Round 1

Reviewer 1 Report

Manuscript Number: processes-1254462

Operation of submerged anaerobic membrane bioreactors at 20ºC: effects of solid retention time on flux, mixed liquor characteristics and performance.

or

Kinetic control of submerged anaerobic membrane bioreactors operating at 20 oC and its influence on membrane flux, mixed liquor characteristics and overall reactor performance

GENERAL COMMENTS

This work evaluates the influence of different cell retention times in an anaerobic membrane bioreactor. It studies different parameters of the mixed liquor to try to discern the causes of the fouling seen in the membranes during an operating period of 242 days. I consider it an interesting job, although there are some operating conditions whose justification has not been entirely clear to me.

I think the work is interesting and it should be accepted after some minor revisions.

SPECIFIC COMMENTS

First, the title of the article does not correspond to the title specified in the text.

Some multi-section paragraphs should be justified like the rest of the text.

2. Materials and methods.

(136-140) It is specified that part of the biogas generated is used to recirculate it and keep the sludge in suspension. Is this generation of biogas obtained from the start of the experiment? If not, what is used to keep the sludge in suspension when no biogas has been produced yet?

Fig. 2. Some labels in the diagram are unintelligible, the size of the font used should be increased.

(151-154) What is the C: N: P ratio and solids concentration of a typical European water?

What method is used to measure the COD in the obtained gas?

3. Results and discussion:

(263) it is said that the SRTs have been reduced, on the other hand, for the S3 there is no reduction, the SRT remains at 60 d.

What reason is there to vary the SRT in S1 and S2 at 60d if they are to be maintained at 30 and 45d, respectively? Could it not have started to purge biomass at 30d in S1 and at 45d in S2?

When S3 goes from 60 to 30d, the value of MLSS is not equal to the value of S1 (SRT 30d), what is this due to? Yes, a similar value is obtained when the SRTs in S4 and S3 are equalized in the EP-3.

(379-409) The authors cite experiments in which the SRT has been varied, on the other hand, the conclusions reached by these studies are not provided.

(420) The standard deviations of the values ​​reported for S3 have not been provided.

Fig. 3 The contribution of TE to the reactors only seems to have an effect on the elimination of COD in S3 and not in S1, which according to the authors was also required.

Have the authors considered making a replica of the experiment or doubling the experimental time to verify that this contribution of SE TE requires again?

Author Response

Operation of submerged anaerobic membrane bioreactors at 20 ^oC: effect of solids retention time on flux, mixed liquor characteristics and performance

Santiago Pacheco-Ruiz, Sonia Heaven and Charles J. Banks

Reviewer 1

GENERAL COMMENTS

This work evaluates the influence of different cell retention times in an anaerobic membrane bioreactor. It studies different parameters of the mixed liquor to try to discern the causes of the fouling seen in the membranes during an operating period of 242 days. I consider it an interesting job, although there are some operating conditions whose justification has not been entirely clear to me.

I think the work is interesting and it should be accepted after some minor revisions.

SPECIFIC COMMENTS

First, the title of the article does not correspond to the title specified in the text.

Authors’ response to comment: Our apologies, the article title was changed as part of some amendments made at the Editor's request before the paper went out for review. The correct title is that currently shown on the paper, i.e. 'Operation of submerged anaerobic membrane bioreactors at 20 ^oC: effect of solids retention time on flux, mixed liquor characteristics and performance'. We hope this can be amended in the submission system to avoid any confusion.

Some multi-section paragraphs should be justified like the rest of the text.

Authors’ response to comment: Thank you - we have amended this as suggested, and hope that all sections are now correctly justified.

1. Materials and methods.

(136-140) It is specified that part of the biogas generated is used to recirculate it and keep the sludge in suspension. Is this generation of biogas obtained from the start of the experiment? If not, what is used to keep the sludge in suspension when no biogas has been produced yet?

Authors’ response to comment: The SAnMBR were operated for 48 days before day 1 of the trial, so biogas production was already well established by the start of the trial. As the system is closed, in principle it is possible to withdraw headspace gas and circulate it even when there is no observable biogas production. After the initial inoculation of the reactors the headspace was purged with N2. Feeding and headspace recirculation then started, and the N2 content of the recirculated gas fell as gas production began. Headspace N2 purge is not only good practice at laboratory and industrial pilot scale but is also a common practice at full-scale for anaerobic reactor start-up even when there is biogas mixing, such as in egg-shaped digesters where rotary sliding vane compressors are used for headspace gas recirculation.

Fig. 2. Some labels in the diagram are unintelligible, the size of the font used should be increased.

Authors’ response to comment: We have modified the font sizes in Fig 1 and Fig 2 to increase them as much as possible, and hope the labels are now readable

(151-154) What is the C: N: P ratio and solids concentration of a typical European water?

Authors’ response to comment: This is a good question, and has made us think hard about what we really meant to say here. In fact these ratios and values can vary considerably, and after reviewing the literature we are not satisfied there is a single authoritative source we can quote that gives reference values for a typical European wastewater. Henze and Comeau (2011) indicate a C:N:P ratio of around 100:22:5 for raw municipal wastewater with minor industrial contributions, with total and suspended solids concentrations of 750 and 450 mg/L respectively. Metcalf and Eddy (2004) suggests a ratio of 100:11:2, with total and suspended solids of 720 and 220 mg/L respectively in medium-strength untreated domestic wastewater; but these values are not specifically related to European wastewaters. The sentence referred to a synthetic wastewater described in a previous paper (Pacheco-Ruiz et al., 2015), itself based on some as-yet-unpublished work at the University of Southampton. We have amended the text to revise this potentially misleading comment, and hope it is now acceptable. We are grateful to the Reviewer for making this point.

References used in response:

Henze, M. & Comeau, Y. 2011. Wastewater Characterization. In: Henze, M., Loosdrecht, M. C. M. V., Ekama, G. A. & Brdjanovic, D. (eds.) Biological Wastewater Treatment: Principles, Modelling and Design. London: IWA Publishing.

Metcalf & Eddy 2004. Water Engineering Treatment and Reuse, New York, McGraw Hill.

What method is used to measure the COD in the obtained gas?

Authors’ response to comment: The COD value of methane was taken as 2.855 g COD L^-1 CH₄ at STP based on stoichiometric considerations. We have added this information to the text in section 2.3.

1. Results and discussion:

(263) it is said that the SRTs have been reduced, on the other hand, for the S3 there is no reduction, the SRT remains at 60 d.

Authors’ response to comment: All four SAnMBR were operated at a 90-day SRT throughout EP-1 (i.e. until day 59). The original text said: "On day 60 the SRT was reduced to 30, 45 and 60 days in S1, S2 and S3 respectively, while remaining at 90 days in S4". We believe the statement is therefore correct, but have modified the text in response to this and a later comment, and hope it is now clearer.

What reason is there to vary the SRT in S1 and S2 at 60d if they are to be maintained at 30 and 45d, respectively? Could it not have started to purge biomass at 30d in S1 and at 45d in S2?

Authors’ response to comment: The SRT was maintained at 90 days in all four SAnMBR until day 59 to ensure that all reactors showed similar performance when operating under the same conditions. As can be seen in Fig 2, although flux, TMP, MLSS and normalised SMP were closely similar in all four reactors during this start-up period, there were slight variations between reactors in COD removal and thus in SMP. It was thought better to allow these to stabilise and ensure that all reactors started from the same conditions on day 60, with the same operating history up to that point. We hope this explanation is satisfactory.

When S3 goes from 60 to 30d, the value of MLSS is not equal to the value of S1 (SRT 30d), what is this due to? Yes, a similar value is obtained when the SRTs in S4 and S3 are equalized in the EP-3.

Authors’ response to comment: This difference is due to the different strategies adopted when changing the SRT in each case. During EP-1 the SRTs in S1 and S3 were respectively 30 and 60 days, so as expected there was a difference in MLSS at the end of EP-1. At the start of EP-2 the SRT in S3 was reduced from 30 to 60 days by increasing the volume of mixed liquor removed each day. No other changes were made with regard to MLSS content at this point. As a result, the MLSS concentration fell progressively, until by the end of EP-2 MLSS concentrations in both reactors with a 30-day SRT were similar (Fig 2 and Table A1b). In contrast, when the SRT in S4 was reduced to 30 days at the beginning of EP-2, as well as increasing the amount of MLSS removed each day the MLSS concentration in S4 was directly reduced. This was done by removing a proportion of the reactor contents over the next four days, and replacing them with low-solids influent. As can be seen in Fig 2, this succeeded in producing a step-change in MLSS concentration in S4 around day 160-165, followed by a slower reduction due to washout. This strategy was deliberately adopted in S4 to provide a contrast to the slower solids washout in S3, and produced some interesting effects on CST and FIC as noted in the subsequent discussion of these parameters. The process used to produce a step reduction in MLSS concentration in S4 is described in section 3.1.1. We have modified the text to clarify this, and have also added a comment on the different effects of these different strategies in the last paragraph of section 3.3.4.

(379-409) The authors cite experiments in which the SRT has been varied, on the other hand, the conclusions reached by these studies are not provided.

Authors’ response to comment: Reviewer 2 also commented on this section, and we have therefore removed this material and placed some of it in the Introduction. We have not provided an extensive discussion of all studies in which SRT has been varied: this is in part because, as noted, in many cases the systems were not operated for the equivalent of 3 SRT (i.e. for long enough to approach steady-state conditions). Wherever relevant, however, we have discussed the findings of the studies carried out by these authors: for example with respect to COD removal rates [30, 31, 33], TE requirements [26, 50, 51]; dissolved methane and COD balances [31]; SMP [24, 32-34]; membrane fouling [33-34]; CST [29, 54, 55-58];EPS [22]. We hope that, together with the changes made, this approach is now satisfactory, and thank the Reviewer for raising this point. (Please see also response to Reviewer 2).

(420) The standard deviations of the values ​​reported for S3 have not been provided.

Authors’ response to comment: We deliberately did not provide standard deviations for COD removal in S3 as removal rates were falling steadily over the period reported, in contrast to the other SAnMBR where conditions had stabilised to within the stated ranges. We hope this explanation is satisfactory.

Fig. 3 The contribution of TE to the reactors only seems to have an effect on the elimination of COD in S3 and not in S1, which according to the authors was also required.

Authors’ response to comment: Before TE addition the COD removal rates in S1 and S3 were around 75%. After addition rates in both reactors rose quite sharply. This can be seen in Fig 2, although precise values are difficult to read in graphical format; and the numerical values are given in Table A1a. We have therefore not amended the text, and hope this clarification is satisfactory.

Have the authors considered making a replica of the experiment or doubling the experimental time to verify that this contribution of SE TE requires again?

Authors’ response to comment: We did consider extending or replicating the experiment as suggested, but unfortunately this was not possible at the time; and after subsequent consideration and discussions we believe that a purpose-designed experimental programme would be the best approach to find out more about TE requirement in AnMBR at different SRT. We fully agree with the Reviewer that this is an important and interesting topic and more research is needed in this area.

Reviewer 2 Report

The paper explored the effect of the SRT on the treatment, membrane performance, methane production of gravity driven submerged AnMBR using low strength synthetic wastewater at psychrophilic temperature.  It present an interesting finding of washout of trace elements at low SRT. A fair design of experiment is there, however, the results sections is too much explanatory and need to be compressed. I recommend the publication of this articles after incorporating the suggestions given below

Major corrections:

Comment1: Results section is too much explanatory and must be compressed.

Comment2: Section 3.1.1 (line 295 – 305): authors should made a table enlisting all phases with operational parameters. It shall summarise the section, instead of explaining it in text in results. Please put all this in under a new heading of experimental plan in materials and methods section.

Comment3: Line 380 – 409; authors have reported extensive literature example but not clear for what purpose? I think it’s pointless to keep reporting the studies. You can report 2 to 3 studies and reference other similar studies. At the end of para you are concluding the gap in literature. I think it should be done in introduction not in results and discussion section. You can compare the results here and possible scientific reasons in this section.

Comment4: Please report the synthetic wastewater composition in materials and method section.

Minor comments

Comment5: It would be better if authors could briefly list the method of EPS i.e. EPS extraction solution composition and centrifugation RFC value and time.

Comment6: What was the phenol concentration used for carbohydrates measurement?

Comment7: It is suggested to explain the figures a bit more. For example, you can explain  what is s1, s2, s3, s4 in each figure with little information on operating parameters in parenthesis (). This shall help readers.

Comment8: Can you please explain about Frozen image centrifugation test in the methods?

Comment 9: As the study involves the evaluation of SAnMBR membrane performance, sludge characterization and fouling phenomena at low temperature. The following recent paper on the topic should be cited in introduction as an attempt to improve AnMBR efficiency using cold adopted at low temperature and exploring the biological reason of fouling at low temperature.

• Tabraiz, Shamas, et al. "Temperature and immigration effects on quorum sensing in the biofilms of anaerobic membrane bioreactors." Journal of Environmental Management293 (2021): 112947.

Author Response

Operation of submerged anaerobic membrane bioreactors at 20 ^oC: effect of solids retention time on flux, mixed liquor characteristics and performance

Santiago Pacheco-Ruiz, Sonia Heaven and Charles J. Banks

Reviewer 2

The paper explored the effect of the SRT on the treatment, membrane performance, methane production of gravity driven submerged AnMBR using low strength synthetic wastewater at psychrophilic temperature. It present an interesting finding of washout of trace elements at low SRT. A fair design of experiment is there, however, the results sections is too much explanatory and need to be compressed. I recommend the publication of this articles after incorporating the suggestions given below

Authors’ response to comment: We thank the Reviewer for these helpful comments and have done our best to address them as described below.

Major corrections:

Comment1: Results section is too much explanatory and must be compressed. 

Authors’ response to comment: We thought very hard about this, even before submitting the paper and receiving comments from the Reviewers, because we appreciate that some parts of the results - particularly Section 3.1.1 on flux and TMP - make heavy reading. Unfortunately it is difficult to reduce the text much further as the information is sequential, and presentation of these results is essential to our point that long studies are needed. In addition, the detailed observation of small changes in flux etc is one of the advantages of the methods used and may provide potentially useful insights for future researchers. We also note that the other Reviewers have not raised this issue, and all Reviewers have in fact requested addition of some information. Other than removing some material from section 3.1.1. we have therefore made only relatively small changes to the results section, but hope that this explanation is acceptable.

Comment2: Section 3.1.1 (line 295 – 305): authors should made a table enlisting all phases with operational parameters. It shall summarize the section, instead of explaining it in text in results

Please put all this in under a new heading of experimental plan in materials and methods section.

Authors’ response to comment: In many cases the changes to operational parameters in each phase were introduced as a direct result of the observed behaviour and performance of the SAnMBR: thus it is difficult to present this information in the format of a plan in the Methods section. Rather than adding a new table as suggested, we have therefore modified the text to direct the reader to Table A2 which contains details of control and operational parameters applied during each phase. We are grateful to the Reviewer for making us think again about this: we believe this makes the paper easier to understand and hope this solution is acceptable.

Comment3: Line 380 – 409; authors have reported extensive literature example but not clear for what purpose? I think it’s pointless to keep reporting the studies. You can report 2 to 3 studies and reference other similar studies. At the end of para you are concluding the gap in literature. I think it should be done in introduction not in results and discussion section. You can compare the results here and possible scientific reasons in this section.

Authors’ response to comment: To date there has been no review of the effects of SRT, and the individual studies are used for comparison later in the paper, so we believe this summary is useful. We initially decided to put the material into this section rather than the Introduction, firstly because we wanted to emphasise the value of maintaining the operating conditions for long enough to reach steady state in studies where the SRT has been varied; and secondly because we felt it might seem too negative to open with a list of papers where this was not achieved. But on re-reading, and in the light of a similar comment by Reviewer 1, we agree this does not belong in this section or achieve what we intended here. We have therefore moved some of the material to the Introduction as suggested, and revised the text here. We are grateful to the Reviewers for raising this point, and hope that the paper is improved as a result. (Please see also response to Reviewer 1).

Comment4: Please report the synthetic wastewater composition in materials and method section

Authors’ response to comment: As the current paper is rather long and the recipe has already been published, in accordance with good practice we have not added it to the Methods section but have included it in Additional Materials in Appendix A for ease of reference. We hope this is satisfactory.

Minor comments

Comment5: It would be better if authors could briefly list the method of EPS i.e. EPS extraction solution composition and centrifugation RFC value and time.

Authors’ response to comment: As the details are quite lengthy, to avoid expanding the main text we have added this information to Appendix A in the form of a flowchart. We hope this is acceptable.

Comment6: What was the phenol concentration used for carbohydrates measurement?

Authors’ response to comment: The phenol concentration was 80% on a weight basis, as specified in the method in reference [38].

Comment7: It is suggested to explain the figures a bit more. For example, you can explain what is s1, s2, s3, s4 in each figure with little information on operating parameters in parenthesis (). This shall help readers.

Authors’ response to comment: We have had extensive discussions on how this can best be achieved. Unfortunately Figures 2-6 show results for all four SAnMBR throughout the full experimental period, so it is extremely difficult to give details of what is happening in each reactor in every case. Values of the control parameters SRT and TMP are shown in Figure 2a and b, making it relatively easy to read this Figure. We have edited the captions and added extra information to Figures 3-6 to make them clearer, and hope this solution is acceptable.

Comment8: Can you please explain about Frozen image centrifugation test in the methods?

Authors’ response to comment: This is a very useful point, as the technique is not widely used or known. We have added a brief explanation of the principle in the methods, and we thank the Reviewer for raising this.

Comment 9: As the study involves the evaluation of SAnMBR membrane performance, sludge characterization and fouling phenomena at low temperature. The following recent paper on the topic should be cited in introduction as an attempt to improve AnMBR efficiency using cold adopted at low temperature and exploring the biological reason of fouling at low temperature.

Tabraiz, Shamas, et al. "Temperature and immigration effects on quorum sensing in the biofilms of anaerobic membrane bioreactors." Journal of Environmental Management293 (2021): 112947

Authors’ response to comment: This is a very interesting and useful paper on quorum effects in biofilms in AnMBR at 4 and 15 ^oC, which has only just become available. As the current study focuses on the effect of SRT, rather than including it in the Introduction as suggested we have added it to section 3.3.2 where there is a brief discussion of the importance of these emerging topics. We thank the Reviewer for this valuable suggestion.

Reviewer 3 Report

The manuscript is original, interesting and in my view, it is a great contribution in the field of study, since it presents a wide range of experimental data that allow to analyze and deepen the application of this technology in the conditions tested.

In my opinion, the manuscript is highly attractive to other readers and the proposed methodology is clear. The manuscript is well written. The experiments have been well conducted and the conclusions are clearly supported by the results.

I suggest this paper should be accepted but after minor revisions that can be seen below.

Results

• In my opinion, the results seem to indicate that the experimental phases should have been longer since most of the parameters measured do not reached stabilization, which makes it difficult to affirm some of the results found. For example, lines 312 and 313 the authors state that the MLSS concentration has been stabilising in a value, however the figure shows a slight decrease in that value.
• The membranes in the last experimental phases have already suffered a previous fouling that could alter the result of the experimentation, because the HRT is related to the J. Why did the authors not carry out a cleaning before each phase? With the cleaning of the membranes, the behaviour could be analysed without taking into account the previous fouling. Do the authors consider that this factor may affect their results?

Conclusions

• In my opinion, the lines 971-975 is an abstract not a conclusion.

Author Response

Operation of submerged anaerobic membrane bioreactors at 20 ^oC: effect of solids retention time on flux, mixed liquor characteristics and performance

Santiago Pacheco-Ruiz, Sonia Heaven and Charles J. Banks

Reviewer 3

The manuscript is original, interesting and in my view, it is a great contribution in the field of study, since it presents a wide range of experimental data that allow to analyze and deepen the application of this technology in the conditions tested.

In my opinion, the manuscript is highly attractive to other readers and the proposed methodology is clear. The manuscript is well written. The experiments have been well conducted and the conclusions are clearly supported by the results.

Authors’ response to comment: We are very grateful indeed for these positive remarks and for the useful comments made.

I suggest this paper should be accepted but after minor revisions that can be seen below.

Results

• In my opinion, the results seem to indicate that the experimental phases should have been longer since most of the parameters measured do not reached stabilization, which makes it difficult to affirm some of the results found. For example, lines 312 and 313 the authors state that the MLSS concentration has been stabilising in a value, however the figure shows a slight decrease in that value.

Authors’ response to comment: We fully agree that it is always better to run experiments of this type for as long as possible: 3 SRT is only a convenient guideline, not an absolute definition. On the other hand, this has to be balanced with the fact that experimental time is also a precious resource! In this case, the reactors were operated for the equivalent of 1.78 SRT at a 90-day SRT, 0.9 HRT at 60 days; 4.1 SRT at 40 days; 3.4 and 4.4 SRT at 30 days and 4.1 SRT at 20 days, giving some confidence that results at 40-, 30- and 20-day SRTs should be representative of steady-state conditions. In the specific case mentioned, at the end of EP-2 the last three values for MLSS concentration in S1 were 4.35, 4.35 and 4.37 mg/L and in S2 were 5.84, 5.78 and 5.94 mg/L respectively, indicating little or no change in both cases. We have therefore not modified the text in this case, and hope this is acceptable.

• The membranes in the last experimental phases have already suffered a previous fouling that could alter the result of the experimentation, because the HRT is related to the J. Why did the authors not carry out a cleaning before each phase? With the cleaning of the membranes, the behaviour could be analysed without taking into account the previous fouling. Do the authors consider that this factor may affect their results?

Authors’ response to comment: This is an interesting point: both options, with and without cleaning of membranes, can provide useful information and insights. In this case we chose not to carry out cleaning before each phase mainly because we were interested to see how changing the SRT affected the TMP required to maintain a target flux in an already-fouled membrane. In fact as discussed in section 3.2 reducing the SRT led to a recovery in flux, implying that the membrane fouling was reversible to some degree. Both CST and FIC testing showed that responses to a change in SRT could be significantly delayed, reflecting the different timescale on which growth-related, metabolic and physicochemical changes occur as well as the interaction between these factors; this confirms once again the importance of long-term operation to ensure that evaluation of system performance and mixed liquor characteristics takes place under steady-state conditions. Operation over the full experimental period, without chemical or external cleaning, not only demonstrated the effects of SRT on performance parameters; but also indicated that membrane fouling could be at least partially reversed if an optimal SRT is applied. The results indicated that the different timescale on which growth-related, metabolic and physico-chemical changes occur may lead to a delayed response to changes in SRT; thus confirming the importance of long-term operation to allow full evaluation of system performance and mixed liquor characteristics under steady-state conditions. Together with earlier findings [25] in this configuration the results enable us to understand that a sustainable / stable membrane flux is achievable under specific operational conditions.

Conclusions

• In my opinion, the lines 971-975 is an abstract not a conclusion.

Authors’ response to comment: We apologise and have modified the text in a way which we hope makes it clearer why we believe this is a useful conclusion from the current work. In this connection, we also made some changes to the last paragraph of section 3.3.4 in response to a point raised by Reviewer 1 which we hope help to highlight the importance of this topic.