Fault Type Diagnosis of the WWTP Dissolved Oxygen Sensor Based on Fisher Discriminant Analysis and Assessment of Associated Environmental and Economic Impact
Round 1
Reviewer 1 Report
The research problem was formulated correctly, and the researcher's methodology was also precisely presented. I have no comments on the scientific part.
The presented material corresponds to the profile of the Journal "Applied Sciences". The scientific value of the submitted material qualifies the article for publication in this Journal.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper proposed a fault type diagnosis method of the WWTP Dissolved Oxygen sensor based on Fisher Discriminant Analysis and assessment of associated environmental and economic impact. The effectiveness of the proposed approach has been validated by different experiments. Some main issues are listed as follows.
1. In the introduction, most of the references are too old to reflect the research status. It is better to get through the current literature in recent three years.
2. Some works have been reviewed. But the deficiencies of the existing techniques are not summarized. It is better to refine the pain points of current research, and highlight the motivations and contributions of the paper.
3. Some defects of DO sensors, such as bias, drift, wrong gain, and loss of accuracy have been simulated. Do those faults follow the real one in the real industrial application? In addition, sometimes, some faults samples are not always available, and how to deal with such challenges.
4. Since different features have different scales. Do the authors implement any preprocessing procedures before using FDA model?
5. There FDA model is adopted. But there is lack of sufficient comparisons. It is better to make a comparison with existing techniques, such as PCA.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
Figure 1 is bit misleading. Authors need to reconsider it in order to provide clear representation/reflection of the work in figure.
A few legends of the Graphs are very blur, can not be read, please make it bold. The insert images are also very blur. Improve the quality of graph images.
Captions are very small, captions must be detailed and should have complete information of the image. There is no mention of insert images in captions. The authors need to spend more time in improving the quality of images.
In my opinion, it would be better to review abstract and conclusions and they must represent the overview of the work concisely.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
The study was conducted to investigate new diagnostic solutions for specific failures caused by faulty DO sensors and to assess the environmental and cost impact of the failures. This study is interesting, but there are some problems.
（1）The introduction part is not complete enough about the research status of fault diagnosis. There are many latest studies, and it is suggested to supplement relevant contents.
（2）Please provide the calculation basis or relevant references of the data in Table 1.
（3） Many parameters have not given specific basis or references. For example, lines 269 to 280.
（4）Please check the spelling of the word on line 281.
（5）The amount of data in the training set is too small. It is suggested that the author use cross validation to verify the method.
（6）The explanation of the fault diagnosis result diagram is not clear enough. The author is requested to supplement relevant contents to make readers better understand.
（7）The author is asked to consider how to diagnose if two or more faults occur at the same time.
（8）Please provide the calculation basis or relevant references of the data in Table 4.
（9）The fourth part of the paper is missing, please revise it.
（10）The accuracy and timeliness of fault diagnosis are not discussed in this study. Please compare with other fault diagnosis methods to prove the superiority of this study.
To sum up, the reviewer considers that this study is not ready for publication, and the author is requested to revise it carefully.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
This is revised version. The authors have made large improvement. Some minor concerns are left.
1. Some key findings and specific diagnostic accuracy can be added in the abstract.
2. The limitations of the proposed method can be summarized and the future work can be added in the conclusion.
3. Although the authors have made some discussions on the compared method PCA. But the reviewer failed to find the main introduction. The results of PCA implemented on the oxygen sensor fault dataset are missing.
Author Response
Response to Reviewer 2 Comments
Manuscript entitled "Fault type diagnosis of the WWTP Dissolved Oxygen sensor based on Fisher Discriminant Analysis and assessment of associated environmental and economic impact "
The authors would like to thank the Reviewer 2 for the new comments and suggestions on the manuscript. We believe that changes brought to the Revised 2 form of the manuscript are in the benefit of the article quality.
The comments have been addressed and the itemized list of responses and actions taken to address these comments are presented in the following (Reviewer2’s comments are italicized).
Note: Changes made to the Revised 1 form of the manuscript, according to the comments and suggestions of the Reviewer2, are presented as track changes in the revised manuscript and added text was highlighted with yellow colour.
This is revised version. The authors have made large improvement. Some minor concerns are left.
Recommendation:
- Some key findings and specific diagnostic accuracy can be added in the abstract.
Response:
Thank you for the recommendation. As a result, the abstract was completed with recommended elements. The new form of Abstract is the following:
“Sensor failures are common events in wastewater treatment plant (WWTP) operation, resulting in ineffective monitoring and inappropriate plant management. Efficient aeration control is typically achieved by the dissolved oxygen (DO) control and its associated sensor becomes critical to the whole WWTP reliable and economic operation. This study presents the Fisher Discriminant Analysis (FDA), used for fault diagnosis of the DO sensor of a currently operating municipal WWTP. Identification of the bias, drift, wrong gain, loss of accuracy, fixed value, complete failure minimum and maximum types of DO sensor fault were investigated. The FDA proposed methodology proved efficiency and promptitude in obtaining the diagnosis decision. The consolidated fault identification showed an accuracy of 87.5% correct identification of the 7 faulty and normal considered classes. Depending on the fault type, the results of the diagnosing time varied from 2.5 h to 16.5 h during the very first day of the fault appearance and only based on observation data not included in the training data set. The latter aspect reveals the potential of the methodology to learn from incomplete data describing the faults. The rank of the fault type detection promptitude was: bias, fixed value, complete failure minimum, complete failure maximum, drift, wrong gain and loss of accuracy. Greenhouse gases (GHGs) such as nitrous oxide (N2O) and carbon dioxide (CO2) emitted during wastewater treatment, electrical energy quantity in association to costs spent in the WWTP water line and clean water effluent quality were ranked and assessed for the normal operation and for each of the DO sensor faulty regimes. Both for CO2 and N2O, the on-site emissions showed the most significant GHGs contribution, accounting for about three quarters of the total emissions. The complete failure maximum, fixed value and loss of accuracy were the DO sensor faults with the highest detrimental impact on GHGs released emissions. The environmental and economic study reveal the incentives of the proposed DO sensor faults identification for the WWTP efficient and environmentally friendly operation.”
Recommendation:
- The limitations of the proposed method can be summarized and the future work can be added in the conclusion.
Response:
Thank you for the recommendation. The following sentences were added in the Conclusions section:
“Complexity of the diagnosis increases when the number of fault types subject to investigation expands or the sensor faults produce comparable effects on the process variables.”
“The identification of simultaneous acting fault types needs specific investigations and this is the subject of future research work.”
Recommendation:
- Although the authors have made some discussions on the compared method PCA. But the reviewer failed to find the main introduction. The results of PCA implemented on the oxygen sensor fault dataset are missing.
Response:
Thank you for the comments and observations.
We considered in our study the FDA diagnosing methodology because of its sound mathematical approach and the scarcity of reported applications for the wastewater treatment plant working in automatic operation mode, especially when a larger number of fault types were taken into consideration. Consequently, we set our target on studying this specific fault identification methodology and the detailed investigation of other diagnosing methods we considered beyond our original scope. This is the reason we did not use the PCA based fault identification methods in our study. However, we conducted bibliographic research to evaluate the benefits of the FDA methodology applied to WWTP, when it is compared to other methods. The penultimate paragraph of section “3.2. Fault diagnosis” presents a set of comparative considerations for the results of our study against reported results in studies using PCA based methods.
Following the comment of the reviewer, the 3.2. paragraph was completed with the following sentences, in order to highlight the contributions of out study. They are:
“Additionally, it did not study in detail the fault identification time. The present study investigated several fault types and the identification time was determined for each of them.”
“The sub-period division strategies combined with multiway principal component analysis for only two faults diagnosis showed that the used methods can manifest false identification results during normal operation period and inability to detect the fault during some time intervals [35]. The present investigations considered six different fault types, that implied increased difficulty in the promptitude and accuracy of the fault diagnosis.”
