Power Regulation and Fault Diagnostics of a Three-Pond Run-of-River Hydropower Plant

Round 1

Reviewer 1 Report

This article looks at integrating a three tank hydraulic system with a Francis turbine and adds two levels of control to power generation. It generates a tool to solve the faults that may occur in terms of the opening of the gates and the input of the flow necessary to obtain the required power. The tolerances and the response of the system are analyzed to avoid any loss of production. Ultimately, the objective is to increase knowledge about the errors that can occur when operating the pond system and find feasible solutions that avoid reducing production capacity.

The authors have developed a very intense work and the mathematical analysis reveals it, with more than one hundred formulas.

However, beyond the novelty of the hydraulic generation system, the solution proposed to optimize its operation is not excessively novel. The control of energy production based on the water level, with the linked set of the gates and the detection and solution of faults does not present any novelty in relation to the normal operation of an installation that works with levels. The simplest SCADAs can collect the conditions discussed in the article and articulating the operation of the installation under more complex conditions than the height of the water and / or the opening of gates.

The special edition of the magazine is titled CFD Modeling and Simulation of Water Turbines and the presented paper is based on developing the mathematics corresponding to the operating events of the system, their errors, tolerances and the way to optimize the operation for achieve the desired energy production. At no time in the article is reference to CFDs or if software has been used or of what kind. It is said that the results are obtained from simulations, but it is not clear whether they come from CFDs or other software that solves the proposed equations according to the analyzed boundary conditions. It is therefore necessary to specify in more detail how these simulations that provide the results on which the conclusions are based have been carried out.

Analyzing the generation system itself on which the article is based and whose origin is [1], it is concluded that it is a very simple system but that in a certain way it is conditioned by the flow provided by the river, being aware of the seasonal variability they usually present. Some reference to the possible environmental difficulties that an installation of this style could entail and the influence that the need to maintain the ecological flow in the riverbed could have when operating the system, conditioning the entry of water, is missing.

In relation to the study of failures and their possible solutions, the analysis is based on identifying leaks and blockages that the malfunction of the gates and the system in general, can lead to a loss of general performance. The mathematical approach seems correct, insofar as it includes the hypotheses raised by the authors, but the depth of these is perhaps not sufficient for the publication of the article. The same authors comment that future work would be to delve more deeply into the errors contemplated and expand those 100 W (operation 800 kW to 2 MW) that they currently correct.

In addition, the entire article needs a thorough review for clarity in writing and grammatical errors in the language. A review by a native person is recommended. To extremely long sentences that cause loss of meaning, misprints are added in numerous words, which results in a difficult reading to understand in some sectors, making it seem much more difficult than it really is.

Author Response

Responses to Reviewer 1

This article looks at integrating a three-tank hydraulic system with a Francis turbine and adds two levels of control to power generation. It generates a tool to solve the faults that may occur in terms of the opening of the gates and the input of the flow necessary to obtain the required power. The tolerances and the response of the system are analyzed to avoid any loss of production. Ultimately, the objective is to increase knowledge about the errors that can occur when operating the pond system and find feasible solutions that avoid reducing production capacity.

The authors have developed a very intense work and the mathematical analysis reveals it, with more than one hundred formulas.

Authors Response: First of all, thank you very much for taking the time to review our paper and commenting on it to improve it. Listed below are the answers to your valuable comments.

Comment:  " However, beyond the novelty of the hydraulic generation system, the solution proposed to optimize its operation is not excessively novel. The control of energy production based on the water level, with the linked set of the gates and the detection and solution of faults does not present any novelty in relation to the normal operation of an installation that works with levels. The simplest SCADAs can collect the conditions discussed in the article and articulating the operation of the installation under more complex conditions than the height of the water and / or the opening of gates."

Authors Response: Thank you for your insightful comments. The novelty of the system can be seen in three ways:

1. The considered three pond hydraulic system to our knowledge is not analyzed for power generation
2. Its integration with the turbine and proposing the nested controls for power generation and regulation is our contribution; previous work on it is done on atraditional hydropower systems with a single head pond.
3. The fault models are presented which are novel when discussed in conjunction with the hydraulic system.

Comment:  "The special edition of the magazine is titled CFD Modeling and Simulation of Water Turbines and the presented paper is based on developing the mathematics corresponding to the operating events of the system, their errors, tolerances, and the way to optimize the operation for achieving the desired energy production. At no time in the article is reference to CFDs or if software has been used or of what kind. It is said that the results are obtained from simulations, but it is not clear whether they come from CFDs or other software that solves the proposed equations according to the analyzed boundary conditions. It is therefore necessary to specify in more detail how these simulations that provide the results on which the conclusions are based have been carried out"

Authors Response: The software used to simulate all the results is Matlab and Simulink. Now it is also mentioned and well presented in the revised manuscript. The simulated results through Matlab and Simulink support the mathematical equations of the system.

 

Comment:  " Analyzing the generation system itself on which the article is based and whose origin is [1], it is concluded that it is a very simple system but that in a certain way it is conditioned by the flow provided by the river, being aware of the seasonal variability they usually present. Some reference to the possible environmental difficulties that an installation of this style could entail and the influence that the need to maintain the ecological flow in the riverbed could have when operating the system, conditioning the entry of water, is missing."

Authors Response: That is very loosely inspired by the canal type irrigation system mostly used in Pakistan and south Asia. Also the focus of the paper is on the control and fault modeling and diagnosis of the presented hydraulic system and is not concerned with the flow of the river before it. Adding the variation in flow and its effects on the proposed system is our future research goal.

Comment: " In relation to the study of failures and their possible solutions, the analysis is based on identifying leaks and blockages that the malfunction of the gates and the system in general, can lead to a loss of general performance. The mathematical approach seems correct, insofar as it includes the hypotheses raised by the authors, but the depth of these is perhaps not sufficient for the publication of the article. The same authors comment that future work would be to delve more deeply into the errors contemplated and expand those 100 W (operation 800 kW to 2 MW) that they currently correct."

 

Authors Response: Thank you very much for your valuable comments. This paper is an initial part of the ongoing work and it will be delved more widely and deeply in the future. However, in accordance to your comments, we have further delved in the mathematical model of one of the faults as an example case in the appendix. We believe the effects of the other faults will require similar steps to reach the results. 

 

Comment: " In addition, the entire article needs a thorough review for clarity in writing and grammatical errors in the language. A review by a native person is recommended. To extremely long sentences that cause loss of meaning, misprints are added in numerous words, which results in a difficult reading to understand in some sectors, making it seem much more difficult than it really is."

 

Authors Response: Thanks for your comment. As per your recommendation, the paper is proofread by a native person all mistakes like extremely long sentences, misprints are added in numerous words, and typos are corrected in the revised manuscript. Besides, the paper is proofread by authors as well, and grammatical errors and errors in language are resolved in the revised manuscript.  

Reviewer 2 Report

processes-1557848-peer-review

In the Introduction I found the following sentence a little strange after the text before which stated that run of river power plants often have a negative impact , so why are these an alternative, and what do you call a traditional hydropower plant?

Due to these reasons run of river hydro plants are an alternative to the traditional hy- 51
dropower plants.

 

In the caption of Fig 7 it should be explained what is x0, x2 and x2.

 

Regarding

During the in- 54 dustrial processes many known and unknown faults or errors can occur, which results in 55 undesirable outputs or off times which in turn can result in monetary loss and loss of 56 service on the consumer side.

Please at this stage provide information on real-life faults that occur in running hydroplants. This is relevant as it is connected to the solutions provided, in particular to what sensors are required and how monitoring can be set up best.

It is not very clear to me what type of faults the authors refer when talking about maintenance: Many a times it is not desired to shut down the system for 57
maintenance at the occurrence of every fault

 

The authors might have missed some relevant papers that have been published in this field, please have a look and take into account useful references and information:

Dong Liang et al 2019 J. Phys.: Conf. Ser. 1176 062058    openaccess

https://journals.sagepub.com/doi/pdf/10.1177/1748006X16689407

Furthermore, references 12 and 13 are from 2010, but the references are not complete. It should be mentioned that ref 12 is a PhD thesis, and an appropriate link to the internet site should be provided for reader to find the thesis.

Same for Ref 13.

Ref 11 seems a PhD thesis from the same German University and Refs 12 and 13. Referring to a thesis is fine, but it is more appropriate to refer to peer-reviewed publications, and it is difficult to believe all three authors of these three theses have no peer-reviewed publications.

 

The following

'For the sake of simplicity, the flow of river is taken as 141
a combination of constant base flow rate and a short term variable flow rate and is mod- 142 eled by a sinusoidal function, while the long term variation is considered constant and is 143 included in the base flow rate' raises two questions: how realistic is it that the short term variation is sinusoidal, I would think there is no reason for this. Secondly, the long term variation is considered constant, so this is no variation? (the fixed parameter m)

 

In Figure 2, and the caption should describe this too more detailed, it should be indicated what these ‘ objects ‘  control. I am missing the ponds, and indication where the controllers are placed, etcetera. I would expect to see in Fig 2 what you mention at the start of 3.1: has the controllable variables ?1 and ?2the valves be- 216
tween ponds 1 and 0 and 2 and 0 and the flow of river in the ponds 1 and 2 through the 217 weir gates.

 

In

'A fuzzy controller is a linguistic and intuitive controller as it takes the input in form 237 of linguistic variables and gives output in the same way. A fuzzy system consists of a set 238 of IF-THEN rules. In the first part, the input variable is a crisp value which is fuzzified to 239 convert it into a linguistic variable.'

the authors jump into a fuzzy logic system , could the authors please elucidate what the objective is to incorporate this at this stage.

 

I am a little surprised that

'As the efficiency of the turbine is not in our control'

As efficiency is not necessarily independent of rotational speed and therefore a potential parameter to steer the output power

 

In the Introduction it is said this manuscript is about ‘to identify the occurring faults’ and in the Introduction ‘Another important part of 53
industrial process and power generation is the field of fault diagnostics’ During the in- 54 dustrial processes many known and unknown faults or errors can occur, which results in 55 undesirable outputs or off times which in turn can result in monetary loss and loss of 56 service on the consumer side’ and later on ‘The magnitudes of both the 777
saturation and leakage faults are arbitarily taken at 10%’ . In particular regarding the latter, but in total with the first two citations, is this a purely imaginary and theoretical exercise then? Where can I see this fault diagnostics has to do with a real life plant?

It is not clear to me how input from sensors is taken into account in the fault diagnosis scheme. And if no sensor input is used, is this indeed a study on an imaginary system?

 

This manuscript contains a huge number of formulae, but to the reader the main story may become a little fuzzy. It is proposed to move most formulae to Supplementary Material, and only put the main, key and used formulae, in the main text. This should improve the readability greatly, as it is often not really transparent where the discussion goes regarding the problem setting formulated at the beginning. The main thread of argumentation should be better expressed, and the main text focussing on that, and therefore shortened to avoid too many details in the main text.

Author Response

Reviewer 2 Comments

Authors Response: Thank you very much for taking the time to read and review the paper and giving your valuable suggestions, these are very much helpful. Now I have addressed all of your points in the revised manuscript.

 

Comment: " In the Introduction I found the following sentence a little strange after the text before which stated that run of river power plants often have a negative impact, so why are these an alternative, and what do you call a traditional hydropower plant?

Due to these reasons run of river hydro plants are an alternative to the traditional hy- 51 dropower plants."

 

Authors Response: The traditional hydropower plants are the ones most commonly used with a dam constructed and have a large reservoir and are generally have rather large outputs in the range of greater than 100 MWs. The negative impact of a traditional hydropower plant is the effect on the environment by flooding the valley for the reservoir and displacing villages and settlements for the construction of the reservoir. The alternative to it is a run of river hydrpower plant which avoids these impacts most of the time as the affected area of its construction is mostly smaller.

 

Comment: " Please at this stage provide information on real-life faults that occur in running hydro plants. This is relevant as it is connected to the solutions provided, in particular to what sensors are required and how monitoring can be set up best."

Authors Response: Thank you for pointing that out, some of these faults are discussed in [2] and are added in the relevant section. The faults discussed in [2] are actuator sluggishness fault, turbine flow fault and turbine speed sensor fault.

Comment: " In the caption of Fig 7 it should be explained what is x0, x2 and x2."

Authors Response: Thank you for pointed that out. These are the pond levels and are now mentioned with the figure in the paper.

Comment: " It is not very clear to me what type of faults the authors refer when talking about maintenance: Many a times it is not desired to shut down the system for 57
maintenance at the occurrence of every fault"

Authors Response: Agreed, that was not clear in the text and is now addressed in the paper..The type of faults which require active maintenance are more serious in nature like loss of power or some serious physical damage to the system as a result of any natural or manmade disaster. Otherwise maintenance is scheduled.

 

Comment: " The authors might have missed some relevant papers that have been published in this field, please have a look and take into account useful references and information:

Dong Liang et al 2019 J. Phys.: Conf. Ser. 1176 062058    openaccess

https://journals.sagepub.com/doi/pdf/10.1177/1748006X16689407"

Authors Response: Thank you for sharing these references, the information from these are now added, and these references are cited in the revised manuscript.

 

Comment: "Furthermore, references 12 and 13 are from 2010, but the references are not complete. It should be mentioned that ref 12 is a PhD thesis, and an appropriate link to the internet site should be provided for reader to find the thesis.

Same for Ref 13.

Ref 11 seems a PhD thesis from the same German University and Refs 12 and 13. Referring to a thesis is fine, but it is more appropriate to refer to peer-reviewed publications, and it is difficult to believe all three authors of these three theses have no peer-reviewed publications"

 

Authors Response: All these three are PhD thesis from University of Duisburg-Essen, German. Now their online link is provided and the references are completed.

Comment: "'For the sake of simplicity, the flow of river is taken as 141 a combination of constant base flow rate and a short term variable flow rate and is mod- 142 eled by a sinusoidal function, while the long term variation is considered constant and is 143 included in the base flow rate' raises two questions: how realistic is it that the short term variation is sinusoidal, I would think there is no reason for this. Secondly, the long term variation is considered constant, so this is no variation? (the fixed parameter m)"

 

Authors Response: For the sake of simplicity to model small scale fast variations compared to long term variations, the river flow is taken as a combination of constant and variable parameter which is sinusoidal. The sinusoidal terms models the short term variations during this interval there is a negligible change in long term variations therefore we kept it constant. All this is done to capture the short term variations and keeping the model tractable. In addition, our assumption is in line with the results from [1] which does not consider the river flow and the results of our paper the results are comparable and having a more accurate flow model of the river in the future will not change the results of the system much. Also due to the time window of the concern the long term variation is considered constant.

 

Comment: " In Figure 2, and the caption should describe this too more detailed, it should be indicated what these ‘objects ‘control. I am missing the ponds, and indication where the controllers are placed, etcetera. I would expect to see in Fig 2 what you mention at the start of 3.1: has the controllable variables?1 and ?2the valves be- 216
tween ponds 1 and 0 and 2 and 0 and the flow of river in the ponds 1 and 2 through the 217 weir gates."

Authors Response: The controllers in figure 2 are now highlighted. Also for the next part of the question please refer to figure 2, 3 and 4 of [1] as it deals with the detail of the level controller.

Comment: "'A fuzzy controller is a linguistic and intuitive controller as it takes the input in form 237 of linguistic variables and gives output in the same way. A fuzzy system consists of a set 238 of IF-THEN rules. In the first part, the input variable is a crisp value which is fuzzified to 239 convert it into a linguistic variable.'

the authors jump into a fuzzy logic system; could the authors please elucidate what the objective is to incorporate this at this stage."

The objective is to describe the purpose of the level controller.

" I am a little surprised that

'As the efficiency of the turbine is not in our control'

As efficiency is not necessarily independent of rotational speed and therefore a potential parameter to steer the output power"

Authors Response: Agreed, the rotational speed of the turbine will be incorporated in the future work. This paper takes the turbine model from [2] and the efficiency of the turbine is considered constant in that model. Our main contribution is to integrate 3 pond system to a turbine and apply fault diagnosis technique.

Comment: " In the Introduction it is said this manuscript is about ‘to identify the occurring faults’ and in the Introduction ‘Another important part of 53
industrial process and power generation is the field of fault diagnostics’ During the in- 54 dustrial processes many known and unknown faults or errors can occur, which results in 55 undesirable outputs or off times which in turn can result in monetary loss and loss of 56 service on the consumer side’ and later on ‘The magnitudes of both the 777
saturation and leakage faults are arbitarily taken at 10%’ . In particular regarding the latter, but in total with the first two citations, is this a purely imaginary and theoretical exercise then? Where can I see this fault diagnostics has to do with a real life plant?

It is not clear to me how input from sensors is taken into account in the fault diagnosis scheme. And if no sensor input is used, is this indeed a study on an imaginary system?"

Authors Response: The data for the system errors is taken from the sensors of the system, as shown in the graphs present. However for the ease to develop the fault model the effects of the faults are taken mathematically rather than through the sensor outputs.

Comment: " This manuscript contains a huge number of formulae, but to the reader the main story may become a little fuzzy. It is proposed to move most formulae to Supplementary Material, and only put the main, key and used formulae, in the main text. This should improve the readability greatly, as it is often not really transparent where the discussion goes regarding the problem setting formulated at the beginning. The main thread of argumentation should be better expressed, and the main text focussing on that, and therefore shortened to avoid too many details in the main text."

 

Authors Response: Your suggestions are taken into consideration, but due to the comments of reviewer 1 who asked to delve more deeply in the mathematical model, the numbers of equations are not reduced. However, to improve the readability of the paper, the fault model and the fault diagnostic section has been divided into subsections to make the story more clear and readable. Also, the summary of the flow of the paper is rewritten more clearly in the introduction.

Reviewer 3 Report

The article is justified and contains the right side of the theoretical considerations of the research problem. An interesting element of the research problem is the presented control system and the failure model, which is mathematically well conditioned. The presented results of the analyzes and the formulated conclusions clearly present the background and stages of solving the research problem. Congratulations!

Author Response

Reviewer 3 Comments

Comment: The article is justified and contains the right side of the theoretical considerations of the research problem. An interesting element of the research problem is the presented control system and the failure model, which is mathematically well conditioned. The presented results of the analyzes and the formulated conclusions clearly present the background and stages of solving the research problem. Congratulations!

 

Authors Response: Dear evaluator, thank you very much for deeply reviewing, appreciating, and recommending this paper for publication in processes. Many thanks to you for the time and effort that you put in the work.

 

Round 2

Reviewer 1 Report

Many thanks to the authors for their correct clarifications and their effort to improve the paper.

I agree with them on the certain degree of novelty that addressing the issue of failures presents. The paper would be more forceful if it focused on it. I want to understand that future studies will delve into it and that this paper would be an introduction.

Congratulations on the improvement in clarity in the presentation. I trust that for future papers they will be more concise. But even so, the paper needs a new wording. I enclose two examples of the difficulty for the reader:

"Where ?1 is the level of water in Pond 1, ? is the cross-sectional area of ​​the pond, ? is the cross-sectional area of ​​the duct connecting the ponds 1 and pond 0,?1is the controllable valve between the ponds 1 and 0, ȵ is the dimensionless water flow constant of value 0.98 [16] and ? is the gravitational constant of 9.8??2⁄."

"Fault tolerance control in its essential form simply corrects the output parameters of a system in the case of occurrence of the fault. In all the faults discussed above, fault tolerant action is not necessary in the case of the occurrence of each and every fault. If the effects of the fault are only apparent during the transient state, they will disappear when the system enters steady state thus negating the need to have dedicated fault tolerance. state, the fault tolerance is directed at the correction of the output which is actually important or critical.In the case of this system, the critical output which needs to have fault tolerance is the output power, while the actual levels of the ponds are not important."

Author Response

Thank you for taking the time and reviewing the paper again. Now we shall be replying to your comments.

 

I agree with them on the certain degree of novelty that addressing the issue of failures presents. The paper would be more forceful if it focused on it. I want to understand that future studies will delve into it and that this paper would be an introduction.

Congratulations on the improvement in clarity in the presentation. I trust that for future papers they will be more concise. But even so, the paper needs a new wording. I enclose two examples of the difficulty for the reader:

We have added some sentences in the introduction and the conclusion to improve the focus on our work and clarify the future directions.

Furthermore, for the clarity in the presentation, we have done English corrections through experts and have rewritten most of the paper for more clarity, and ease of reading. In our corrections, we have changed many of the sentences to be more concise and meaningful and removed ambiguity.

Reviewer 2 Report

the authors jump into a fuzzy logic system; could the authors please elucidate what the objective is to incorporate this at this stage."

The objective is to describe the purpose of the level controller.

" I am a little surprised that

'As the efficiency of the turbine is not in our control'

As efficiency is not necessarily independent of rotational speed and therefore a potential parameter to steer the output power"

Authors Response: Agreed, the rotational speed of the turbine will be incorporated in the future work. This paper takes the turbine model from [2] and the efficiency of the turbine is considered constant in that model. Our main contribution is to integrate 3 pond system to a turbine and apply fault diagnosis technique.

REFEREES COMMENT IN ROUND 2:

The authors did not reply to the question about why fuzzy logic was incorporated at this stage.

I feel the reply of the authors to the other point is appropriate and understandable, but please also mention this in the text as this will make it all more transparent for the reader why things are done they were done.

Comment: " In the Introduction it is said this manuscript is about ‘to identify the occurring faults’ and in the Introduction ‘Another important part of 53
industrial process and power generation is the field of fault diagnostics’ During the in- 54 dustrial processes many known and unknown faults or errors can occur, which results in 55 undesirable outputs or off times which in turn can result in monetary loss and loss of 56 service on the consumer side’ and later on ‘The magnitudes of both the 777
saturation and leakage faults are arbitarily taken at 10%’ . In particular regarding the latter, but in total with the first two citations, is this a purely imaginary and theoretical exercise then? Where can I see this fault diagnostics has to do with a real life plant?

It is not clear to me how input from sensors is taken into account in the fault diagnosis scheme. And if no sensor input is used, is this indeed a study on an imaginary system?"

Authors Response: The data for the system errors is taken from the sensors of the system, as shown in the graphs present. However for the ease to develop the fault model the effects of the faults are taken mathematically rather than through the sensor outputs.

REFEREES COMMENT IN ROUND 2:

Point taken, but please also explain this in the text of the manuscript.

Author Response

Thank you very much to take the time to review our paper again. We shall be replying to your comments now.

The authors did not reply to the question about why fuzzy logic was incorporated at this stage.

I feel the reply of the authors to the other point is appropriate and understandable, but please also mention this in the text as this will make it all more transparent for the reader why things are done they were done.

Apologies, that question was mistakenly overlooked by us. The main reason for using fuzzy logic for level control was the flexibility of the fuzzy logic controller and its ability to independently regulate the ponds level at three different levels. The part having different levels for the head ponds is also part of future direction and part of future proofing the system.  Also as it was found out in designing the system, fuzzy control has more flexibility and better results than compared to PID control in specific to our system application. This also has been added to the paper as:

For the level control, a fuzzy controller is used due to its flexibility, speed and higher accuracy as compared to a traditional PID controller

For the second part of the question, these are now added to the paper.

According to equation (34), the rotational speed of the turbine can also be used as a control parameter but that is kept as constant and is left for future work.

Point taken, but please also explain this in the text of the manuscript.

These are now added to the paper.

The graphs of the residues are generated through the system sensors in Matlab/Simulink, while to aid in calculations the fault effects were taken mathematically.

Round 3

Reviewer 1 Report

The review of this article has been divided into three stages.
A first review, where the low quality of the article is detected, in terms of the little novelty and of course an extremely low level in English.
In a second review, a minimal increase in novelty is seen in the issue of errors that can save the article and the authors state that an English native has modified the wording. Something that is totally false because the level of its writing was still unacceptable.
In a third review, the increase in the level of English is detected. This is how it should have been sent the first time. So that we reviewers could dedicate our time to analyze the theory and practice of the proposed research and not act as language reviewers.
Now that it can be read in a logical way and analyzing the article itself, the only thing that is salvageable is the part of the failures that can be explained in 10 pages. Those that remain up to 39 all they achieve is to divert attention from the only minimally valuable part of the paper.

Author Response

Dear reviewer, 
Thanks for the efforts and time that you put into deeply reviewing our paper.
Your comments in rounds 1 and 2 really improved the manuscript as per the requirement of the journal. Your comment in round 3 is to reduce the length of the paper. However, We are not reducing the manuscript because if we reduce the manuscript length then major discontinuities in writeup will be created, and it will be difficult for readers to understand. Therefore, you are requested to approve the manuscript for publication. Your cooperation in this regard will be highly appreciable. Thanks for understanding the situation