Pilot Plant Test of Single-Pass Electrodialysis Reversal System

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

This is a well-structured, experimental study on improving electrodialysis reversal (EDR) performance through single-pass operation, custom spacer design, and pressure control. The work is relevant to the field of membrane-based desalination, presents novel engineering solutions, and includes a thorough pilot-scale validation and economic analysis. However, the following points need to be improved before acceptance.

1 The spacer design is highlighted as novel, but its geometry and fabrication are only briefly mentioned. A diagram or detailed description in the main text would help readers understand its innovation.

2 Table 3 (daily pilot results) is dense and difficult to interpret. Consider summarizing the key trends based on Table 3.

3 Figure 3 references previous work ref. [20] but is not clearly linked to the current study’s findings. Explain how the spacer performance metrics (Sh vs. Pn) relate to the observed improvements in desalination.

4 The economic comparison with the Stalowa Wola plant is valuable, but some assumptions (e.g., membrane cost, equipment cost per m², lifetime) are not fully justified or referenced. Sensitivity analysis on key parameters would strengthen the conclusions.

5 A brief comparison with reverse osmosis (RO) or electrodeionization (EDI) in terms of energy consumption, membrane lifetime, and operational complexity would highlight the advantages of electrodialysis reversal.

6 Add a conclusion section summarizing the key findings (recovery, deep demineralization, cost reduction) of the study.

Author Response

Thank you for the review. We have revised our manuscript to address the issues you pointed out. The changes from the previous version are highlighted in yellow. Our detailed point-by-point response is below:

“1 The spacer design is highlighted as novel, but its geometry and fabrication are only briefly mentioned. A diagram or detailed description in the main text would help readers understand its innovation.”

We have added the discussion of the spacer mesh geometry and how we think it links to the mass transfer improvements in section 2.1.2 Novel intermembrane spacers.

“2 Table 3 (daily pilot results) is dense and difficult to interpret. Consider summarizing the key trends based on Table 3.”

We have re-arranged the data in Tab. 3 to make it more clear what we were trying to achieve. We also add a paragraph in section 3.2 explaining how we were gradually increasing work/spec off time and recovery to push our system and test its stability.

“3 Figure 3 references previous work ref. [20] but is not clearly linked to the current study’s findings. Explain how the spacer performance metrics (Sh vs. Pn) relate to the observed improvements in desalination.”

We added discussion to section 2.1.2 on how Sh and Pn are linked to EDR properties (limiting current density, pressure drop).

“4 The economic comparison with the Stalowa Wola plant is valuable, but some assumptions (e.g., membrane cost, equipment cost per m², lifetime) are not fully justified or referenced. Sensitivity analysis on key parameters would strengthen the conclusions.”

We have added the sensitivity analysis to section 3.3 with a discussion of why these values were chosen.

“5 A brief comparison with reverse osmosis (RO) or electrodeionization (EDI) in terms of energy consumption, membrane lifetime, and operational complexity would highlight the advantages of electrodialysis reversal.”

We have added to section 1 a brief explanation of why EDR can be chosen over RO in brackish water demineralization.

“6 Add a conclusion section summarizing the key findings (recovery, deep demineralization, cost reduction) of the study.”

We have added the missing conclusions.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

This manuscript reports a pilot-scale investigation of a single-pass electrodialysis reversal (EDR) system with pressure regulation via outlet throttling to enable different linear flow velocities in diluate and concentrate compartments. While the study presents interesting operational results, the manuscript still lacks sufficient mechanistic justification and several issues related to clarity, consistency, and rigor of the economic assessment. Major revision is required before the manuscript can be considered for publication.

1. The novelty of this approach is not sufficiently clarified. Please provide adequate description for the novelty of this work.
2. The authors should clearly explain how this strategy differs from existing methods used in industrial ED/EDR systems.
3. A comparison with previously reported pressure control approaches (if any) should be added to better position the contribution.
4. The manuscript repeatedly claims operation in the overlimiting current regime and attributes deep demineralization to electroconvection and spacer-induced effects. However, this discussion remains largely speculative.
5. No direct evidence (e.g., current–voltage curves, visualization, impedance data, or limiting-current determination under pilot conditions) is provided to confirm overlimiting operation.
6. Some tables (notably Table 3) are dense and difficult to interpret; reorganization or partial graphical representation is recommended.
7. There is a lack of references for this work. Some demonstrations are required with references.
8. The economic analysis presented here is based on the central assumption, any cases to support your assumption?
9. Where is the conclusion of this manuscript?

Author Response

Thank you for the review. We have revised our manuscript to address the issues you pointed out. The changes from the previous version are highlighted in yellow. Our detailed point-by-point response is below:

“1. The novelty of this approach is not sufficiently clarified. Please provide adequate description for the novelty of this work.” 

We have rewrote the last paragraphs of section 1 and added brief description to section 2.1 to clarifiy the novelty.

“2. The authors should clearly explain how this strategy differs from existing methods used in industrial ED/EDR systems.”

We have added an explanation of which elements of our EDR unit are not used at all in the industry (e.g. the hexagonal mesh spacer and the pressure control strategy).

“3. A comparison with previously reported pressure control approaches (if any) should be added to better position the contribution.”

We are not aware of any research on pressure control strategies. The industrial practice is just to keep flow rate more or less similar and apply feed-and-bleed configuration if higher recovery is needed. We have added statements to the section 1 of the manuscript to clarify that.

“4. The manuscript repeatedly claims operation in the overlimiting current regime and attributes deep demineralization to electroconvection and spacer-induced effects. However, this discussion remains largely speculative.”

We agree our assessment of overlimitting regime is not directly proven. We have added statements to the manuscript to clarify that we are basing our speculations on the fact we have worked at much higher voltage drop per membrane pair that is usually required to obtain limiting current in the existing literature and the fact the voltage drop in the EDR pilot was much higher the voltage drop at the limiting current density that we measured for the same spacer/membrane/feed combination during the initial, auxiiliary tests. We agree that the direct determination of LCD in the pilot plant would have been more convincing for the reader.

“5. No direct evidence (e.g., current–voltage curves, visualization, impedance data, or limiting-current determination under pilot conditions) is provided to confirm overlimiting operation.”

See the response for issue #4.

“6. Some tables (notably Table 3) are dense and difficult to interpret; reorganization or partial graphical representation is recommended.”

We have re-arranged the data in Tab. 3 to make it more clear what we were trying to achieve. We also add a paragraph in section 3.2 explaining how we were gradually increasing work/spec off time and recovery to push our system and test its stability.

“7. There is a lack of references for this work. Some demonstrations are required with references.”

We have expanded the section 1 to provide some examples of pilot-scale EDR being investigated.

“8. The economic analysis presented here is based on the central assumption, any cases to support your assumption?”

Our central assumption is based on the industrial practice – the EDR manufacturers recommend that the pressure difference should be kept below specific value. The novelty here is that instead keeping the diluate pressure drop low, we are deliberatly throttling the concentrate to increase the pressure drop in that channel, thus decreasing the concentrate/diluate pressure difference. We are not aware of any research on the membrane life-time when this type of pressure control is applied. In our research, we have not observed any effect on the membranes after the module was disassembled; however, we acknowledge that the membrane degredation may creep in after prolonged operation (3 years? 5 year? Or maybe never? It’s anyone’s guess) and our tests were simply not long enough. We have added the sensitivity analysis to the economic analysis to show that even 20% decrease in the membrane life-time would still make our solution economically competitive.

“9. Where is the conclusion of this manuscript?”

We have added the missing conclusions.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

Overall Comments

This manuscript deals with pilot-scale experimental results of a river water desalination system based on single-pass electrodialysis reversal (EDR). The topic itself is practically relevant, and the operation of single-pass EDR may be of interest to the water treatment community.

However, in its current form, the manuscript does not meet the requirements of a scientific research article. In particular, the research objectives are not clearly defined in introduction, and the experimental strategy is insufficiently explained. Moreover, the presentation and interpretation of results lack scientific rigor.

At present, the manuscript reads more like a technical pilot test report than a hypothesis-driven research paper. Substantial revision is required before the manuscript can be considered for publication.

 

Major Comments

1. The manuscript investigates a single-pass EDR system for river water desalination. However, it is unclear about the following points:

- What specific scientific or engineering question this study aims to answer

- What hypothesis is being tested

- What is the novel point compared to existing studies on EDR or single-pass operation

The Introduction should explicitly define the research objectives and the novelty of the work.

 

2. Multiple operating conditions are tested. However, the relationship between experimental conditions and research objectives is unclear. The purpose of comparing different cases is not systematically described. A clearer explanation of the experimental design and strategy is required.

 

3. Plots shown in Figure 6 seem to be simulation results rather than experimental data. If these are experimental results, there should be uncertainty or variability measurement. Also, repetition of experiments should be clarified.

Without this information, it is difficult to assess the reliability and reproducibility of the results.

 

4. While the tables contain a large amount of data, it is unclear:

- Which parameters are most important

- Which comparisons are intended

- What conclusions should be drawn from the tables

Each table should be accompanied by clear explanations in the text that guide the reader toward the key findings. 

Or, these results should be shown in several diagrams to efficiently convey authors' interpretation of them.

 

5. The Discussion section mainly presents speculative explanations, but it is not clearly shown which experimental results support these interpretations. The authors should clearly distinguish what is directly supported by the data from hypotheses or future research directions.

 

6.The logical linkage between objectives, methods, results, discussion, and conclusions is weak.
The manuscript would benefit from a comprehensive restructuring so that the reader can clearly understand

- What was investigated

- How it was investigated

- What was found

- Why the findings are important

 

While the topic is potentially interesting, the manuscript does not currently satisfy the standards of a scientific research paper. Therefore, Major Revision is recommended.

Comments on the Quality of English Language

The overall English language quality requires significant improvement. While the manuscript is generally understandable, there are frequent issues with sentence structure, word choice, and clarity that hinder readability. In particular, several long and complex sentences obscure the intended meaning, and technical terms are sometimes used imprecisely. Substantial language editing by a proficient English speaker or a professional editing service is strongly recommended before publication.

Author Response

Thank you for the review. We have revised our manuscript to address the issues you pointed out. The changes from the previous version are highlighted in yellow. Our detailed point-by-point response is below:

 

“1. The manuscript investigates a single-pass EDR system for river water desalination. However, it is unclear about the following points: - What specific scientific or engineering question this study aims to answer – What hypothesis is being tested – What is the novel point compared to existing studies on EDR or single-pass operation The Introduction should explicitly define the research objectives and the novelty of the work.”

 

We have added the statement to the end of section 1 to excplitly define the research objectives and novelty.

 

“2. Multiple operating conditions are tested. However, the relationship between experimental conditions and research objectives is unclear. The purpose of comparing different cases is not systematically described. A clearer explanation of the experimental design and strategy is required.”

 

We have rearranged the presentation of pilot data and added the explanation of operational strategy to section 3.2 to make clear what we were trying to achieve. We have also added an explanation to the end of section 1 to clearly separate the presented research into related parts.

 

“3. Plots shown in Figure 6 seem to be simulation results rather than experimental data. If these are experimental results, there should be uncertainty or variability measurement. Also, repetition of experiments should be clarified.

Without this information, it is difficult to assess the reliability and reproducibility of the results.”

 

The Fig. 6 is indeed the experimental data; however, we treated it only as auxilliary experiment to confirm whether our estimation of the required length/time made sense before we started building the pilot-scale unit. Since the batch test confirmed we can reach deep demineralization, we did not repeat this particular experiment and the main focus was placed on the pilot unit operation. We have added explicit statement about that to avoid misleading the reader.

 

““4. While the tables contain a large amount of data, it is unclear:

- Which parameters are most important

- Which comparisons are intended

- What conclusions should be drawn from the tables

Each table should be accompanied by clear explanations in the text that guide the reader toward the key findings. 

Or, these results should be shown in several diagrams to efficiently convey authors' interpretation of them.”

 

We have rearranged the tables so hopefully they are clearer. We have also expanded the description of the tables to make it clear what parameters are presented and what they mean.

 

“5. The Discussion section mainly presents speculative explanations, but it is not clearly shown which experimental results support these interpretations. The authors should clearly distinguish what is directly supported by the data from hypotheses or future research directions.”

 

We have rewritten the discussion to make it clear what is our speculation and what is our basis for the assumptions we made.

 

“6.The logical linkage between objectives, methods, results, discussion, and conclusions is weak.
The manuscript would benefit from a comprehensive restructuring so that the reader can clearly understand

- What was investigated

- How it was investigated

- What was found

- Why the findings are important

 While the topic is potentially interesting, the manuscript does not currently satisfy the standards" of a scientific research paper. Therefore, Major Revision is recommended.”

 

We have added some explanations to make the manuscript clear for the reader, indcluding summarizing our goal and procedure in the introduction and elaborating more on the details throughout sections 2-3. We have also added the conclusions to summarize our finding and clearly explain what we why think they are important. We hope that the manuscript reads more like a research paper now.

 

“Comments on the Quality of English Language

The overall English language quality requires significant improvement. While the manuscript is generally understandable, there are frequent issues with sentence structure, word choice, and clarity that hinder readability. In particular, several long and complex sentences obscure the intended meaning, and technical terms are sometimes used imprecisely. Substantial language editing by a proficient English speaker or a professional editing service is strongly recommended before publication.”

 

We have re-read the manuscript and hopefully fixed the errors affecting the readability.

Round 2

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

All the comments are well addressed

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents a novel single-pass electrodialysis reversal (EDR) system that achieves high recovery (70-75%) and deep desalination (as low as 8.52 µS/cm) of river water desalination with a 50% reduction in unit cost by optimizing the design of the septum (0.35 mm thin septum) and employing an over-limit current density operation (1.75V per membrane pair). Studies have shown that the single-pass mode of operation reduces concentrate residence time to minimize the risk of fouling.However, there are still many problems in the article, and it is recommended to modify or explain：

1、The conclusion is inconsistent with the content data of the article. The conclusion said that “up to 7 V per membrane pair”, but it is stated in section 3.2 that “7 V per membrane stack (1.75V per membrane pair)”, please confirm and unify.

2、It only ran for 81 min, and its feasibility in real long-term practical applications has yet to be verified.

3、The economic estimate is not reasonable, as mentioned in the previous article, the author only ran for 81 min before fouling, but the economic estimate of membrane life reached 80,000 h. Whether dynamic depreciation should be taken into account and recalculated.

4、The paper tested only a single type of water source (river water with a conductivity of ~500 µS/cm), which does not indicate the So will it still be applicable in river water with a large amount of organic pollutants and a conductivity in the range of conductivity ~500 µS/cm?

5、The study mentions the control of differential pressure by means of a throttle valve to prevent membrane bulging, but does not provide a specific range of differential pressure, or an adjustment strategy.

6、The authors hypothesize that the spacer design promotes electro-convective eddies, but no direct evidence is provided for verification, perhaps some flow field simulations could be performed to verify this.

Comments on the Quality of English Language

Need to strengthen and improve language descriptions in articles.

Author Response

Response to Reviewer 1

Thank you for reviewing the manuscript. We believe we fixed the issues you mentioned, please find below the detailed response.

“1、The conclusion is inconsistent with the content data of the article. The conclusion said that “up to 7 V per membrane pair”, but it is stated in section 3.2 that “7 V per membrane stack (1.75V per membrane pair)”, please confirm and unify.”

These two values describe two different experiments made on two different modules. During the initial batch-mode experiments conducted in lab scale, we worked at 1.75V per membrane pair and that was the only voltage that was tried. Next, we built new ED, pilot-scale one, using which we also tried to deliberately exceed the limiting current density on some days, hence the “up to 7V per membrane pair”. We have rewritten part of the text to make it more clear which voltage was used when.

“2、It only ran for 81 min, and its feasibility in real long-term practical applications has yet to be verified.”

The 81 minutes was not the total time of experiments, but the run time of a single electrode polarity (81 min of work -> polarity reversal -> 1.5 min spec-off flush -> 81 min of work -> polarity reversal -> 1.5 min spec-off flush -> 81 min of work… - repeat for the entire day). Additionally, there were no signs of of scaling when the EDR run for multiple days at shorter work time. We have added clarification to the text to stress that values in Table 2 are daily averages of EDR working in continuous mode, not just a few minutes long batches.

“3、The economic estimate is not reasonable, as mentioned in the previous article, the author only ran for 81 min before fouling, but the economic estimate of membrane life reached 80,000 h. Whether dynamic depreciation should be taken into account and recalculated.”

As we mentioned in the previous response, the EDR was ran for far longer than 81 minutes. There were no signs of of scaling when the EDR ran at shorter work time (i.e. first row of Tab. 2 means 24 h operation with electrode polarity switch every 5 + 1.33 min throughout the day). The fouling only appeared when after a few weeks of operation we have tried to substantially increase the time between polarity reversal. While we admit few weeks of operation is still much shorter than 80 000 h, we believe our results prove long-time operation is feasible as long as polarity reversal is performed at least once per hour (see last 3 rows of Tab. 2).

“4、The paper tested only a single type of water source (river water with a conductivity of ~500 µS/cm), which does not indicate the So will it still be applicable in river water with a large amount of organic pollutants and a conductivity in the range of conductivity ~500 µS/cm?”

We agree we only tested single type of water source, but this is a very common water source in the region of the power plant that hosted the pilot study. Electrodialysis is generally less affected with organics than pressure-driven methods such as reverse osmosis. While industrial practice says EDR has TOC limit of 15 ppm (Reahl E.R. Half a Century of Desalination with Electrodialysis, Ionics Technical Paper, 2004), there are studies showing ED can handle TOC of even 70 ppm (Zhao et al, Sep Purif Technol 213 (2019) 339-347). It is highly uncommon for a natural river water to have TOC high enough to be a problem in ED. We do not expect EDR would remove organic from water, unless the said organics were ionized at the river water pH.

“5、The study mentions the control of differential pressure by means of a throttle valve to prevent membrane bulging, but does not provide a specific range of differential pressure, or an adjustment strategy.”

We have added the description of the range of differential pressures and the adjustment strategy to section 2. Materials and method.

“6、The authors hypothesize that the spacer design promotes electro-convective eddies, but no direct evidence is provided for verification, perhaps some flow field simulations could be performed to verify this.”

We agree that it just as an unconfirmed hypothesis. Thank you for suggesting flow field simulation, we currently lack the ability to perform such research, but we will certainly look into cooperating with research groups that do that. We have rewritten the mentioned part to stress that this is only our unconfirmed hypothesis.

Reviewer 2 Report

Comments and Suggestions for Authors

I find this work investigates a very interesting problem: whether a single-pass ED with different velocities in the diluate and concentrate channels for different water recoveries is better than the commonly known feed-and-bleed operation (concentration recirculation) for controlling water recoveries. The author argues that they found a way to make the single-pass ED more competitive with the “feed-and-bleed,” which seems to involve a critical “outflow throttling” and a better spacer. Their main support for this argument is that they obtained better economic analysis outcomes for their single-pass operation, on a pilot system they proposed, than a feed-and-bleed operation, on a previously reported pilot plant system. While I think this work picks up an interesting and significant problem to investigate, I find it hard to comprehend the approaches in the work and the outcome is in doubt. Major questions and concerns regarding the authors’ conclusion include:

1. It is unclear how the economic analysis was conducted.
2. It is unclear what exactly “throttling the outflow” is and what its implication is.
3. It is unclear what exactly the better spacer does.
4. It is questionable to translate parameters evaluated in their bench experiment to the pilot setup directly without any scaling justification.
5. As the authors also pointed out, the key challenge for operating a single-pass ED with a velocity difference between the diluate and concentrate channel is the pressure difference causing faster membrane wearing or destruction. However, in their economic analysis, they did not consider the effect of membranes’ lifetime in a dramatically different operation mode, yet this will be a critical factor for the overall cost. No investigation on membranes durability and lifetime in the single-pass operation makes the entire argument and conclusion implausible.
6. The authors stated their under- or over-limiting current density operations without evaluating the level of the limiting current density. How can one verify if the current is under or over the limiting current without knowing what the latter is?

More specific comments through the lines:

 

L30: I recommend that an explanation or reference be added to support the statement that “EDR is also effective for feed waters with high silica (SiO2) content.”

 

L43: This statement is inaccurate for the feed-and-bleed mode (concentration recirculation) of operation, which is a common operation in the industry as suggested by the authors as well.

 

L48-49: I recommend that references be added for the concentration recirculation mode, and maybe refer to this as the commonly known “feed-and-bleed” operation.

 

L53-60: This part is difficult for me to read. Many nomenclatures, such as “train,” “line,” “stack,” and “circuit,” were randomly used without definition. A diagram could be helpful to assist in the reading of the plain text. Grammar errors are present in some lines. I recommend that the authors improve the writing of this part to assure that a general reader may understand their point.

 

L62-63: What is the evidence or support for the “increased risk of scaling” for a longer residence time of the concentrate flow?

 

L68-70: It is hard to comprehend what “throttling the outflow of the two media” is. How is “throttling” operated exactly? What is exactly happening with “throttling the outflow”? What is the “media”?

 

L78: Same question as above for the statement that “the concentrate outflow is throttled.” What does this mean? If it is a reduction in flow rate, how will that affect the flow rate in the entire channel? It is completely unclear about this critical operation proposed here.

 

L79: What is “residence time variance”?

 

L83: What is the “waste solution”? “Reduced” compared to what?

 

Eq 1: The authors should fully define all symbols in the equation. The basic assumption underlying the validity of this equation should be provided.

 

L177-185: I find it difficult to access the economic analysis method used by the authors. Citation does not have a doi and seems to be written in a non-English language. I also cannot find citation by common searching. In this case, the authors should provide more direct access to the method they used for the economic analysis, which serves as the most critical approach in this work.

 

L195-196: What is the limiting current density level? What verifies that it is over-limiting current?

 

L198: Please provide reasoning for “the overlimiting current regime is said to potentially provide a smaller required footprint of electromembrane units.”

 

L200-201: Please check for writing typos.

 

L215: Please provide a basic explanation of “shielding the boundary layer” and “electroconvective vortices.”

 

With all these concerns, I respectfully reject the publication of this work.

Comments on the Quality of English Language

I recommend that the authors carefully proofread the manuscript. I also suggest that they present the manuscript with the assumption that the reader is not familiar with the terminology commonly used in the subject area.

Author Response

Response to Reviewer 2

Thank you for reviewing the manuscript. We believe we fixed the issues you mentioned, please find below the detailed response.

“1. It is unclear how the economic analysis was conducted.”

 

We have provided additonal explanation and equations describing the economic analysis.

 

“2. It is unclear what exactly “throttling the outflow” is and what its implication is.”

 

Throttling the outlow was performed by a valve. Constriction of flow cross-section means the pressure must be increased to maintain the volumetric flow rate, which in effect changes the pressure differences between the different points of the electrodialyzer. We have added the explanation to the Section 2. Materials and methods.

 

“3. It is unclear what exactly the better spacer does.”

 

The better spacer exhibits higher mass transfer coefficient at lower pressure drop. We have added the comparison of Power number-Sherwood number relation for our spacer and commercial spacer to show the improvement.

 

“4. It is questionable to translate parameters evaluated in their bench experiment to the pilot setup directly without any scaling justification.”

 

The parameters evaluated in bench-scale were only a starting point for designing the pilot-scale EDR. Once the pilot-scale EDR was up and running, its operation was evaluated separately.

 

“5. As the authors also pointed out, the key challenge for operating a single-pass ED with a velocity difference between the diluate and concentrate channel is the pressure difference causing faster membrane wearing or destruction. However, in their economic analysis, they did not consider the effect of membranes’ lifetime in a dramatically different operation mode, yet this will be a critical factor for the overall cost. No investigation on membranes durability and lifetime in the single-pass operation makes the entire argument and conclusion implausible.”

 

Our analysis is based on the central assumption(1) that two membranes working at the same pressure difference between diluate and concentrate should exhibit the same life time. We know what is the life time of membranes when the pressure difference is kept at 25 kPa – this is the data from the industrial-scale river water demineralization station in Stalowa Wola, which operates in feed-and-bleed configuration so the pressure difference is relatively low. Based on the assumption(1) and the industrial data we assert the life time of the membranes should be similar. Of course, we only ran the pilot for a few weeks and we agree the fact we did not observe decreased life-time of membrane is not a direct proof that the membrane would behave the same for the full 80 000 h. However, the results show that even if outflow throttling cuts the life-time of the membrane in half, our solution is still economically competitive with the commercial EDR station. We have added this discussion to the manuscript.

 

“6. The authors stated their under- or over-limiting current density operations without evaluating the level of the limiting current density. How can one verify if the current is under or over the limiting current without knowing what the latter is?”

 

We have evaluated whether we are over or under the limitting current density based on the voltage drop per membrane pair. Typically, at low feed conductivities voltage drop over ca. 2 V/membrane pair indicates over-limitting regime (Desalination 264 (2010) 268-288); in some of our experiments, we have reached 7 V/membrane pair, which clearly indicates working above the limiting current density.

L30: I recommend that an explanation or reference be added to support the statement that “EDR is also effective for feed waters with high silica (SiO2) content.”

We have added the reference statiting EDR can work with “unlimited to saturation” silica.

“L43: This statement is inaccurate for the feed-and-bleed mode (concentration recirculation) of operation, which is a common operation in the industry as suggested by the authors as well.”

We have specified that we meant “unless the feed-and-bleed mode is used”.

“L48-49: I recommend that references be added for the concentration recirculation mode, and maybe refer to this as the commonly known “feed-and-bleed” operation.”

Corrected.

“L53-60: This part is difficult for me to read. Many nomenclatures, such as “train,” “line,” “stack,” and “circuit,” were randomly used without definition. A diagram could be helpful to assist in the reading of the plain text. Grammar errors are present in some lines. I recommend that the authors improve the writing of this part to assure that a general reader may understand their point.”

We have rewritten this fragment to make it clear.

“L62-63: What is the evidence or support for the “increased risk of scaling” for a longer residence time of the concentrate flow?”

The scaling risk stems from the slow kinetics of crystallization. For example, gypsum, which is often responsible for scaling, does not precipitate immediately, the time between reaching supersaturation and observing macroscopic crystallization strongly depends on how saturated the solution is (i.e. at 140% saturation gypsum is stable for days, at 400% saturation it precipitates in minutes). If we keep the supersaturated concentrate inside the ED for too long, we risk scaling on the membrane surface. This severily limits the feed-and-bleed configuration, as some growing crystal nuclei can end up being recirculated.

“L68-70: It is hard to comprehend what “throttling the outflow of the two media” is. How is “throttling” operated exactly? What is exactly happening with “throttling the outflow”? What is the “media”?”

We have added explanation of the throttling.

“L78: Same question as above for the statement that “the concentrate outflow is throttled.” What does this mean? If it is a reduction in flow rate, how will that affect the flow rate in the entire channel? It is completely unclear about this critical operation proposed here.”

We have clarified that by throttling we mean restricting the flow cross-section at the valve but maintaining the same flow rate, which results in the increased pressure required by the pumps.

“L79: What is “residence time variance”?”

It is the variance of the residence time distribution (E(t) function generated by the stimulus-response method of RTD measurement).

“L83: What is the “waste solution”? “Reduced” compared to what?”

In this case we meant that concentrate can be more saline and have less volume compared to the case when its concentration is limited by the concentration of sparingly soluble salts.

“Eq 1: The authors should fully define all symbols in the equation. The basic assumption underlying the validity of this equation should be provided.”

We have added the missing explanation of the symbols. The underlying assumption is that there is a non-linear current density distribution along the membrane, we have added references discussing that.

“L177-185: I find it difficult to access the economic analysis method used by the authors. Citation does not have a doi and seems to be written in a non-English language. I also cannot find citation by common searching. In this case, the authors should provide more direct access to the method they used for the economic analysis, which serves as the most critical approach in this work.”

We have provided the equations we used to calculate the costs.

“L195-196: What is the limiting current density level? What verifies that it is over-limiting current?”

We have verified whether we work over or under limitting current by observing the voltage drop per membrane pair.

L198: Please provide reasoning for “the overlimiting current regime is said to potentially provide a smaller required footprint of electromembrane units.”

The smaller footprint can be a result of increasing the ion flux across the membrane, less membrane is needed to transport the same amount of ions.

“L200-201: Please check for writing typos.”

Fixed.

“L215: Please provide a basic explanation of “shielding the boundary layer” and “electroconvective vortices.”

We have added the explanation to the text.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All my questions have been addressed, and the revised manuscript can be published.

Author Response

Thank you for the review.

(No further response needed)

Reviewer 2 Report

Comments and Suggestions for Authors

I appreciate the revision and acknowledge its improvements. Remaining key concerns are the following:

I still find it impossible to reproduce the TEA described in the work. There is no clear reference to the method, no costing parameter or variable values provided, and some equations are doubtful, with unbalanced units across the "=" (e.g., CAPEX and OPEX calculations) and missing symbol definitions (e.g., η). The authors should ensure that a reader can reproduce their calculation outcomes by following the described methods. Supplementary data and files should be provided if necessary.

I still do not understand from the article how the throttling affects the flow rate in the channels and, in turn, the water recovery. If the throttling increases the flow rate in the concentrate channel, will that decrease the water recovery? How does that differ from directly controlling the inlet flow rate? Will the throttling undermine the theoretical basis of equation (1)?

Several comments from the first review were not addressed in the revision.

Author Response

Thank you fot the review. Please find below our response to the issues you pointed out:

"I still find it impossible to reproduce the TEA described in the work. There is no clear reference to the method, no costing parameter or variable values provided, and some equations are doubtful, with unbalanced units across the "=" (e.g., CAPEX and OPEX calculations) and missing symbol definitions (e.g., η). The authors should ensure that a reader can reproduce their calculation outcomes by following the described methods. Supplementary data and files should be provided if necessary."

We have added Appendix A in which we describe the OPEX/CAPEX calculations step-by-step to ensure the reader knows how we came up with our numbers. We have also revised data presented in Tab.3, as we noted some errors when we were repeating the calculations

"I still do not understand from the article how the throttling affects the flow rate in the channels and, in turn, the water recovery. If the throttling increases the flow rate in the concentrate channel, will that decrease the water recovery? How does that differ from directly controlling the inlet flow rate? Will the throttling undermine the theoretical basis of equation (1)?"

We apologise, we omitted the crucial point of the pressure regulation: the throttling is not there to affect the flow, it is to increase the pressure drop by constricting the outlet pipe. Once the flow cross-section decreases, inlet pump increases the pressure to maintain the same volumetric flow rate in the concentrate channel. This makes the concentrate side of the module slightly pressurized, but the pressure difference between two sides of the membrane is kept low.

"Several comments from the first review were not addressed in the revision."

We have re-read the first review to make sure we did not omit anything and we believe now we have adressed all the comments.