Review Reports - Modeling Dark Fermentation of Coffee Mucilage Wastes for Hydrogen Production: Artificial Neural Network Model vs. Fuzzy Logic Model

Round 1

Reviewer 1 Report

The manuscript used two simulation models to predict the H2 production from the bio-conversion of coffee mucilage waste. Many practical parameters are considered in the modelling processes. The modelling results from ANN and FLB approaches were compared and validated as well. The two models were presented in a logic manner. This research is helpful and interesting for the field of bio-hydrogen production. It is suggested to be published after addressing some minor issues. Specific comments are as below:

Some details about the experimental design with MiniTab should be added.
Authors should give some explanations/understandings about the reason why FLB model shows better fitting results than that of ANN model.
Line293, to achieve better fitting results, the bi-lateral parameters are excluded. Specifically, what parameters are removed?

Author Response

March 27, 2020

General comments: We have addressed the specific comments on a point-by-point basis and modified the manuscript appropriately, as referred to below. We thank the reviewers for the helpful feedback and assure the reviewers that we take all manuscript revision, particularly for Energies, quite seriously. The original submission has been revised so that it concisely presents our significant research of hydrogen production from coffee mucilage in dark fermentation with organic wastes. We believe that data to support this finding are more clearly presented, thanks to careful reviews and comments of the reviewers, which we have incorporated into the revised manuscript.

Responses to Reviewers' comments:

Reviewer #1:

The manuscript used two simulation models to predict the H₂ production from the bio-conversion of coffee mucilage waste. Many practical parameters are considered in the modelling processes. The modelling results from ANN and FLB approaches were compared and validated as well. The two models were presented in a logic manner. This research is helpful and interesting for the field of bio-hydrogen production. It is suggested to be published after addressing some minor issues. Specific comments are as below:

1) Some details about the experimental design with MiniTab should be added.

Reply: Thanks to the reviewer comment. We have added a brief experimental design with MiniTab in lines 136-139 on page 3 as follows, “Briefly, a two-level factorial experimental test was designed with three different ratios (w/w) of coffee mucilage and organic wastes mixture (8:2, 5:5, and 2:8) were prepared, and additional control runs with only coffee mucilage or organic wastes were added [13]”

2) Authors should give some explanations/understandings about the reason why FLB model shows better fitting results than that of ANN model.

Reply: The reviewer is correct that we have not provide the exact reasons why the fuzzy-logic-model was better than those results from the ANN model. While we addressed possible rational and description in lines 306-310 on page 9 with the related references and Table that clarify the manuscript. Furthermore, the approaches of ANN and fuzzy-logic-model were described in lines 90-106 on pages 2 and 3. We believe that these explanations have clarified our research study in support of our experimental approach.

3) Line293, to achieve better fitting results, the bi-lateral parameters are excluded. Specifically, what parameters are removed?

Reply: Thanks to the reviewer for the constructive comments. Following the reviewer’s suggestion, we have added some explanations accordingly in lines 294-295 on page 9. We believe these explanations have clarified our research study in support of our technical and conceptual novelty.

Author Response File: Author Response.pdf

Reviewer 2 Report

The purpose of the manuscript was the prediction of hydrogen production and yield in the batch system varying parameters of mucilage/organic waste ratio, pH, acidification time, chemical oxygen demand, and temperature. The authors tested two-level factorial experiments, and modeled their hydrogen profiles with the ANN and fuzzy-logic-based approaches. The article is an interesting piece of work and may be interesting to the broader scientific community. The manuscript does not bring scientifing novelty. There are literature data for anaerobic fermentative hydrogen production from coffee mucilage, but the subject is important from ecological and biotechnological point of view. The experimental analysis was well designed, the content of the state of art provides useful information about the topic. As was confirmed by the authors the results of the study show that using experimental results for bio digestion of waste substrate in batch reactor with ANN and fuzzy model was predictive. The fuzzy model had a greater predictive capacity than the ANN model with the response to the interaction of independent factors. As was mentioned by the authors using of coffee waste may be considered as an alternative source for energy production. My recommendation is to accept the article for the possible publication in “Energies” in present form.

Author Response

March 27, 2020

Responses to Reviewers' comments:

Reviewer #2:

The purpose of the manuscript was the prediction of hydrogen production and yield in the batch system varying parameters of mucilage/organic waste ratio, pH, acidification time, chemical oxygen demand, and temperature. The authors tested two-level factorial experiments and modeled their hydrogen profiles with the ANN and fuzzy-logic-based approaches. The manuscript does not bring scientific novelty. There are literature data for anaerobic fermentative hydrogen production from coffee mucilage, but the subject is important from ecological and biotechnological point of view. The experimental analysis was well designed, the content of the state of art provides useful information about the topic. As was confirmed by the authors the results of the study show that suing experimental results for bio digestion of waste substrate in batch reactor with ANN and fuzzy model was predictive. The fuzzy model had a greater predictive capacity than the ANN model with the response to the interaction of independent factors. As was mentioned by the authors using of coffee waste may be considered as an alternative source for energy production. My recommendation is to accept the article for the possible publication in “Energies” in present form.

Reply: We appreciate the reviewer’s warm comments. We believe that the current study will be interesting to the readers of Energies journal for providing helpful information and potentials for further research.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper analyzes and estimates the hydrogen production from coffee mucilage with organic wastes in different ratios by dark anaerobic fermentation. It uses an artificial neural network and a fuzzy logic model to calculate the production of hydrogen gas. The major contribution of this paper is on providing the information that coffee mucilage is a potential resource as the renewable energy carrier and that the fuzzy logic model is more precise to predict hydrogen production than the artificial neural network model. Some specific suggestions are given as follows.

Please describe the reason why you use these methods including artificial neural networks and a fuzzy logic model to calculate the production of hydrogen gas.
In lines 139-140, is there any basis to formulate the range of independent variables? For example, why does the range of the temperature set up at 30-40 °C?
Figure 1 and figure 2 are blurry. Please replace them with clearer figures.
In page 5, it lacks equation (2) and duplicate equation (1) while it mentions equation (2) in lines 85. Besides, it is suggested that all equations in this paper need to be centered or left-aligned uniformly and in lines 164, the sentence lacks the subject.

Author Response

March 27, 2020

Responses to Reviewers' comments:

Reviewer #3:

1) Please describe the reason why you use these methods including artificial neural networks and a fuzzy logic model to calculate the production of hydrogen gas.

Reply: We appreciate the reviewer’s thoughtful comments. There are several major parameters for anaerobic fermentation performance for hydrogen production, it is complicated to determine the final yield of molecules and compare the most critical factors with numbers. Based on these considerations and issues, we have chosen ANN model and fuzzy-logic-model. ANN is similar to humans’ neurons in the brain that would be able to make the network system and be utilized for estimating and calculating input/out-put parameters. While the fuzzy-logic-model is capable of applying for the predictive capacity without accurate knowledge of the process system the interactions of the parameters. The detailed explanations and sentences are clearly presented in lines 90-93 and 101-106 on pages 2 and 3. We believe that our intention was to fully address the reviewer’s comments.

2) In lines 139-140, is there any basis to formulate the range of independent variables? For example, why does the range of the temperature set up at 30-40 °C?

Reply: Thanks to the reviewer for pointing this out with this impressive comment. We have addressed the range of independent variables for coffee mucilage fermentation in our earlier study (Orrego et al., 2018, Energies, 11, 786: Optimization and scale-up of coffee mucilage fermentation for ethanol production). We have added some paragraphs here to support the reviewer’s comment as follows “In order to determine the optimal experimental condition for ethanol fermentation from the coffee mucilage, a full factorial design method was considered and designed with different levels of temperature, pH, and initial inoculum concentration. Generally, the Saccharomyces cerevisiae strain has its optimal growth condition at the temperature range of 25^oC - 30^oC [32] or 30^oC - 33^oC [33] with the acid pH range of 3.0 to 7.0, and the optimal pH condition is around 5.0 [33]. Another study found that S. cerevisiae was able to grow at 4^oC with the minimum growth while the maximum cell growth was achieved at 38^oC - 39^oC [34]. More recent work with the utilization of coffee mucilage reported that the reducing sugars in the coffee mucilage could effectively be fermented to ethanol at 32^oC, pH 5.1, and initial reducing sugar concentration of 61.8 g/L with the cell density of 1.0x10⁶ cfu/mL [28]. In the light of previous publications, the different levels of temperature (28^oC - 35^oC), pH (4.0 - 7.0), and initial cell concentration (3.0 g/L - 6.0 g/L) were selected to determine the optimal experimental condition. Because the substrate inhibition (reducing sugar) would cause > 250 g/L concentration [31], we have not considered the reducing sugar concentration and directly used the initial mucilage as a substrate.” We believe that we do not need to describe this again in the manuscript and this is fully address the reviewer’s comment.

3) Figure 1 and figure 2 are blurry. Please replace them with clearer figures.

Reply: Thank you for the correction; we have corrected these.

4) In page 5, it lacks equation (2) and duplicate equation (1) while it mentions equation (2) in lines 85. Besides, it is suggested that all equations in this paper need to be centered or left-aligned uniformly and in lines 164, the sentence lacks the subject.

Reply: We appreciate the reviewer’s thoughtful comments and suggestions. Based on the reviewer’s comments, we have corrected and reworded in the manuscript.

Author Response File: Author Response.pdf