Determination of Nitrogen and Sulphur Mineralization in Batch and Semi-Continuous Anaerobic Digestion Using an Artificial Fiber Bag Technique

Round 1

Reviewer 1 Report

The paper is well organized, generally easy to read, and contains very valuable original results. The paper deals the validation of a mento through batch anaerobic digestion with and without artificial bags. However, the following minor comments can help improve the manuscript.

Section 2.1. A picture of the bags could be included at least in supplementary materials. In my opinion could be a helpful information for the readers.

Section 2.2. A summary table with the different operational conditions could allow a quick distinction between the experiments.

Table 1 and section 2.8 should be moved to results section.

Section 3.1.2 and from Line 489 to 494. Results obtained by other authors should be included. I know that the aim is the validation between the reactors and bags, but I think it is interesting to provide values obtained by other authors with the same or similar substrates.

Author Response

The paper is well organized, generally easy to read, and contains very valuable original results. The paper deals the validation of a mento through batch anaerobic digestion with and without artificial bags. However, the following minor comments can help improve the manuscript.

Section 2.1. A picture of the bags could be included at least in supplementary materials. In my opinion could be a helpful information for the readers.

A picture of a single bag and multiple bags on a chain, as used in the continuous experiment, have been added as supplementary material and can be seen at the end of the revised manuscript.

Section 2.2. A summary table with the different operational conditions could allow a quick distinction between the experiments.

A table has been added.

Table 1 and section 2.8 should be moved to results section.

The table and section have been moved to the results section, apart from the first few lines describing methods.

Section 3.1.2 and from Line 489 to 494. Results obtained by other authors should be included. I know that the aim is the validation between the reactors and bags, but I think it is interesting to provide values obtained by other authors with the same or similar substrates.

There is very little information available, mainly H2S measurements, but not much about S mineralization in batch assays, possibly for the reasons tackled in this work. Most S mineralization studied in literature has been in the soil. However, I found and added a reference about the amount of S that is typically lost during AD:

"The S degradation of CM and OFMSW seemed realistic reaching 40% and 80% degradation in the thermophilic digester, as it has been shown that less than 50% of added S leaves a digester in the digestate [32]."

Reviewer 2 Report

This manuscript describes the implementation of a new method into the AD method toolbox. More specifically, the authors aimed to introduce the use of artificial fiber bags (AFB) as a tool for measuring the degradation rates of different substrate fractions within a co-digestion system, by monitoring mass loss, as well as N and S concentrations to evaluate their rate of mineralization. The feasibility of using this method was evaluated in batch bottles as well as in a semi-continuous reactor systems. This approach is really interesting for AD, since it allows for a separate study of substrate hydrolysis in a co-digestion system, where otherwise the combination of different substrates tends to obscure the effects of each individual one.

Overall, the manuscript is superbly written and thoroughly describes a well-designed experiment. The topic is very interesting, novel, and of high relevance to the scientific community. Some parts, however, should be improved. In particular, the authors should be careful with aspects related to statistics:

• Lines 228-230 the authors state that “each treatment was compared by a t-test of the model parameters fitted to individual replicates”. This sentence is somewhat difficult to understand – one could interpret it as if each replicate of each treatment was compared to each replicate of each other treatment, which is probably not the case. Please revise this to make it clearer to the readers how the values were compared. Furthermore, if several means were compared (i.e. each treatment against each other treatment), then the t-test is not appropriate due to an increased chance of a type I error occurring. An ANOVA and/or a Tukey HSD test should be used instead.
• Figure 1: The experiment with no AFB appears to have performed better than the rest, since as the authors mention it produced about 10% more methane than the others, while also having faster kinetics. The authors mention that there is no statistical significance in the difference between mean methane productions and move on, which is problematic, since the lack of significance does not confirm that there is no difference in reality. One possible explanation could be a too low statistical power for the detection of such an effect, since there were only 3 replicates in the experiment. In any case, the authors should be careful in the phrasing of their arguments related to this observation. It would be better to include a few sentences of discussion acknowledging that the lack of significance does not necessarily mean a lack of true difference and discussing what such a result could mean and why it could happen (similarly as was done later on when the authors discuss the kinetics differences). It would also be better if the p values were written out instead of just being below or above 0.05, since this will provide the readers with information on the probabilities that the observed difference is due to chance (i.e. p > 0.05 could mean p=0.9 or p=0.051, which is a large difference that gets obscured when it is summarized into lower or higher than 0.05).
• Figure 1 – AFB only: Why are there no error bars on this curve? Were all replicates very similar?
• Line 296: The phrase greater significance is not appropriate. It would be better if the authors just state that the differences were found to be statistically significant and report the actual p values.

Minor comments:

• Line 46: The authors are advised to avoid the term ”producing energy”, since that is not possible. A better alternative could be “extracting energy”.
• When the AFB method is performed in semi-continuous reactors, the solid retention time is decoupled from the hydraulic retention time. This will have implications for several parameters in the reactor, particularly the gas production and extent of degradation, which means one needs to be careful in which conclusions one can make from the data. The authors did this well, but it would still be interesting to see some discussion of these aspects in the text (i.e. what can or cannot be done with this method). This could add additional value to the readers who aim to implement the method in their own experiments.

Author Response

This manuscript describes the implementation of a new method into the AD method toolbox. More specifically, the authors aimed to introduce the use of artificial fiber bags (AFB) as a tool for measuring the degradation rates of different substrate fractions within a co-digestion system, by monitoring mass loss, as well as N and S concentrations to evaluate their rate of mineralization. The feasibility of using this method was evaluated in batch bottles as well as in a semi-continuous reactor systems. This approach is really interesting for AD, since it allows for a separate study of substrate hydrolysis in a co-digestion system, where otherwise the combination of different substrates tends to obscure the effects of each individual one.

Overall, the manuscript is superbly written and thoroughly describes a well-designed experiment. The topic is very interesting, novel, and of high relevance to the scientific community. Some parts, however, should be improved. In particular, the authors should be careful with aspects related to statistics:

• Lines 228-230 the authors state that “each treatment was compared by a t-test of the model parameters fitted to individual replicates”. This sentence is somewhat difficult to understand – one could interpret it as if each replicate of each treatment was compared to each replicate of each other treatment, which is probably not the case. Please revise this to make it clearer to the readers how the values were compared. Furthermore, if several means were compared (i.e. each treatment against each other treatment), then the t-test is not appropriate due to an increased chance of a type I error occurring. An ANOVA and/or a Tukey HSD test should be used instead.

We agree with the reviewer and thank them for their input, in that it was better to use ANOVA as there were more than two means. There have been considerable changes made that relate to the comments below as well. Fortunately, changing from t-test to ANOVA did not change whether anything was or was not statistically significant at the p>0.05.

• Figure 1: The experiment with no AFB appears to have performed better than the rest, since as the authors mention it produced about 10% more methane than the others, while also having faster kinetics. The authors mention that there is no statistical significance in the difference between mean methane productions and move on, which is problematic, since the lack of significance does not confirm that there is no difference in reality. One possible explanation could be a too low statistical power for the detection of such an effect, since there were only 3 replicates in the experiment. In any case, the authors should be careful in the phrasing of their arguments related to this observation. It would be better to include a few sentences of discussion acknowledging that the lack of significance does not necessarily mean a lack of true difference and discussing what such a result could mean and why it could happen (similarly as was done later on when the authors discuss the kinetics differences). It would also be better if the p values were written out instead of just being below or above 0.05, since this will provide the readers with information on the probabilities that the observed difference is due to chance (i.e. p > 0.05 could mean p=0.9 or p=0.051, which is a large difference that gets obscured when it is summarized into lower or higher than 0.05).

We have added text regarding the low statistical power and how the results should be used with caution. We have also written the actual p values, as suggested. The added text is in red in the revised manuscript and is spread throughout the section.

• Figure 1 – AFB only: Why are there no error bars on this curve? Were all replicates very similar?

The “AFB only” sample was only performed in a single bottle as it was not expected that an empty bag would have any effect on the yields. If an effect was measured, we would have investigated it further. We have made this clearer wherever it is mentioned in the manuscript, as well as adding a table of the batch bottles setup, as was suggested by another reviewer.

• Line 296: The phrase greater significance is not appropriate. It would be better if the authors just state that the differences were found to be statistically significant and report the actual p values.

This has been done, as suggested.

Minor comments:

• Line 46: The authors are advised to avoid the term ”producing energy”, since that is not possible. A better alternative could be “extracting energy”.

This has been changed, as suggested.

• When the AFB method is performed in semi-continuous reactors, the solid retention time is decoupled from the hydraulic retention time. This will have implications for several parameters in the reactor, particularly the gas production and extent of degradation, which means one needs to be careful in which conclusions one can make from the data. The authors did this well, but it would still be interesting to see some discussion of these aspects in the text (i.e. what can or cannot be done with this method). This could add additional value to the readers who aim to implement the method in their own experiments.

This was a very good point that has been considered for future experiments. We have added the following section regarding a possible method of compensating for the washout effect, or lack thereof:

“It should be noted that despite the reactors being operated semi-continuously, the substrates held in the bags are not subject to the reactor HRT. Thus, the reactor HRT and solids retention time (SRT) are decoupled and the substrate remaining in the bag is spending longer in the reactor than the bulk fluid. It is possible to adjust the mass loss-es in the bags to compensate for this effect. In the study conducted by Ward et al. [28], calculations were made to calculate mass loss over time, based on HRT and an as-sumption of perfect mixing (i.e. complete homogenization of reactor contents). One could apply such calculations of mass loss over time due to washout caused by the continuous operation and apply corrections to the measured masses remaining in the bags. This could be used to improve the method with continuous reactors, but should be used with some caution due to the assumption of perfect mixing.”

Reviewer 3 Report

Only a little comment to the introduction chapter; the authors could have given an overview of earlier work on determining mineralization of N and S. How has this been done earlier? 

Author Response

Only a little comment to the introduction chapter; the authors could have given an overview of earlier work on determining mineralization of N and S. How has this been done earlier? 

We thank the reviewer for this suggestion. Although it was a “little comment” it was still challenging as there is not a lot of information regarding measuring S during a batch test. The problem is about having to measure the inoculum at the same time as the substrate and we have added the following section and citation in the introduction and a short section and citation in the Results and Discussion regarding S mineralization in AD:

“In conventional batch assays, it is not possible to effectively separate the substrate from the inoculum, and any analysis of chemical parameters will be to some extent contaminated by the inoculum. Normally, chemical parameters such as S concentration are measured in the combined substrate with inoculum, if measurements after digestion has proceeded are required [9].”

“The S degradation of CM and OFMSW seemed realistic reaching 40% and 80% degradation in the thermophilic digester, as it has been shown that less than 50% of added S leaves a digester in the digestate [33].”