.

Dear Reviewer, following your recommendation we added the figure 18 on the page 18 and its interpretation on the pages 18- 20 to show and explain the difference in computation times. Indeed, preliminary tests shows much slower convergence with the heuristics embedded in MATLAB. The proposed formulation and faster simulations and systematic convergence allows to test a wide range of scenarios in reasonable computational times.

The corresponding modification are highlighted by green color.

3. In Figure 14, why there’re flat DR power and power shedding share between Reserve 30%-45%? Can it be addressed?

Dear Reviewer, we considered your suggestion and added the explanation for the flat DR power and power shedding in the Figure 14. This flat curves can be explained by the thermal capability of transformer to withstand the connected load alone therefore no DR is needed. In other words, if any load, corresponding to reserve margins below 45%, would be connected to transformers, the transformer total load will not violate any temperature or ageing limits.

The corresponding modifications are presented on the page 16 and highlighted by cyan color.

4. There are a few grammar mistakes/typos need to be corrected, e.g., (Ln 109): “A specific attention is attached the problem formulation for integrated DR design and management.”, (Ln 162): “the Table 2”, (Ln 210): “allows”, etc.

Dear Reviewer, we considered your remark in revised paper. Thus, we corrected all abovementioned grammar mistakes/typos. We admit that not being native speakers we could miss errors. Thus, we asked a professional translator, whose working language is English, to read and correct the revised paper. Therefore, some improvements are made in different parts of the paper. However, if there are still errors especially those you found in first review, we would appreciate if you could specify section in revised paper to let us improve it

The corresponding modifications are presented on the pages 3, 5, 6 and highlighted by grey color.

Author Response File: Author Response.pdf

Reviewer 2 Report

The reviewed paper is interesting and it is devoted to important scientific and technical problem. This paper is well organized and clearly written.

The Authors proposed the methodology to assess the available reserve for transformer using demand response and dynamic thermal rating. Results of investigations presented in this papers show that this methodology makes it possible to reduce maximum value transformer temperature and to ensure high reserve margins of the transformers.

In my opinion it can be published after minor revision. In the revised version of this paper the Authors should take into account following remarks:

1. In Section 3.1 explain the meaning of parameters presented in Table 3 and in Equations (1) - (3).
2. In Section 3.4 the ageing model should be described in more detail.
3. The wrong symbol in Equation (13) should be corrected.

Author Response

The Reviewer 2

The reviewed paper is interesting and it is devoted to important scientific and technical problem. This paper is well organized and clearly written.

Dear Reviewer, we want to express all our gratitude for all your comments and advices that help us improve the proposed paper.

The Authors proposed the methodology to assess the available reserve for transformer using demand response and dynamic thermal rating. Results of investigations presented in this papers show that this methodology makes it possible to reduce maximum value transformer temperature and to ensure high reserve margins of the transformers.

In my opinion it can be published after minor revision. In the revised version of this paper the Authors should take into account following remarks:

1. In Section 3.1 explain the meaning of parameters presented in Table 3 and in Equations (1) - (3).

Dear Reviewer, we summarized all symbols as well as their meanings in Table 1 (on the page 1). After rereading the initial version we admit that it was not probably clearly explained that the reader can see the meanings of each symbol, including those given in the equations, in the Table 1. Thus, we improved the text by properly referring the Table 1.

The corresponding modifications are presented on the page 8 and the modified text is highlighted by red color.

2. In Section 3.4 the ageing model should be described in more detail.

Dear Reviewer, to provide the additional details, we added a new figure 8a on the page 12 and provided its explanation. The corresponding text is highlighted by violate color on the same page.  

3. The wrong symbol in Equation (13) should be corrected.

Dear Reviewer, we considered your remark by using the correct brackets. The modified equation (13) is presented on the page 13.

Author Response File: Author Response.pdf