Towards Semi-Automatic Generation of a Steady State Digital Twin of a Brownfield Process Plant

Round 1

Reviewer 1 Report

In the introduction and literature review of the article, the authors presented a very wide study of literature which is connected to the main subject of the paper.

In paper should be much more highlighted the difference between the available in literature methodology and presented by authors. During the preparation of the pre-review, It was hard to conclude if the presented methodology gives a better result. Also, it is hard to find the science part of the article, which is obligatory for a Journal of Applied Science.

During the pre-review process was also noted below pointed signification disadvantages:
- The quality of figure 1 is low, especially in the section "System and documents at a brownfield site" (P&ID symbols are unreadable);
- Literature reference are incorrect noted in case of the notion of 2 or more position (for example line 189: should be [12, 54, 66, 67]
- line 328: is hard to conclude that E-100 is in figure 12, 2 pages after
- Faults in references - double number o each position;

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper shows a model transformation from a graph (steps 1 and 2) to a specific simulator (Balas) (steps 3-7) Steps 1 and 2 are not the subject of the paper as are available as previous research.

The described work in this paper, basically 3 and 4, are quite simple. A match from the nodes and edges of the graph to a flowsheet. Some rules and algorithms are described using UML, these could have been described using any flowchart, no explanation on how to execute the UML models to derive the final model from the graph. After the matching a "tuning" of the flowsheet is done using some rules to remove valves and changing some equipment. 

The case shown is very simple and it appears that the simulation expert that is necessary for steps 5 and ahead would just make the flowsheet directly in the simulator program in much less time that the needed for this automatic application. In chemical engineering there are many other unit operations not considered, nor explained that are not so simple (distillation columns, extractions, absorptions, complex reactors).

If new rules are needed for equipment that it is not in the library of the simulator and new rules are needed to transform the model to other simulators the utility of the application is of not very high.

The paper focus on a brownfield process plant, although the case presented is of a pilot plant whose P&I is developed digitally, which usually is  not the case in brownfield plants, besides the main difference is in steps 1 and 2. The methodology seems general enough for any process plant, why focus on saying is a browfield plant?

The section about steady state simulation is not very useful and the references are but any means too much and they confuse more than support the remarks made by the authors.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This article proposes a semi-automatic methodology for generating a digital twin of a brownfield plant. The topic is potential and interesting. I have only some minor remarks as detailed below.

• Introduction section. I suggest to add some more line at the end of the section to explain how the rest of the paper is structured.
• In general Figures are not well justified. I suggest adding some clarifying details at least in the first figures (Figure 2 and Figure 3)
• What is the main novelty of the research respect the state of art? And what is the main contribution. Please clarify it.
• I suggest to add a Results section in order to esponds to the question "What have you found?" Hence, only representative results from your research should be presented. The results should be essential for discussion.
• Add a Discussion section. Here you must respond to what the results mean. 
• Write a clear and compelling conclusion. What is the main limitation of your research? How the problem was solved?
• Check English and grammar.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

The paper addresses a topic of great interest for brownfield plants. A digital twin is a necessary tool for both optimization of the plants, but even more for upgrading and technology updates.

The authors contribute with solutions for partial automatic generation of digital twins of brownfield plants, mainly in what they define as steps 3-5.

As there is a great variety of brownfield plants with various states of documents available, I suggest the authors review the types of brownfield plants and pinpoint the categories of plants their research best applies, e.g. chemical, utilities, etc.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The paper has improved, I still think more work should be included in order for it to be used in practice, but as a proof of concept can be ok.