A Fuzzy Inference System for Detection of Positive Displacement Motor (PDM) Stalls during Coiled Tubing Operations

Round 1

Reviewer 1 Report

In this work, a Fuzzy Logic Inference System (FIS) with 3 inputs and 2 outputs is proposed which will monitor the surface parameters of a coiled tubing operation and predict the PDM stall, giving enough time for pumping to be halted before the PDM completely stalls. After reviewing carefully, the reviewer found that the paper is written very poorly. As a result, it cannot be accepted for publication in the current state. A comprehensive major revision is required. The comments are as follows:

1. The abstract is poorly written. The abstract needs to be written by mentioning the advantages of the proposed method and numerical results at the end. The abstract needs style and form corrections since the ideas sound lost/disconnected. Please try to smooth the connections in between and organize them logically, to make them sound as a solid single paragraph.  

2. The quality of the figure 1, 2, 4, 5, 6, 7, and 8 needs to be improved.

3. Introduction is poorly written. Rewrite the introduction, covering background, problem statement, a literature review of current efforts, motivation, aims, contribution, and paper organization.

4. Cleary state the gap and novelty of the research in the introduction along with advantage of the proposed research work.

5. Add paper organization at the end of introduction.

6. Figure 5 is written twice for two figures in page 8 and 9.

7. How come Figure 1 is written in page 11?

8. In Figure 9, 10, 11, and 12, axis titles are missing.

9. There is no literature review is found.

10. A comparative study is required to present in the discussion section to prove the effectiveness of the proposed method compared to the existing approaches which are proposed from 2019-2022.

11. Conclusion should be rewritten by explaining the proposed algorithms performance for different cases along with future research or further extension of the current work.

12. The references are poorly cited and written in the paper.

13. Only few references are found from recent years. However, most of the references are older than 2016.

14. Add discussion section separately by mentioning the significance of the results. 

15. Avoid use of we/our/us.

16. There are lots of grammatical, lingual and punctuation error has been observed throughout the paper. The paper required extensive proofreading.

Author Response

1. The abstract is poorly written. The abstract needs to be written by mentioning the advantages of the proposed method and numerical results at the end. The abstract needs style and form corrections since the ideas sound lost/disconnected. Please try to smooth the connections in between and organize them logically, to make them sound as a solid single paragraph.  
- Abstract has been changed as requested.

2. The quality of the figure 1, 2, 4, 5, 6, 7, and 8 needs to be improved.
- All figures have been fixed.

3. Introduction is poorly written. Rewrite the introduction, covering background, problem statement, a literature review of current efforts, motivation, aims, contribution, and paper organization.
- Introduction has been expanded, a section about the currently literature has been added.

4. Cleary state the gap and novelty of the research in the introduction along with advantage of the proposed research work.

5. Add paper organization at the end of introduction.
- Added as requested.

6. Figure 5 is written twice for two figures in page 8 and 9.
- Corrected

7. How come Figure 1 is written in page 11?
- I did not find this issue, maybe it has already been fixed by the review team.

8. In Figure 9, 10, 11, and 12, axis titles are missing.
- All 3 axis titles have been added. Text font sized increase for better readability.

9. There is no literature review is found.
- Added as related works

10. A comparative study is required to present in the discussion section to prove the effectiveness of the proposed method compared to the existing approaches which are proposed from 2019-2022.
- there is no similar approach in Scopus or SPE. The similar works are described in the Related Works section. They are about the use of AI to detect fails or misoperation of PDMs in the drilling scenario. Even so, none of them address the same problem as our research, which is detect the PDM stall using only surface data.

11. Conclusion should be rewritten by explaining the proposed algorithms performance for different cases along with future research or further extension of the current work.

12. The references are poorly cited and written in the paper.

13. Only few references are found from recent years. However, most of the references are older than 2016.
- Newer references used to expand the introduction section.

14. Add discussion section separately by mentioning the significance of the results. 
- Discussion section added at the end of the results, as after all the results for each dataset a small discussion already takes place.

15. Avoid use of we/our/us.
    - A word search for we/our/us was perfomed and text was corrected.

16. There are lots of grammatical, lingual and punctuation error has been observed throughout the paper. The paper required extensive proofreading.
- Paper was revised again for errors.

Reviewer 2 Report

Although the subject of the article is interesting, it has shortcomings in terms of contribution to the literature. I have made some suggestions to increase the readability of the article and raise its level. These recommendations should be met with care.

1) The success of the fuzzy inference system should be explained with numerical information in Abstract Section.

2) Figure 1 and Figure 2  have very poor resolution.

3) What makes this work innovative? It should be given in the Introduction Section.

4) Membership functions other than Trapezoidal and Triangular should also be examined.

5) Discussion section should be added.

6) The conclusion section is very short. It should be expanded.

7) The literature of the study is insufficient. It should be supported by current literature.

Author Response

1) The success of the fuzzy inference system should be explained with numerical information in Abstract Section.

- Added system accuracy of 94%

2) Figure 1 and Figure 2 have very poor resolution.

 - All figures have been fixed. Figure 1 was redone.

3) What makes this work innovative? It should be given in the Introduction Section.

- Better explained in the Introduction.

4) Membership functions other than Trapezoidal and Triangular should also be examined.

- I had a conversation with my professor about what the literature discuss regarding which MF should be the best. Some of the papers describe the triangular being fairly accurate and also having a lesser computational demand to be computed. We decided to kept the study only using triangular MFs for now. When we get more data and expand the study, we can compare different MFs. As this is an inference system and the output is not directly connected to any actuator as of now, it is not critical that the output has an exact value, as long as it is above the defined threshold for 1 datapoint (1 second) the alarm will trigger and the system will succeed.

5) Discussion section should be added.

- Added

6) The conclusion section is very short. It should be expanded.

- Conclusion expanded as requested.

7) The literature of the study is insufficient. It should be supported by current literature.

- Current literature added, although papers describe the use of AI to assess PDMs, there is no specific paper that has an algorithm to detect stalls and we have nothing at the moment to compare our study to.

Round 2

Reviewer 1 Report

The authors addressed al the issues. 

Reviewer 2 Report

All suggestions were taken into account by the authors and updates were made. I think it would be appropriate to accept.