Pyrolysis of Mixed Plastic Waste: II. Artificial Neural Networks Prediction and Sensitivity Analysis

Round 1

Reviewer 1 Report

The manuscript “ Pyrolysis of Mixed Plastic Waste: II. Artificial Neural Networks Prediction and Sensitivity Analysis” is interesting and within the scope of the journal but some minor changes should be addressed:

1. Please insert the error references (lines 67,75, 80…).
2. Please update the state-of-the-art by discussing briefly about the plastics used for fuels obtaining in comparison with conventional fuels (please see https://doi.org/10.37358/MP.19.4.5259 and https://doi.org/10.37358/MP.19.1.5119).
3. Please try to make figures 1 and 2 more readable.
4. Please try to keep the same font and size for all figures.
5. I recommend to extend the Conclusions section.
6. Please rewrite the reference according to journal instructions.

Author Response

Dear Respected Reviewer,

Greetings

Thanks for your kind valuable comments that lead to a significant improvement of the manuscript. Please find attached the authors' response to your kind comments.

Thanks again with best regards

Author Response File: Author Response.pdf

Reviewer 2 Report

- There is an error with the refs numbers in the Introduction section.

- In section 3.1 Thermogravimetric Analysis of Mixed Polymers, authors must review and correct the writing, based on the repetitions.

- Improve the presentation and clarity of Figures 1, 2, 3 and 4;

- Table 8, use scientific notation for describing temperature values.

- As with the other statistics, Garson correlation should be shown in the Experimental section.

- Edit section 3.3 Sensitivity analysis without using subsections. Deal with Pearson and Garson in a single section. Why is it important to work with both forms of correlation? What did the authors want to show with this? Why is it important to quantify the importance of parameters by Garson, complementing Pearson's? The second result is expected by the dynamics of the equation. The discussion of these results does not follow the same quality as previous discussions.

• The authors state in section 3.3.1, Pearson correlation, that “Among the five input parameters, PS (wt 236%) and T (K) had a negative impact, while PP (wt %), HDPE (wt %), and LDPE (wt %) had 237 a positive impact on weight left %.”.
  But, most parameters returned statistical test results close to zero, if compared to the order of magnitude of the maximum impact values ​​(1 and -1); The authors only claim that there is "positive impact" and "negative impact" without discussing these results based on the degree of interference returned, which can lead the reader to superlative (or even erroneous) interpretations of what the result actually returned and disseminate this fact in future works. Authors should review the presented analysis and discussion regarding the result obtained in the statistical test and provide the reader with a more authoritative discussion.
  - Next, no less important, in the analysis of the statistical significance of the input parameters, the authors state that “the parameter was considered statistically important if the corresponding p-value was less than 0.05.”. On the other hand, according to the results obtained, only temperature has a statistically proven influence. This is a usual statistical test, safe and with practical applications in research and development for decision making in various industrial sectors. Thus, based on these assumptions, the authors should not just state that "this statistical tool did not allow quantifying the contributions of each variable" because this is widely known. The test allows for qualitative analysis and the authors suppress the information that the other variables are not significant, which was proven by the test.
  The combined analysis of the values ​​of the obtained correlation parameters (test statistics close to zero) and the p-values ​​confirm that the variables are not important for the best results obtained. Next, the quantification of each importance seems unnecessary and beyond the scope of the statistical discussion that governed the manuscript.
  Since the authors have given relevant importance to these statistical tests for the conclusion that the alternative method is reliable and safe for its intended use, it is important that the authors review and clarify these aspects to the reader and to the discussion of the results for method validation.

Author Response

Dear Respected Reviewer,

Greetings

Thanks for your kind valuable comments that lead to a significant improvement of the manuscript. Please find attached the authors' response to your kind comments.

Thanks again with best regards

Author Response File: Author Response.pdf

Reviewer 3 Report

The paper has some original results, and is well organized. The introduction mut be improved and some figures need to be remaid to be acceptable.

Author Response

Dear Respected Reviewer,

Greetings

Thanks for your kind valuable comments that lead to a significant improvement of the manuscript. Please find attached the authors' response to your kind comments.

Thanks again with best regards

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

1. The methods used clearly showed responses that the authors insist on discussing erroneously and/or superficially to minimize the effect of the results.
The issue here is not about merit for the work. The work is well written and may appear to the reader as a scientific work with correct descriptions of facts. What does not happen in the main points of discussion that govern the purpose of the proposal.
It's a question of what the results actually prove.

2. The results prove that among the tested variables , only temperature is significant for the process. Why don't the authors clarify this question to the reader? What is the problem with exposing the conclusion that of all the variables tested as a hypothesis, only temperature returned significant results? Just as it is highlighted that temperature has a great influence on the response, based on the test statistic, it should be said that the others have no influence, based on the same test parameter.

3. In this sense, the authors corroborate this problem and only minimize this discussion, as in:
"Based on the SAI value, the relationship between an input parameter and the output parameter can be classified as follows: negiligible (0 ≤ SAI ≤ ±0.2), weak (±0.21 ≤ SAI ≤ ±0.35), moderate (±0.36 ≤ SAI ≤ ±0.67), strong (±0.68 ≤ SAI ≤ ±0.90), and very strong (±0.91 ≤ SAI ≤ ±1.00) [41].
In this excerpt, the authors present the range of SAI values ​​which allows us to state that the importance of a parameter is negligible, but at no point do they say that, based on these values, it is concluded that the variables are not significant to the process .
The data in table8 returns that all parameters are negiligible (only temperature has significance)!!! In addition, the authors still do not cite this discussion.
Following Table 8, shortly thereafter, the authors simply discard this discussion and start another discussion and try to show the matter as closed. Nothing was discussed!!! This cannot happen.

4. In another excerpt, the authors write:
"The p-values ​​provides a qualitative idea about the most significant parameters. A significance level of 5% (ie, α = 0.05) was considered for the analysis. than 0.05 [42-45]. The results of the analysis are summarized in Table 9".
And they discuss this: "Results imply that temperature has the dominant impact on the pyrolysis process of the small sample of mixed plastic." Right! But by no means do the authors talk about the other variables! Probably because, from the results of Table 9 returned by the statistical test, it is clear and visible that all the other variables highlighted for the study were not significant, proving that only the temperature interferes in the process.

Thus, in my opinion, given the data returned from the exploration of the authors, these results could be used for future decisions in production processes of the authors, but the data presented and their discussions are not configured for purposes of scientific study (as they were discussed and shown to the reader).

Given the above, I do not indicate the publication of the manuscript since the authors continued showing only what they wanted to show from the initial hypothesis, without clarifying that practically all the variables informed are not significant to the study.

Authors should rethink their writing in order to clarify these critical points for the reader and the community in the field.

Author Response

Dear Respected Reviewer,

Greetings & thanks for your kind feedback

Please find attached the authors' response to your comments.

Thanks

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

The authors expressed the perspectives and limitations of the work. Scientific writing was present. Some punctual writing errors are still present. Please proofread the text for publication.