Genetic Algorithm for Energy Commitment in a Power System Supplied by Multiple Energy Carriers

Round 1

Reviewer 1 Report

The proposed manuscript proposed a genetic algorithm for energy commitment in a power system supplied by multiple energy carriers.

I suggest the following modification to improve the quality of the manuscript.

• I suggest reducing the introduction length, by focusing more on the novelties of the proposed paper with respect to literature findings that should be developed more.
• Line 97: “ This model is a mixed-integer linear program (MILP) based on Linear Program (LP) models” check this sentence.
• Is it necessary to use both Figure 3 and Table 4?
• I suggest putting Section 3.5 in a dedicated Appendix.
• I suggest putting some tables such as Table 5 and Table 6 in a dedicated Appendix (if possible).
• A dedicated section on results is better than including it in the section of the same section where the description of the case study and/simulation is provided.
• Please, put some quantitative findings in both abstract/conclusions.
• In my opinion, (the weak point of the manuscript) simulation results should be more critically described.

Author Response

Reviewer #1: The proposed manuscript proposed a genetic algorithm for energy commitment in a power system supplied by multiple energy carriers.

Authors do appreciate your consideration and useful comments on the manuscript. It surely improves the quality of the paper. Based on these valuable comments, the article has been revised. The authors hope that the revised manuscript will be accepted.

I suggest the following modification to improve the quality of the manuscript.

1. I suggest reducing the introduction length, by focusing more on the novelties of the proposed paper with respect to literature findings that should be developed more.

Response: Thank you for your observation. Based on this, authors have made some changes. Please see lines 132 to 149:

Lines 132 to Lines 149:

In this research work, based on dynamic programming (DP) and genetic algorithm (GA) concepts, the solution of energy commitment (EC) in multi-carrier energy systems is proposed. Moreover, considering the required information from the final energy consumption, in the present study a mathematical model is developed to estimate: the amount of electrical energy demands in the country, the combination of input fuel to power plants and the optimum combination of production with a view to satisfying the demand. Thus, it offers the best integration of primary energy carriers to supply energy consumption.

This work enables a better understanding for the following benefits:

• Determination of the most appropriate pattern of using energy carriers to satisfy energy consumption.
• Study of integrated energy network.
• Integrated optimization of energy carriers instead of independent optimization of each carrier separately.
• Mathematical modeling of energy network from bottom to up (from the lowest energy level to the highest energy level).
• Application of DP and GA in EC.
• Impact of crude oil refining and its products on EC.
• Distribution of electrical energy as a subset of EC study.

1. Line 97: “ This model is a mixed-integer linear program (MILP) based on Linear Program (LP) models” check this sentence.

Response: Thank you for your remark. To address this valuable comment, authors checked this sentence. This sentence has been modified to: “ This model is based on mixed-integer linear program (MILP) and Linear Program (LP) [29] that includes the features of the demand side management (DSM) [30, 31] ”.

Line 97:

“ This model is based on mixed-integer linear program (MILP) and Linear Program (LP) [29] that includes the features of the demand side management (DSM) [30, 31] ”.

3. Is it necessary to use both Figure 3 and Table 4?

Response: Thank you for your comment. Figure 3 and Table 4 show the energy consumption information in the energy grid.

Table 4 contains detailed information and data on energy consumption in different sectors of the energy network. Figure 3 is also used for visual representation and intuitive perception. Readers can see the peak time of energy demand as well as changes in energy consumption at different hours by referring to Figure 3.

The authors thank the dear reviewer for this in-depth attention. The authors suggest that Figure 3 and Table 4 be included in the article, however, if the dear Reviewer does not consider it appropriate, the authors will remove these items.

4. I suggest putting Section 3.5 in a dedicated Appendix.

Response: Thank you for your suggestion. In this section, the process and steps of implementing the proposed study are described. This section provides important information. A flowchart has also been added for more explanation. The authors ask the dear Reviewer for permission to have this section in the same way in the article.

5. I suggest putting some tables such as Table 5 and Table 6 in a dedicated Appendix (if possible).

Response: Authors do appreciate your suggestion. To address it, the authors move this tables to Appendix A. Thank you.

Lines 427 to 435

6. A dedicated section on results is better than including it in the section of the same section where the description of the case study and/simulation is provided.

Response: Authors do appreciate the Reviewer’s remark. This comment is very valuable and accurate. This comment greatly contributes to improving the quality of the article. Based on this valuable comment, a new section named '' 4. Simulation Results ‘’ has been added.

Line 336

4. Simulation Results
5. Please, put some quantitative findings in both abstract/conclusions.

Response: Thank you. The authors have modified both abstract and conclusions based on this valuable comment.

Abstract: Lines 34 to 37

Conclusions: Lines 391 to 394

8. In my opinion, (the weak point of the manuscript) simulation results should be more critically described.

Response: Thank you for your observation. This comment greatly contributes to improve the quality of the article especially for the Simulation Results Section. Based on this valuable comment this section has been revised and detailed description has been added.

Lines 343 to 369

UC is one of the important outputs of EC study which determines the on / off status and output of units for each hour of the study period. Table 5 presents the results of the UC. In this table, a power plant with zero production is off. Also, the power plant that its production rate has been determined is on. In the last column of this table, the value of the objective function for each hour of the study is presented. Figure 5 shows the production profiles of the different units for the entire 24-hour period. According to this figure, in the seventh hour of the study period, which is related to the highest peak of demand, the fourth unit turns on and then, with the decrease of demand, shut down in the 12^th hour of the study period. The processes of achieving an optimal distribution of electrical energy between the units are shown in Figures 6 and 7. Figure 6 is for the first to twelfth hours of the study period and Figure 7 is for the thirteenth to twenty-fourth hours of the study period. In these figures, the convergence curves of the performance of the GA in the optimization of the objective function are shown as the best solution in terms of the iteration of the algorithm. Also, the “state” in these figures shows the number of power plants on.

Another important outcome of the energy commitment study is to determine the appropriate pattern of use of energy carriers, which is presented in Table 6. In this table, the amount of need for nine energy carriers for each hour of the study period is specified separately. In this table, negative numbers indicate the excess of energy carrier production and positive numbers indicate the remaining need for energy carrier, which is supplied based on its domestic production.

Based on the amount of need for energy carriers per hour of the study period, the total amount of need for energy carriers for the entire study period can be calculated, which is presented in Table 7. In fact, Table 7 identifies the need for each energy carrier for the entire 24-hour study period. The meaning of negative and positive numbers in this table is similar to Table 6.

The need for energy carriers is identified in Tables 6 and 7, which must be supply using domestic products. But if domestic production is not enough to supply each of the energy carriers, that energy carrier must be supplied in the form of imports. Also, energy carriers that have surplus production enter the export sector. Accordingly, the export and import volumes of different energy carriers as another output of the EC study are presented in Table 8.

Author Response File: Author Response.pdf

Reviewer 2 Report

Paper: Genetic Algorithm for Energy Commitment in a Power System Supplied by Multiple Energy Carriers

Authors: M. Dehghani, M. Mardaneh et al.

After studying the content of the paper, the reviewer found the following:

Strong points:

The subject is of interest, in the global effort of energy efficiency and optimal use of primary resources.

The methodology used by the authors int heir research is described extensively and backed up by a well-covered state of the art presentation.

A case study is presented, with relevant results.

Concern points and questions for the authors:

In section 2, EC Problem Formulation, and in Algorithm 1, the steps used for solving the problem are described. However, a supplementary graphical representation would help to follow more easily the method.

In the case study, it can be seen that the test system uses only 4 power plants, of classical types.Their aggregate maximum generation capacity does not seem to be sufficient for covering the peak load from Figure 3. How is the difference accounted for? No clear distinction is made between electricity and other energy carriers availability and this should be better explained.

On the other hand, newer generation sources such as PV and wind power plants were not included in the electricity generation mix. How would their presence affect the results of the  study?

In Table 6, the T23 parameter (transmission, distribution and consumption efficiency) has values of 1 or above. What is the significance of these values? Efficiency is usually less than 1, to account for energy conversion and transmission losses.

In Table 10 (Need for energy carriers), what is the significance of the negative values? They could translate into surplus, but the values for the coke gas export from Table 11 do not match. More explanation should be provided.

Results are provided in Figure 4 only for electricity. For a complete picture, the other branch from Figure 2 should also be presented in the same manner.

The authors are kindly asked to comment on the relevance of the information from Figures 5 and 6. For a meaningful evaluation, other details should be provided, such as execution time and GA setup parameters such as population size, chromosome length, number of iterations.

Author Response

Herein in attached file the response to reviewer No. 2 comments

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors kindly replied to all my requests.

Author Response

The authors appreciate dear reviewer for the carefully consideration and useful comments on the paper. It surely improves the quality of the paper.

Author Response File: Author Response.pdf

Reviewer 2 Report

The reviewer thanks the authors for their extensive comments regarding the issues raised for the initial version of the paper. It mostly agrees with the clarifications and modifications performed int he paper. In this form, in its opinions, some supplemental clarifications should be made:

In modeling the problem of energy commitment, because mathematical modeling starts from consumption and continues to production, the T23 parameter (transmission, distribution and consumption efficiency) has values of 1 or above.

If this is the case, since at line 171, the parameter T23 is labeled as efficiency, the authors are kindly asked to add in the paper, possibly in fewer words, the explanation provided in the  reviewer comments. This would help readers to better understand the intent of the authors.

The need for energy carriers to provide final energy consumption over 24 hours is given in Table 6. Similar figures can be drawn for all energy carriers identified for the entire study period in Table 6.
Therefore, 9 new figures must be drawn.

A simpler approach would be (in conjunction with the discussion about the electricity peak and total energy demand peak) to show the electricity and non-electricity curves. Also, Figure 2 seems to suggest this separation. If this is not true, the meaning of the two main arrows should be better explained.

3.6. Experimental Setup
The proposed study is simulated on the mentioned energy network for a 24-hour study period. The proposed EC is implemented in Matlab R2017b version using a 64-bit Core i7 processor with 3.20 GHz and 16 GB main memory.

This information is relevant only if the computation time is a factor in the optimization. Thus, could the authors provide this information?

Genetic algorithm has been used as the optimization technique to solve the objective function. The number of chromosomes (population size) of the algorithm is equal to 50, the length of each chromosome is equal to 8 and the number of iterations is 50.

The figures suggest that the number of iterations is sufficient, since the optimization stops after 10-20 iterations. But, usually, the standard genetic algorithms uses larger populations. Why was 50 used? Also, again for clarity reasons, could the authors explain what are the 8 variables used in the GA chromosome? There are 9 energy carriers used in the paper, as Table 6 shows. This seems to be a little unclear.

Thank you!

Author Response

Comments from reviewer 2:

The reviewer thanks the authors for their extensive comments regarding the issues raised for the initial version of the paper. It mostly agrees with the clarifications and modifications performed in the paper. In this form, in its opinions, some supplemental clarifications should be made:

 The authors appreciate dear reviewer for the carefully consideration and useful comments on the paper. It surely improves the quality of the paper. Based on these valuable comments, the article has been revised. The authors hope that the revised paper will be accepted by dear reviewer.

Concern points and questions for the authors:

1. “In modeling the problem of energy commitment, because mathematical modeling starts from consumption and continues to production, the T23 parameter (transmission, distribution and consumption efficiency) has values of 1 or above.”

If this is the case, since at line 171, the parameter T23 is labeled as efficiency, the authors are kindly asked to add in the paper, possibly in fewer words, the explanation provided in the  reviewer comments. This would help readers to better understand the intent of the authors.

Response: Thank you so much to the dear reviewer for his valuable and accurate comment. To address this valuable comment, a brief explanation has been added.

Lines 171 to 173

It should be noted that due to the fact that mathematical modeling is considered from demand to production (down to up), the  parameter (transmission, distribution and consumption efficiency) has values of 1 or above.

2. “The need for energy carriers to provide final energy consumption over 24 hours is given in Table 6. Similar figures can be drawn for all energy carriers identified for the entire study period in Table 6.

Therefore, 9 new figures must be drawn.”

A simpler approach would be (in conjunction with the discussion about the electricity peak and total energy demand peak) to show the electricity and non-electricity curves. Also, Figure 2 seems to suggest this separation. If this is not true, the meaning of the two main arrows should be better explained.

Response: Thank you so much to the dear reviewer for his valuable and accurate comment. The authors thank the dear reviewer for this careful attention. Based on this valuable comment, the authors have added a description of Figure 2.

Lines 260 to 265

“In other words, Figure 2 shows the flow of energy from primary energy carriers to energy demand. Part of the energy demand is met directly from the primary energy carriers, which is indicated by arrow 2. Part of the energy demand is related to secondary energy carriers that become available after the energy conversion process in refineries or power plants which this concept is indicated by arrow 1. For example, electrical energy is a secondary energy carrier that is generated based on the conversion process of energy carriers in power plants.”

3. 6. Experimental Setup

The proposed study is simulated on the mentioned energy network for a 24-hour study period. The proposed EC is implemented in Matlab R2017b version using a 64-bit Core i7 processor with 3.20 GHz and 16 GB main memory.

This information is relevant only if the computation time is a factor in the optimization. Thus, could the authors provide this information?

Response: Thank you so much to the dear reviewer for his valuable and accurate comment. Based on this valuable comment, the authors deleted this sentence.

Lines 337 to 344

“The proposed study is simulated on the mentioned energy network for a 24-hour study period.

Genetic algorithm has been used as the optimization technique to solve the objective function. The number of chromosomes (population size) of the algorithm is equal to 50, the length of each chromosome is equal to 8 and the number of iterations is 50. The objective function has 8 variables which the 4 variables are related to the on or off status of power plant units and the other 4 variables are related to the amount of production of each of these units. In GA, the length of each chromosome is selected based on the number of problem variables. Therefore, the length of each chromosome is considered to be equal to 8.”

4. “Genetic algorithm has been used as the optimization technique to solve the objective function. The number of chromosomes (population size) of the algorithm is equal to 50, the length of each chromosome is equal to 8 and the number of iterations is 50.”

The figures suggest that the number of iterations is sufficient, since the optimization stops after 10-20 iterations. But, usually, the standard genetic algorithms uses larger populations. Why was 50 used? Also, again for clarity reasons, could the authors explain what are the 8 variables used in the GA chromosome? There are 9 energy carriers used in the paper, as Table 6 shows. This seems to be a little unclear.

Response: Thank you so much to the dear reviewer for his valuable and accurate comment. Based on this valuable comment, more detailed explanations have been provided.

As the dear reviewer has stated, the algorithm has reached convergence and a quasi-optimal solution in sufficient repetition.

An important parameter in optimization algorithms is the number of population members. In various articles, the number of population members for the genetic algorithm has been observed from about 20 chromosomes to 200 chromosomes.

In many of these cases, the number of chromosomes is experimentally suggested by the paper's supervisor.

A general rule for selecting the number of population members based on the number of problem variables is also suggested by researchers:

N ≥ 5 * (problem variables)

So, in this paper:

N ≥ 5* 8    N ≥ 40

Set: N=50

Lines 338 to 344

Genetic algorithm has been used as the optimization technique to solve the objective function. The number of chromosomes (population size) of the algorithm is equal to 50, the length of each chromosome is equal to 8 and the number of iterations is 50. The objective function has 8 variables which the 4 variables are related to the on or off status of power plant units and the other 4 variables are related to the amount of production of each of these units. In GA, the length of each chromosome is selected based on the number of problem variables. Therefore, the length of each chromosome is considered to be equal to 8.

All these modifications are highlighted in the paper

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

The reviewer thanks the authors for their supplementary comments.

After language proofreading, the paper can be considered now for publishing.

Good luck!