Numerical Analysis of Gas Hold-Up of Two-Phase Ebullated Bed Reactor

Round 1

Reviewer 1 Report

The manuscript con be publish ni Ehem Engineering with minor revision.

Comment

1. The topic talk about numerical analysis, but the author do not describe the numerical method used in this manuscript. Moreover, the equations described the flow phenomenar should be added in the paper. Is this paper studied the numerical method?

2. What is the main response obtain and used for the machine learing? Graphical represenation the effect of each parameter studied to the response should be shown?

3.  Why the experiment used the PMC as liquid for studied?

4. What is the standardized data used in this study?

5. The equation or function used should be shown.

6. All figure need to be improved the resolution.

7. Fig 9, each feature should be labeled. Which F-value is the critical value that can be used to decise important or not important parameter?

Author Response

We thank Reviewer 1 for the time and effort in reviewing our manuscript. Thank you for recommending minor corrections.

Author Response File: Author Response.docx

Reviewer 2 Report

The overall level of the paper is acceptable, demonstrating a certain degree of innovation. However, there is room for further improvement regarding image clarity and aesthetics used in the article. Moreover, there are specific issues with language expression. The number of cited references falls short, and previous studies have not been sufficiently investigated in this paper. It is advisable to provide additional supplementation in these areas.

There are a few grammar errors in the article, and it is recommended to make improvements

Author Response

We thank reviewer 2 for the time and effort taken to review our manuscript. We have further made improvement on the revised manuscript as indicated in yellow highlights.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript titled "Numerical Analysis of Gas Hold-Up in a Two-Phase Ebulliated Bed Reactor" offers a valuable contribution to the field of chemical engineering and industrial processes. By applying machine learning algorithms, specifically Support Vector Machine (SVM) and Gaussian Process Regression (GPR), to predict gas hold-up in ebulliated bed reactors, the authors address a significant challenge in heavy feedstock processing and hydrocracking. These findings have practical implications for optimizing reactor performance in industrial settings, potentially leading to increased efficiency and cost savings. However, the following comments should be addressed before considering this manuscript for publication.

1. The title of the manuscript has a typo. It is not ebulliated but "ebullated bed reactor." The authors should consider correcting this typo.
2. If possible, quantify the statement that "liquid velocity has the most impact on predicted gas holdup" by providing some measure of this impact.
3. The authors in the introduction section briefly mentioned some experimental studies related to EBRs but didn't delve into their findings or their relevance to the current study. The authors should consider summarizing key findings and explaining how they relate to the present research.
4. The authors should emphasize why gas holdup is a critical parameter to study in EBRs and how it directly impacts their performance and efficiency.
5. The authors should clearly state the research gap that this study aims to address. Explain why studying the inter-relationship between operating parameters and gas holdup is important. Provide a clear statement of the research objectives.
6. The authors in the introduction mentioned using machine learning algorithms (support vector machine regressions and Gaussian process regressions) in this study. The authors should briefly explain why machine learning is suitable for this problem and what it adds to the existing research.
7.  The first paragraph in the experimental setup begins by discussing the behavior of bubbles, but it is essential to connect this information directly to the experimental setup and objectives. It is suggested to delete the first paragraph of the experimental part.
8. Figure 1 needs to be clarified, and it is hard to read the parts of the experimental setup. Therefore, the authors should consider improving the resolution of this figure.
9. The authors should explain the significance of the recycle ratios (1, 1.5, and 2) and how they were selected. Provide reasoning for why these specific values were chosen.
10. The authors mentioned the essential use of kernel functions, but they could consider elaborating on the role of kernel functions in SVM regression. The authors could explain how different kernels can capture different types of relationships in the data, as this is crucial for readers to understand the model's flexibility.
11. In the section on Model Configuration, the authors briefly mentioned configuring SVM and GPR with different kernel functions. It would be beneficial to explain why these specific kernel functions were chosen and how they relate to the research objectives. 
12. The authors should consider explaining the concept of box constraints and their role in preventing overfitting. Providing a brief rationale for setting them to automatic for all models can help readers appreciate their significance.
13. The interpretation of SVM model findings, particularly the observation that the choice of kernel function significantly influenced model performance, is insightful. However, providing some insights into why the quadratic kernel function performed the best could be beneficial. 
14. The authors should correct the sequence of figures. For example, on page 14 and row #311, the figure number should be 9, not 8.
15. The provided manuscript has some typographical errors or inconsistencies, which should be addressed.

The quality of the English language in the manuscript seems to be generally good. However, there are a few minor suggestions to enhance the overall readability and clarity.

Author Response

We thank Reviewer 3 for the efforts and time expended on manuscript to improve its quality. We have addressed the comment raised as indicated in yellow highlights throughout the manuscript.

Author Response File: Author Response.docx

Reviewer 4 Report

Attached please find my comments.

Comments for author File: Comments.pdf

There are a few minor grammatical errors. 

Author Response

We thank Reviewer 4 for the efforts and time expended on manuscript to improve its quality. We have addressed the comment raised as indicated in yellow highlights throughout the manuscript.

Author Response File: Author Response.docx

Reviewer 5 Report

The paper presents a machine learning based modeling approach to predict gas hold-up in ebullated bed reactors, which is valuable given the importance of these reactors in industry. The objectives are clearly outlined, the methodology covers model development and evaluation metrics, and the results are presented through tables and plots. However, there are some areas that need improvement:

- The introduction, while providing basic background, could be expanded with more context and motivation from literature. Ensure all cited references are directly relevant.

- The model development and training process needs more elaboration with technical details on procedures, parameters, and techniques used.

- Though results are organized well, include some supporting discussion and sample predictions to aid interpretation.

- Conclusions are properly supported by the data. But relate them better to practical implications.

Overall, this is a good concept but the authors need to enhance the literature background, provide more comprehensive explanations of the methodology, supplement the results with discussion, and improve the quality of the language and presentation. Addressing these aspects will significantly strengthen the paper and make it suitable for publication after revisions. I hope these recommendations are useful for the authors moving forward.

- The paper requires proofreading and editing to address issues with grammar, word usage, sentence structures, tense consistency, and style to improve clarity and flow.

Author Response

Thank you for the comment. The introduction of the revised manuscript has been improved further.

Author Response File: Author Response.docx

Round 2

Reviewer 4 Report

Authors must clearly indicate where they addressed comments in the revised manuscript.

The manuscript still lacks a clear statement of novelty. Why this research is important. What is their contribution to this field, and how they achieved their objectives including experimental and numerical methodologies.

There is a consistent typo in the manuscript with ' Ebullated reactors' that needs fixing before publication.

Page 3, lines 89 to 96 are unclear. It is not easy to understand what the authors are trying to convey.

On page 7, after ' some studies have shown that...' they must include a few citations.

Overall, the manuscript is well prepared, and no additional comments on the results section. 

The quality of English should be improved especially in the portions added in the revised manuscript. The use of tense in the manuscript is not consistent. 

Author Response

We thank the Editor and the reviewer for time and efforts to further improve the quality of our manuscript. The manuscript has been revised based on the comments raised.

Author Response File: Author Response.docx