Comparison of Optimal Hedging Policies for Hydropower Reservoir System Operation
Round 1
Reviewer 1 Report
The authors present the results from the application of 7 hedging policies rules linked with the hybrid PSO-GA (4 customary forms - 1PHP, 2PHP, 3PHP, DHP - and 3 new forms - SOPHP, BSOPHP, SHPHP) to the optimization of power generation in Batang Padang's hydropower reservoir system, Malaysia. The aim of the paper is newsworthy in relation to the perspective of optimizing the water available use efficiency in order to generate energy. This is a very important goal in the context of global warming mitigation, making this item a source of huge interest to the readers of this journal.
In a global appreciation, some points are listed:
- The stability of power supply along the year shouldn't be the most important objective, because it is essential to look for energy demand. What about considering the objective function as the difference between energy demand and energy generation?
- It would be necessary to discuss the differences between the 7 forms of hedging rules. Once the results are very similar, we may question the importance of the 7 different approaches.
- It would increase the interest of the paper if the reader could have a better understanding of the study area and the standard of the demand, evaporation and water level data.
- The reservoir storage volumes constrain (2.3.3) is a consequence of the main and first constrain: the reservoir head.
- In Figure 9 it would be helpful to represent the TNB, too. It is not possible for me to understand the SHPHP result. It would be an added value if the authors could discuss these results more deeply.
- The presentation of the different plots of water balance equation along the study period would be very interesting, even in Supplementary File.
- There are some duplications in the text (e.g. several lines in the beginning of results' chapter, in which the objectives of the paper are presented unnecessarily for the second time).
Additional minor concerns:
e.g. Line 111 – replace “KW” for “kW”. There are other cases.
Line 304 – replace “Figure.S2” for “Figure 1”
Line 358 – TNB never appear before
Author Response
Reviewer 1
Thank you for your comments.
Please note:
Comments and Suggestions for Authors
The authors present the results from the application of 7 hedging policies rules
linked with the hybrid PSO-GA (4 customary forms - 1PHP, 2PHP, 3PHP, DHP
- and 3 new forms - SOPHP, BSOPHP, SHPHP) to the optimization of power
generation in Batang Padang's hydropower reservoir system, Malaysia. The
aim of the paper is newsworthy in relation to the perspective of optimizing the
water available use efficiency in order to generate energy. This is a very
important goal in the context of global warming mitigation, making this item a
source of huge interest to the readers of this journal.
In a global appreciation, some points are listed:
- The stability of power supply along the year shouldn't be the most important
objective, because it is essential to look for energy demand. What about
considering the objective function as?
1.The stability of the system is not an objective function. Producing the maximum power generation in operational time (2003-2012) was taken as an objective function.
In an overall, the main reason of using hedging policies is to uniformly distribute the water deficiency to decrease the future severe shortage. In terms of hydropower reservoir systems, the concept of hedging and rationing factors are used to maintain the water at present to increase the water storage and water head in the future, which are the key factors for power generation.
So based on the above explanation, hedging policy could be an effective policy in hydropower generation by keeping the water in higher level, which lead to produce more power generation with less release (because power generation is a function of head and release together). So, it logically can produce the target power most of the time as well, while the system is also stable.
2.With respect to your opinion, the difference between energy demand
and energy generation can be a good objective function. We can consider this in
the future work.
- It would be necessary to discuss the differences between the 7 forms of
hedging rules. Once the results are very similar, we may question the
importance of the 7 different approaches.
This is completely true. Because we optimize all policy and find the best shape of hedging policy for each type with the objective of maximum power generation over certain time. That is why all of policy seems the same in a case of power generation. but the rules of releasing water to produce hydropower were completely different. This differences could highly effect on the stability of the system, which will be discussed later in Figure 12, by presenting the box plot of reservoir head.
- It would increase the interest of the paper if the reader could have a better
understanding of the study area and the standard of the demand, evaporation
and water level data.
The required data for implementing this research are demand data, Evaporation data, water elevation-storage capacity and daily inflow data, which should be gathered or estimated. The first parameter is target demand data. In terms of water demand, Reservoirs with the purpose of hydropower are completely different with other types of reservoirs. To explain more, in a case of reservoir operation for municipal, the amount of water demand is fixed and release is based on water availability and demand. In a case of irrigation requirements, the demand varies from season to season depending on weather conditions and the annual demand pattern is repeated year after year. However, in the case of operation of reservoir for hydropower generation, the water demand is not fixed and varies though the constant power demand. It means that when stored in the reservoir is more, the head is also more and hence a smaller discharge from the reservoir may be sufficient to produce the power demand. However, if the storage is less, to produce the power demand, a large quantity of water must be discharged. To sum up the foregoing, the water demand for hydropower production is changeable and depends on the water head or water availability. So, there is not any specific target demand pattern available for hydropower reservoir system. Meanwhile, in order to determine the target demand in the case study, the historical data of release were analyze and the average of reservoir release during the operation period (2003-2012) was taken as a target demand (constant value:1 Mm3).
The daily evaporation data (mm) was collected from site 1030 (Latitude: 4.466, Longitude: 101.366, Altitude: 1545m) Cameron Highlands at Tanah Rata, Pahang, Malaysia.
The relation between water elevation-storage capacity is presented in Figure S1 (Supplementary file).
The next parameter that is so essential for study of reservoir operation is stream flow. However TNB did not have any stream flow gages to record river discharge coming into reservoir systems, some estimation has made for stream flow data based on the daily power station, which pass the turbines. It is remarkable to note that all the data were taken from Tenaga Nasional Berhad (TNB), Malaysia.
- The reservoir storage volumes constrain (2.3.3) is a consequence of the
main and first constrain: the reservoir head.
Reservoir storage volume and the reservoir head have direct relation. It seems that by make a limitation for one of this parameter (for example reservoir storage volume), the other one is also limited.
As you see the Figure S1. Relationship between elevation and storage capacity of Jor Reservoir in Supplementary file, you can find their relation.
Storage volume limited between minimum water level (1.6 Mm3) and storage at maximum operating level (2.8 Mm3).
minimum water level is around 486 m
maximum operating level is around 493.4 m
- In Figure 9 it would be helpful to represent the TNB, too. It is not possible for
me to understand the SHPHP result. It would be an added value if the authors
could discuss these results more deeply.
Box plot of TNB is also added.
The second aspect to point out is whenever the optimal operating policies apply, the water storage in the reservoir is strictly limited and did not exceed the surcharge storage (493.4 m) and inactive zones (486.1 m) in critical drawdown and fulfilled periods. While, the box plot of reservoir head in TNB operation shows the wide variation (481.9 m <head< 494 m) in water head and instability in reservoir operation. The results demonstrated that the reservoir head sometimes decreased blow the minimum operating level (486.1 m) and sometimes pass the maximum operating level (493.4 m), which could be dangerous for dam safety. So, this research could help the Jor reservoir manager to operate the reservoir system more efficiently and safety.
Another issue is all types of operating policies try to maintain some storage in a whole year for flood control. While, the maximum operating level of Jor reservoir located at the elevation of 493.4 m, all policies kept the mean water level (Q2) less than 493.4 m, except SHPHP policy, which maintain the water level in the highest permissible level (493.4m) to produce target power with the least release.
- The presentation of the different plots of water balance equation along the
study period would be very interesting, even in Supplementary File.
Rebuttal:
by calculating the water balance equation, the amount of storage in each time were obtain.
St = St-1 + It - Et - Rt - SPt
Since reservoir storage and the reservoir head have direct relation (as I discuss in the above questions), by discussing one of this parameter (boxplot reservoir head in this research (Figure 12), the other parameter (reservoir storage) is consequently considered. Because they follow the same results.
- There are some duplications in the text (e.g. several lines in the beginning
of results' chapter, in which the objectives of the paper are presented
unnecessarily for the second time).
The lines were removed
Additional minor concerns:
e.g. Line 111 – replace “KW” for “kW”. There are other cases.
I found 4 “KW” in the text and I change all of them to “kW”.
Line 304 – replace “Figure.S2” for “Figure 1”.
Done
Line 358 – TNB never appear before
The output of monthly mean power by using optimal operating policies and also Tenga Nasional Berhad (TNB) operation (Current operation policy) is shown in Figure 8.
TNB stands for Tenaga Nasional Berhad: Current operation policy
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In this study, the authors compared 7 different hedging policies for hydropower reservoir system operation. As case study, a single reservoir system in Malaysia has been considered.
The topic and the application fall within the scope of MDPI Water. The paper is well structured. It is an high quality work, but its readability and clarity of exposition can be improved. The manuscript, in its present form, contains few minor issues which need to be better discussed. Appropriate revisions to the following points should be undertaken in order to justify recommendation for publication.
1) The main issue is relative to the style of text and figures. Moderate changes of English language and style are required (see specific comments for some examples). Using acronyms, remember to report the full name the first time you used the acronym (e.g. PSO, TNB). Be uniform in the use of “,” as thousands separator (check all the occurrences both in the text and in tables/images). The quality of some of the figures included should be improved (avoid low resolution figures, use a uniform style for the hedging rules plots, remove the black outline from figures 3, 6, 8). See specific comments for further suggestions.
2) Clearly state if there is some innovative contribution (e.g. in the hedging rules or in the optimization algorithm), if any. If not, please, insert one or more references for each hedging rule shape and for the hybrid optimization algorithm.
3) Include a paper recently published on MDPI Water to enrich the bibliography:
Line 46:
Sangiorgio, M.; Guariso, G. NN-Based Implicit Stochastic Optimization of Multi-Reservoir Systems Management. Water 2018, 10, 303. https://doi.org/10.3390/w10030303
The sentence should become “… hedging rules, space rule, pack rule, linear decision rule, neural networks based rule [number related to Sangiorgio and Guariso, 2018], etc [3].”
4) Looking at figure 8, it is clear that all the 7 proposed control laws provide nearly the same performances. A comparison with the TNB shows that the optimized operating rules, in a certain sense, remove the seasonality from the power generation trajectory. In my opinion, this could be caused by the fact that in this study the authors considered only a single objective (power generation) whose demand is usually almost constant within the year. In the other hand, the TNB considers also other objectives with high dependence on the seasonality (e.g. minimize agricultural water, minimize the risk of floods, …). In this case, the comparison between the policies here identified and the historical one is not .It is necessary to better investigate this issue and declare the presence or absence of agricultural districts and/or flood risk areas.
Specific comments:
a) Lines 7-16. Align the authors affiliations.
b) Line 22. “Rarely applied for operation of hydropower reservoir”. Include this sentence only if you are sure. Don’t you find any other papers where hedging policies has been used to optimize hydropower production?
c) Line 32. “increase by 13%”: specify that the variation is with respect to the historical management”.
d) Line 89. Why “Figure. S1”? is it “Figure 1”?
e) Figure 1. Arrow missing under “Water level-Storage capacity data”.
f) Figure 1. Align properly the arrows directed to the block “Model constraints”.
g) Figure 1, section “Model Output”. I do not understand why the water head is considered apart from release and power generation. In fact, the power generation is a function of both the water head and the release (see equation 1) and thus the proposed scheme seems to be inadequate. Please, explain the reason why you did this or fix the figure.
h) Figure 2. Low quality. The ratio between height and width is probably not the one of the original image. Restore it.
i) Figure 3. Remove the black outline.
j) Figure 3. Remove the uppercase “R” in Rivers (last word of the caption).
k) Section 2.2. Report for each shape of the hedging rules how many and which are the parameters to optimize.
l) Section 2.2.1. Specify that K is the reservoir capacity.
m) Lines 132-134. Rephrase the sentence (“which the angle”, “the fraction”)
n) Line 143. “lines are smoother” does not have sense since we are considering straight lines. Use curves instead of lines.
o) Lines 161-162. Rephrase the sentence “The discrete …… gradation option”.
p) Figure 4. Poor quality. In particular, the tick labels on the axis are not readable.
q) Figure 4. Include a line which underline which is the actual release curve (as you did in figure 4 d). Use the same style for all the 7 hedging rules.
r) Figure 4. In my opinion, it is better to slit the figure in 4 different figures placed near the related section and remove the outline. The same goes for figure 5 and 6. Use a uniform and consistent style in all these figures is fundamental to have an high quality article.
s) Figure 4 d. There is a dashed horizontal line missing (HF3*Dt).
t) Lines 176-177. Rephrase the sentence (“which the main difference”, “Because. In …”).
u) Lines 183-184. Rephrase the sentence (“The mentioned….. non-linear way”)
v) Figure 5. The ratio between height and width is probably not the one of the original image. Restore it.
w) Figure 5. Report R0 and Rf on the y-axes instead of over the dashed lines.
x) Figure 5. When the available storage is greater than D+K, the curve has to be a constant straight line.
y) Figure 5. When the available storage is between D and D+K, how do you determined the shape of the curve? If it is linear, fix it on the plot. If not, explain which is the actual shape. There is the same issue in figure 6, after each T.
z) Figure 6. Why is the label of the y-axis different (release + spill)?
aa) Figure 6. Available water and Available storage are the same variable. Remove one of them consistently with figure 4 and 5.
bb) Lines 276-279. Rephrase the sentence. Advantages and disadvantages (or pros and cons) can be used instead of merits and demerits. “but around global optimum” is not clear.
cc) Line 281. “Garg was compared” should probably be “Garg compared”
dd) Lines 287-291. These lines contain lots of repetitions of the last paragraph of the previous page. Merge the two in a single paragraph.
ee) Figure 7. What are P (update P best) and G in the scheme?
ff) Line 319. “Afterwards, In …”. Remove the uppercase.
gg) Table 1. Fix the table style. Text inside rows with more than one line is not properly aligned.
hh) Table 1. Change the order of the columns and add/remove columns in order to made the table self-explaining. In the current form, it is hard to figure out the meaning of the columns and the relations between them.
ii) Table 1. The active storage is 1,205,290 m3. I cannot understand how the Jor reservoir storages are computed and how is the relation between them and the active storage.
jj) Lines 350-351. Rephrase the sentence.
kk) Table 2. Text inside rows with more than one line is not properly aligned.
ll) Line 399. What is the difference between inner outlier and outer fences? A short sentence which clarify this point should probably be included.
Author Response
Please see the answers on the attached file.
Thank you.
Author Response File: Author Response.pdf
Reviewer 3 Report
This paper investigates the applicability of seven competing hedging policies including four (4) customary forms of hedging (1PHP, 2PHP, 3PHP, DHP) and 3 new forms of hedging rules (SOPHP, BSOPHP, SHPHP) for hydropower reservoir operation. The models are constructed for simulation of the Batang Padang hydropower reservoir system, Malaysia, and optimised using a hybrid PSO-GA for maximising power generation. The paper is a novel study and is generally well-structured as it explains the methodology, the mathematical framework and the assumptions used, and the justification of the results and the conclusions. However, there are few critical points that should be addressed in the revised manuscript. Addressing these comments will improve the quality of the paper and help the general reader of the paper.
Major comments:
1. Novelties of the method. Please discuss the novelties of the manuscript in comparison with the previous studies of the authors at the study area. Please also highlight the similarities and differences of the previous works of the authors for the paper objectives.
2. The method is missing a validation part of the system. The authors have used a 10-year reference period for optimisation period of the hybrid PSO-GA. My suggestion to the authors is to use a validation procedure (i.e 1/3 of the data period = 3years) in the revised manuscript. This is quite important in demonstrating that the optimised hedging policies are able to produce acceptable and adequate results in the near future. Please address this critical issue in the revised manuscript.
3. Superiority of the hybrid optimisation algorithm. Please provide evidence that the used hybrid PSO-GA provide scientific merit when compared with other optimisation algorithms at the study area using the same principles and constraints. My suggestion to the authors is to compare their employed algorithm with another optimisation algorithm to demonstrate the superiority of the hybrid PSO-GA. This test will guarantee that the hybrid PSO-GA model is more reliable and could be used by other researchers for similar applications.
4. Please discuss the uncertainty bounds of the seven hedging policies and not only the mean and total power generation in the Discussion section. The results with mean and +/- values (min-max or at 95%) are quite useful for operational purposes and should be given in the revised manuscript.
For the motivations listed above, the paper in its present form needs revisions in order to evaluate the innovative character of the manuscript. However, the paper is of general interest for international audience and merits publication in Water Journal when the major comments are addressed. Addressing these comments will improve the quality of the paper and help the general reader of the paper.
Author Response
Please see the answers on the attached file.
Thank you.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Dear Authors,
The revised paper has marginal improvements over the previous submitted version. Based on my 1st review, the authors responses to reviewer and the revised version there are still few major points that should be clarified and addressed (see major comments to authors from my 1st review). My major comments are not fully addressed in the revised version (see also authors responses to reviewer and revised manuscript). Hence, again I recommend major revisions for the revised Manuscript No: water-383094. I strongly recommend to the authors to address my comments and to resubmit their manuscript as a new submission. The editor should guarantee that my comments are fully addressed in a potential re-review of the manuscript.
Major Comments
1. “1st review: Novelties of the method. Please discuss the novelties of the manuscript in comparison with the previous studies of the authors at the study area. Please also highlight the similarities and differences of the previous works of the authors for the paper objectives.” In the revised manuscript, the authors discuss several issues but they not compare their previous studies (No 6,11,17 of the reference list) with the novelties of this study. Hence, this comment is not addressed in the revised manuscript.
2. “1st review: The method is missing a validation part of the system. The authors have used a 10-year reference period for optimisation period of the hybrid PSO-GA. My suggestion to the authors is to use a validation procedure (i.e 1/3 of the data period = 3years) in the revised manuscript. This is quite important in demonstrating that the optimised hedging policies are able to produce acceptable and adequate results in the near future. Please address this critical issue in the revised manuscript.” This comment is not addressed and I completely disagree with the authors. A validation part will always reveal the capabilities of any model for operational purposes. Furthermore, nowadays the problem of limited data could be handled using cross validation techniques. Hence, I strongly believe that a validation part is needed to reveal the potential superiority of the developed method.
3. “1st review: Superiority of the hybrid optimisation algorithm. Please provide evidence that the used hybrid PSO-GA provide scientific merit when compared with other optimisation algorithms at the study area using the same principles and constraints. My suggestion to the authors is to compare their employed algorithm with another optimisation algorithm to demonstrate the superiority of the hybrid PSO-GA. This test will guarantee that the hybrid PSO-GA model is more reliable and could be used by other researchers for similar applications.” Again, in the revised manuscript the authors discuss several issues but they did not address my comment. I would like to see a comparison with another optimisation algorithm to demonstrate the superiority of the hybrid PSO-GA.
4. “1st review: Please discuss the uncertainty bounds of the seven hedging policies and not only the mean and total power generation in the Discussion section. The results with mean and +/- values (min-max or at 95%) are quite useful for operational purposes and should be given in the revised manuscript.” This comment has been addressed in the revised manuscript.
Author Response
Comments and Suggestions for Authors
Dear Authors,
The revised paper has marginal improvements over the previous submitted version. Based on my 1st review, the authors responses to reviewer and the revised version there are still few major points that should be clarified and addressed (see major comments to authors from my 1st review). My major comments are not fully addressed in the revised version (see also authors responses to reviewer and revised manuscript). Hence, again I recommend major revisions for the revised Manuscript No: water-383094. I strongly recommend to the authors to address my comments and to resubmit their manuscript as a new submission. The editor should guarantee that my comments are fully addressed in a potential re-review of the manuscript.
Major Comments
1. “1st review: Novelties of the method. Please discuss the novelties of the manuscript in comparison with the previous studies of the authors at the study area. Please also highlight the similarities and differences of the previous works of the authors for the paper objectives.” In the revised manuscript, the authors discuss several issues but they not compare their previous studies (No 6,11,17 of the reference list) with the novelties of this study. Hence, this comment is not addressed in the revised manuscript.
The references that the reviewer mentioned (No 6,11,17 of the reference list) are my previous publication.
Ref 6: Tayebiyan, A.; Mohammad, T.A. Optimization of cascade hydropower system operation by genetic algorithm to maximize clean energy output. Environ. Health Eng. Manag. J. 2016, 3, 99–106
In this research only one of the opertaional policy ( Discrete hedging policy) was discussed.
Case study: Cameron Highland Hydropower reservoir system
Optimization technique: Real coded genetic algorithm
Ref11: Tayebiyan, A.; Ali, T.A.M.; Ghazali, A.H.; Malek, M. Optimization of exclusive release policies for hydropower reservoir operation by using genetic algorithm. Water resour.manag. 2016, 30, 1203-1216.
In this research, exclusive hedging policies BSOPHP, SOPHP, SHPHP were compared and discussed.
Case study: Cameron Highland Hydropower reservoir system
Optimization technique: Real coded genetic algorithm
Ref 17: Tayebiyan, A.; Ali, T.A.M.; Ghazali, A.H.; Megat Mohd Noor, M. J. Simulation and optimization of multiple points hedging policies for operation of Cameron Highland hydropower reservoir system, Malaysia. Desalination Water Treat.2018, 120, 205–216.
In this research, continuse hedging policies namely 1PHP, 2PHP, and 3PHP were compare and discuss discussed.
Case study: Cameron Highland Hydropower reservoir system
Optimization technique: Real coded genetic algorithm
This research.
In this research, the applicability of seven competing hedging policies including 4 customary forms of hedging (1PHP, 2PHP, 3PHP, DHP) and exclusive form of hedging policies (SOPHP, BSOPHP, SHPHP) for hydropower reservoir operation were investigated, which have never seen before for any reservoir system.
Case study: Batang Padang Hydropower reservoir system
Optimization technique: Hybrid PSO-GA
In an overall, the similaritis and differences of this work with my previous works are
Although the rule curve of release are similar but the previous models are constructed based on the characteristics of cameron hghland hdro system, while in this research the models were constructed based on the characteristics of Batang Padang hydro system.
Objective function of previous researches and this research are the same.(Maximizing the total power generation during time period)
Differences:
The case study are different
In this research, three optimization techniques namely PSO, GA, and Hybrid PSO-GA were investigated. While, in the previous researches, only GA was used as an optimization technique.
Novelty: at the end of introduction
In this research, the applicability of seven competing hedging policies including 4 customary forms of hedging policy (1PHP, 2PHP, 3PHP, DHP) and 3 new forms of hedging rule (SOPHP, BSOPHP, SHPHP) for hydropower reservoir operation were investigated and compared. The proposed methodology has never been applied to any reservoir system particularly for maximization of power generation. In order to compare the competing hedging policies, Batang Padang hydro scheme, Malaysia was selected as a case study. Meanwhile, daily mathematical models that are based on hedging rules and optimization of power generation from operation of reservoir system were constructed in Matlab R2011b. Afterwards, three optimization algorithm namely particle swarm optimization (PSO), genetic algorithm (GA), and combination of PSO and GA, which is called hybrid PSO-GA were linked to one of the constructed model (1PHP as a test) in order to find the most effective algorithm. Since the results demonstrated the superiority of the hybrid optimization algorithm compared to either PSO or GA, the hybrid PSO-GA were linked to each constructed model in order to find the global maximization. Maximizing the total power generation over the operational periods is chosen as an objective function, while physical and operational limitations were satisfied.
The hybrid PSO-GA was used in water resources management particularly for finding optimum release from impounding reservoirs [23, 28]. However, in this study, it is used for constrained optimization problems including maximization of power generation.
2. “1st review: The method is missing a validation part of the system. The authors have used a 10-year reference period for optimisation period of the hybrid PSO-GA. My suggestion to the authors is to use a validation procedure (i.e 1/3 of the data period = 3years) in the revised manuscript. This is quite important in demonstrating that the optimised hedging policies are able to produce acceptable and adequate results in the near future. Please address this critical issue in the revised manuscript.” This comment is not addressed and I completely disagree with the authors. A validation part will always reveal the capabilities of any model for operational purposes. Furthermore, nowadays the problem of limited data could be handled using cross validation techniques. Hence, I strongly believe that a validation part is needed to reveal the potential superiority of the developed method.
Author comment:
Three years data (2010-2012) are used for model validation. This also is discussed in the manuscript text (see pages 14 and 15).
The results of monthly mean power generation (MW) from 2003-2009 (the time period that the models were constructed) were compared with the results of monthly mean power generation (MW) from 2010-2012 as shown in (Figure 12). Figure 12 shows that the all models namely 1PHP, 2PHP, 3PHP, DHP, SOPHP, BSOPHP, SHPHP are able to produce acceptable and adequate results and there are no significant difference between the output of monthly mean power generation during 2003-2009 and 2010-2012.
3. “1st review: Superiority of the hybrid optimization algorithm. Please provide evidence that the used hybrid PSO-GA provide scientific merit when compared with other optimization algorithms at the study area using the same principles and constraints. My suggestion to the authors is to compare their employed algorithm with another optimization algorithm to demonstrate the superiority of the hybrid PSO-GA. This test will guarantee that the hybrid PSO-GA model is more reliable and could be used by other researchers for similar applications.” Again, in the revised manuscript the authors discuss several issues but they did not address my comment. I would like to see a comparison with another optimization algorithm to demonstrate the superiority of the hybrid PSO-GA.
In order to choose the best algorithm among PSO, GA, and hybrid PSO-GA, the algorithms were linked to one of the constructed model (1PHP) in order to find the most appropriate technique. The comparison results showed that the hybrid PSO-GA can simultaneously find a promising solution and speeding up the convergence (Figure 11). Based on the given results, the hybrid PSO-GA algorithm found the highest fitness value in less number of iterations and this demonstrate the superiority of the hybrid optimization algorithm compared to PSO or GA. Since the performance of hybrid PSO-GA is better than PSO and GA, so this algorithm was used in this research to optimize other constructed hedging policies.
Author Comments:
As your request, hybrid PSO-GA, PSO and GA are linked to one of the operational policy [one point hedging policy (1PHP)] model in order to compare the output results of these three optimization algorithm based on the amount of global optimization (fitness value) and the number of iteration. The comparison results is shown in figure 11.
4. “1st review: Please discuss the uncertainty bounds of the seven hedging policies and not only the mean and total power generation in the Discussion section. The results with mean and +/- values (min-max or at 95%) are quite useful for operational purposes and should be given in the revised manuscript.”
This comment has been addressed in the revised manuscript.
Author Response File: 
Author Response.doc
Round 3
Reviewer 3 Report
The paper in its present and revised form merits publication in Water journal. The authors have done a remarkable effort to address all my major and minor comments, which are explained and analysed in the authors’ response document and the revised manuscript. As a conclusion, based on the authors’ response to reviewers and the revised manuscript, all major and minor comments have been addressed and I recommend accepting for publication the revised manuscript in its present form.
