A Fuzzy Model to Manage Water in Polymer Electrolyte Membrane Fuel Cells
Round 1
Reviewer 1 Report
The authors present a fuzzy model to detect water content in PEMFC. The reviewer recommends the acceptance of the paper for publication in Processes following the revision of the points detailed below.
- Can author correlate the model with real fuel cell polarization data? Otherwise, it is hard to quantitatively analyze the fuel cell performance with this fuzzy model. In other words, can author provide concept-proof that by using this fuzzy method, researchers can regulate the water transport better?
- The author expressed their results in “flood”, “normal”, and “dry” states. What is the relative humidity for these three modes? Can author provide more data points from 0% RH to 100%RH?
- For EIS, the author needs to provide the unit for Fig 6. Can author scan the sample with a longer frequency rang? The EIS data looks incomplete. What is the equivalent circuit to extract the data? Why flood PEMFC looks similar to normal PEMFC?
Author Response
Please see attachments
Author Response File: 
Author Response.docx
Reviewer 2 Report
The article presents issues related to Proton Exchange Membrane (PEM) fuel cell, and in detail, the water management issue. The research carried out by the authors is part of a global trend of attempting to improve fuel cell technology to enable its introduction on an industrial scale and to replace other energy sources operating on the basis of consumption of fossil fuels. The developed fuzzy-logic-based model is intended for determining the hydration degree of the cell's membrane which is one of the key parameters for proper operation of the cell. The amount of the water inside the membrane is intended to be assessed by measuring and analyzing two parameters i.e., voltage slope changes during the step changes of the load, and changes of voltage amplitude while forcing the current oscillations.
The organization of the article is adequate. Extensive introduction and "state of the art" chapter, with numerous references to literature related to the topic of the article, allow to learn about other approaches to the issue considered in the article and present the solutions proposed by the authors. The next parts are the description of the experiment setup, methodology and proposed fuzzy model for determining the membrane hydration. The next section contains the experiment results and the validation of the obtained results performed with the use of electrochemical impedance spectroscopy. Two final chapters present the discussion of research results and directions of future work.
The content presented in the article is suitable for publication in the special issue of Processes, i.e. "Computation and numerical modeling of fuel cell". However, despite good organization of the paper there are some parts that need to be supplemented with additional information.
The first one and the most important, is related with the membrane hydration states. It is advised to put somewhere in the paper the definitions of DRY, NORMAL and FLOOD states of the membrane along with some measurable factor or factors which are used to assess in which state is the membrane.
Another information that should be included in the article is the method that was used to assess the hydration level of the cell membrane and also to regulate it. In the article there are only fuzzy values describing the humidity of the membrane, but nowhere in the paper I can find what crisp quantity was used to create these fuzzy values and how the membrane humidity was measured during the operation of the fuel cell. Moreover, in the paper it was stated that the DRY and FLOOD states were forced without giving any information on how it was performed.
Another important aspect that should be expanded is the description of the fuzzy model. In order to fully understand it and the possibility of reperforming the experiment, it is necessary to include complete information regarding the membership functions belonging to individual linguistic variables and their values. Moreover the “Fuzzy Decision Tree” that was used for the research require some wider description. Additionally, the proposed heuristic rules (which typically are presented in form: IF smth THEN smth else) described by equations (8) - (11) seem to be incomplete and also need to be supplemented.
Finally in the results and discussion chapter, it is not clearly presented how the fuzzy model was used to determine the hydration state of the fuel cell membrane. As I understand that chapter, it presents the measurements of the voltage oscillations and the slope changes in three different operating conditions of the membrane (flood, dry, normal) but does not show how the fuzzy model assess the hydration state of the membrane using the information about slope change and voltage oscillation.
Detailed remarks to the paper are listed below:
Table 1 - it is not easy to follow the data in the table - it requires to be rearranged.
L256: What are the standard operation conditions for PEM fuel cell?
L258: How the water content in the membrane was regulated/measured?
L264-266: Why the other components of the ohmic losses of the cell (i.e., resistance of the GDL, bipolar plates, resistance of the connectors and other) was neglected?
L267: Current oscillation and the unit mV does not correspond each other.
L248, 396 - using battery instead cell is inadequate in terms of fuel cell.
L217: instead of diffuse rather fuzzy(?)
Author Response
Please see attachments
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The paper can be accepted if the author address the following comment:
1. The equivalent circuit used to extract the data from Nyquist plot is hidden by Figure 8. Please separate these two figures.
Author Response
The graph is shaken
Author Response File: Author Response.docx
Reviewer 2 Report
L74 (and in whole paper): I believe it should bye fuzzy not fussy.
Figures 8 covers figure 9
Author Response
I correct fussy by fuzzy and move figure 8
Author Response File: Author Response.docx
