Computational Tools for Supporting the Operation and Management of Water Distribution Systems towards Digital Transformation
Round 1
Reviewer 1 Report
Dear authors,
I have read the manuscript ‘Computational tools for supporting the operation and management of water distribution systems towards digital transformation’ submitted to Water. This study presents tools developed in two research projects aimed at improving the water management and the operations activities of water utilities.
I would like to start by saying that the article is well written, pleasant to read with a smooth English and a clear structure. Given the scope of the paper, the structure of the manuscript is appropriate, with a sequence of ‘small’ sections that present all the developed tools, without going too much into detail in the developed methodologies. Therefore, although I would be intrigued to see something more about some methodologies, I also understood that it would not be the scope of the paper.
I enjoyed reading the discussion proposed by the authors, who were firstly honest when they showed the real limitations of their tools, and then made a digression over their vision of ‘what will be’ the future of water management. I found it interesting, although the authors could provide some more details alongside methodologies. Still, I appreciated the discussion over imputation (that strangely has been tackled by scientific literature only in the last few years in our field and please put some references about it) and general AI. Although the scope of the paper is not a review of current state of water management, still the authors provided a good view of what is going on in the field.
Finally, the article proposes a close view to what is the state of the art in water management, and could be a very nice source of inspiration for many water utilities worldwide. Therefore, I believe that the paper should be published.
Author Response
The authors would like to thank the Reviewer for the time taken to review our manuscript and for the valuable comments and suggestions on the manuscript. The manuscript has been thoroughly revised in order to attend to suggestions, namely, by incorporating some more details alongside methodologies.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Computational tools for supporting the operation and management of water distribution system towards digital transformation
In this manuscript, the author presents several computational tools to aid the management of water distribution systems, focusing on digital transformation. The content of the paper is clear and concise. Some additional details should be addressed as per the comments below.
Introduction
· Page 1, lines 89; KPI is first mentioned here, and it is not described. Please provide the description, and remove the description in the next mentions.
Data integration and analytics platform
· Page 3, line 117; “Portuguese Water and Wastewater Authority” already abbreviated as “ERSAR” in the introduction.
· Figure 1; for visual presentation, it is advisable to have nothing highlighted in the screenshot unless the highlighted part contributes to the discussion.
· The water balance is an interesting part of the application; more explanation should be provided—more on its input data and calculations.
Flow rate time series processing
· Automatic processing is applied in this part, but is there a possibility to modify the method or parameters for a more experienced user? Please deliberate on the customization capabilities of the tool
Optimal number and location of pressure sensors
· Optimization is typically a computationally taxing process; how is the performance of the tool in terms of time and accuracy?
· Depending on the algorithm and parameter settings of the optimization, the result can significantly vary; if this is an automated process, how does the tool ensure that it is not trapped in a local optimum and deliver adequate result?
Identification of critical areas for pipe burst location
· Page 6, line 210; please list the available algorithm that can be used
· The procedure is only repeated for all possible locations of burst; the time of the burst should also be considered since the demand pattern contributes to detection.
Discussion
· The tools are “tailored specifically” for the five participating utilities. Some details of the utilities (demand, size, topology) should be described to know the ranges of network it is applied to.
· Page 8, line 281; mentioned here is that “server may not be capable of handling the expected traffic”, is the tool only available online? Are there plans to deploy the tools locally?
General
· As the tools are important for managing WDS, and if it is intended for wide use, some highlights for its practicality should be added.
Author Response
Please find our responses written in blue and highlighted in yellow.
In this manuscript, the author presents several computational tools to aid the management of water distribution systems, focusing on digital transformation. The content of the paper is clear and concise. Some additional details should be addressed as per the comments below.
Response: The authors would like to thank the Reviewer for the time taken to review our manuscript and for the valuable comments and suggestions. The manuscript has been thoroughly revised to attend to suggestions. Responses to Reviewers' comments are provided below point-by-point and were updated in the manuscript accordingly.
Page 1, lines 89; KPI is first mentioned here, and it is not described. Please provide the description, and remove the description in the next mentions.
Response: The description has been included in its first appearance and removed from the following.
Page 3, line 117; “Portuguese Water and Wastewater Authority” already abbreviated as “ERSAR” in the introduction.
Response: The abbreviation ERSAR was used as recommended.
Figure 1; for visual presentation, it is advisable to have nothing highlighted in the screenshot unless the highlighted part contributes to the discussion.
Response: The Figure was corrected as recommended.
The water balance is an interesting part of the application; more explanation should be provided—more on its input data and calculations.
Response: A new paragraph was included with information on the input data of both water and energy balances, as well as the functional procedure of the platform for the balance calculation.
Automatic processing is applied in this part, but is there a possibility to modify the method or parameters for a more experienced user? Please deliberate on the customization capabilities of the tool.
Response: A new paragraph was included explaining how the method’s parameters can be customized for a more experienced user ; otherwise, default values are suggested and can be used
Optimization is typically a computationally taxing process; how is the performance of the tool in terms of time and accuracy?
Response: The implemented tool uses NSGA-II as the optimization algorithm and a complete sensitivity analysis with several real case studies has been carried to ensure that the best parameters (e.g., mutation and crossover operator) for the optimal sensor location problem are used (this research is currently under a review process in an international journal). The values (attributes) of the parameters that demonstrated to lead to robust solutions for the analysed networks have been considered in the optimization algorithm and cannot be changed by the user. At the moment, the only costumable parameter is the number of generations. Future upgrades of the tool will integrate the possibility to change the NSGA-II parameter values/attributes, as well as will incorporte other alternative heuristic optimization algorithms.
Depending on the algorithm and parameter settings of the optimization, the result can significantly vary; if this is an automated process, how does the tool ensure that it is not trapped in a local optimum and deliver adequate result?
Response: Please see the previous answer.
Page 6, line 210; please list the available algorithm that can be used.
Response: Different algorithms for pipe burst location were listed as suggested.
The procedure is only repeated for all possible locations of burst; the time of the burst should also be considered since the demand pattern contributes to detection.
Response: Thank you for the valuable comment. We have included the information that the procedure should also consider that the burst may start at distinct periods of the day (since the demand patterns may vary considerably and which may affect the performance of the burst location method).
The tools are “tailored specifically” for the five participating utilities. Some details of the utilities (demand, size, topology) should be described to know the ranges of network it is applied to.
Response: A new paragraph was included in the introduction section with overall details of the participating water utilities. A reference to the complete characterization of the water utilities was also included.
Page 8, line 281; mentioned here is that “server may not be capable of handling the expected traffic”, is the tool only available online? Are there plans to deploy the tools locally?
Response: The several tools were developed as web-based applications on cloud services since this allows universal availability using web-connected devices, and no further requirement than a web browser is needed to use the tools. This is critical in water utility computers which, due to security concerns, may not allow the use of third party software. Thus, there are no plans of deploying these tools locally.
As the tools are important for managing WDS, and if it is intended for wide use, some highlights for its practicality should be added.
Response: At the end of each section, there is a dedicated paragraph concerning the practicality of that specific tool. Also, an adicional paragraph with major highlights is included in the conclusion section.
Author Response File: Author Response.docx
Reviewer 3 Report
The article is really interesting and well written. My only concern is that it does not present any scientific advance, since the techniques inside the tools have been previously presented in other articles.
The organisation of the paper could be improved by reducing and reorganising the number of sections (now, there are 9 sections). For example, you can join some of them in the form of subsections 2.2, 2.3, etc. Furthermore, an initial flowchart or scheme of the steps and/ or blocks of the two computational tools would be useful to have a complete vision and to better understand the paper.
In addition, I noticed a minor grammatical error in line 65, where the word “used” is repeated.
Finally, I have some curiosities:
- As you have used the tools for several companies, did you impose the data format to the companies, or did you have to "clean" and "organise" the data after uploading it to the web/desktop tool? From my experience, and as you mentioned in the paper this is a great problem in this industry, since each company has different data and with different formats.
- Do you have previously published the approach to support pipe rehabilitation?
Author Response
Please find the responses written in blue and highlighted in yellow.
The article is really interesting and well written. My only concern is that it does not present any scientific advance, since the techniques inside the tools have been previously presented in other articles.
Response: The authors would like to thank the Reviewer for the time taken to review our manuscript and for the valuable comments and suggestions. The paper not only presents the set of computational tools but also offers a discussion on how water utilities can use computational tools developed under R&D projects. Additionally, it presents an envisioned journey towards digital transformation in water utilities. The manuscript has been thoroughly revised to attend to suggestions. Responses to the Reviewers' comments are provided below point-by-point and were updated in the manuscript accordingly.
The organisation of the paper could be improved by reducing and reorganising the number of sections (now, there are 9 sections). For example, you can join some of them in the form of subsections 2.2, 2.3, etc.
Response: The developed computational tools are independent from one another and, thus, the organization of the paper is set as a chapter for each computational tool. As suggested, sections 7 and 8 were merged under a common Discussion section, with a subsection for each topic.
Furthermore, an initial flowchart or scheme of the steps and/ or blocks of the two computational tools would be useful to have a complete vision and to better understand the paper.
Response: As suggested, a Table was included in the introduction section listing the computational tools that were developed under each R&D project.
In addition, I noticed a minor grammatical error in line 65, where the word “used” is repeated.
Response: The error was corrected.
As you have used the tools for several companies, did you impose the data format to the companies, or did you have to "clean" and "organise" the data after uploading it to the web/desktop tool? From my experience, and as you mentioned in the paper this is a great problem in this industry, since each company has different data and with different formats.
Response: For each tool, the main rationale was to identify the required type of data to perform the required analysis. Then, it was assessed in each water utility how these required data are structured. Finally, the software data structure was established following the most common data structure within the participating utilities (3 utilities in 5). In this way, these water utilities can use the developed tools with minor changes to their own data structures, whereas other utilities may have to organise data structures to comply with computational tool data requirements. In short, the correct data structure must be ensured by the user, and utilities might have to clean and organize data accordingly.
Do you have previously published the approach to support pipe rehabilitation?
Response: The research paper detailing the approach to support pipe rehabilitation is currently under development and will be submitted to a peer-reviewed journal in the first trimester of 2023.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The comments from the reviewer are addressed well; thus the manuscript is ready for publication.
Reviewer 3 Report
Thank you for your answers to all my questions and suggestions.
