the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Nov 2022
Quantification of In-situ Remediation of Deep Unsaturated Zone and Groundwater
Abstract. In-situ bioremediation techniques are cost-effective, environmentally friendly, and sustainable. This study examined a large-scale in-situ treatment of an unsaturated zone and a local groundwater system that were heavily contaminated with perchlorate and other co-contaminants, including nitrate, chlorate, and RDX. Principally, the upper section of the unsaturated zone was used as a bioreactor for treating the deep unsaturated zone and groundwater. The treatment was based on a cyclic process that included pumping contaminated groundwater, adding an essential electron donor, and injecting the amended water back into the top-soil, which was used as a bioreactor in the treatment process. In the shallow soil, the local bacteria reduced the perchlorate to chloride and water, and the treated water continued to displace the major pollutants from the deep part of the vadose zone, where the biological potential for contaminant degradation is low, towards the water table. The contaminated leachates were pumped back to the surface with polluted groundwater as part of the cyclic treatment process. Results show that the other co-contaminants, including nitrate, chlorate, and RDX, were removed. Water flow and reactive transport models were calibrated and validated against a time series of the water contents and bromide and perchlorate concentrations that were obtained across the unsaturated zone using the VMS. The calibrated models enabled quantifying the clean-up process and estimating the required time for full perchlorate removal. According to the model's predictions, after 700 days of continuous operation, all the perchlorate, in a total amount of 7754 kg, would be removed from the unsaturated zone. To obtain full removal, the modelling simulations suggest that the in-situ bioremediation should be implemented for an additional 200 days. Ultimately, we present a low-cost, efficient method for treating perchlorate contamination and potentially that of other pollutants in the subsurface.
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(1195 KB)
-
Supplement
(570 KB)
-
This preprint has been withdrawn.
- Preprint
(1195 KB) - Metadata XML
-
Supplement
(570 KB) - BibTeX
- EndNote
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1179', Anonymous Referee #1, 19 Dec 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-1179/egusphere-2022-1179-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Ilil Levakov, 09 May 2023
We would like to express our sincere gratitude to the reviewers for their valuable feedback. We believe that their comments and suggestions have significantly improved the quality of the manuscript. We are truly appreciative of their time, effort, and expertise. Please see our response in the attached file.
-
AC1: 'Reply on RC1', Ilil Levakov, 09 May 2023
-
RC2: 'Comment on egusphere-2022-1179', Anonymous Referee #2, 07 Apr 2023
This interesting manuscript follows a relatively large scale experiment in which contaminated water (perchlorate) is pumped, and the shallow subsurface acts as a reactor for the remediation. The presented work include experimenta work (some of it already presented, but its extent is not clear from the presentation), and a numerical model using the HYDRUS platform. Overal the manuscript is easy to follow, presents an interesting approach (even if somewhat questionable), and is of value. Most of the conclusions make sense. However. too many aspects need to be improved before the manuscript can be accepted for publication. Primarily, the authrs need to convince that their model is reliable - I was not convinced and therefore I could not evaluate the model results and the conclusions drawn by these results
1. While the idea is certainly interesting, one should wonder why use the shallow vadose zone, where it is far from being trivial to control pH and oxygen levels, instead using a controlled reactor and let cleane water percolate. After all, the water is pumped anyway. The authors should at the very least to discuss their alternative vs. more cassic pump & treat approach
2. overall the text is "slopy", especially the introduction and materials and methods - see too many examples below - and should be polished
3. acronyms should be defined, even if the authors believe they are common (they are not)
4. L19 and on This part of the abstract lacks clarity. 70 days or additional 200 days?
5. L33 There are so many "conventional methods", that accuracy s needed. The most conventional method is pump and treat, but it is not trivial to call it in-situ method
6. L45 if the final product is water, the chemical equations should contain hydrogen and its source (and the consequence of its use - pH change) should be discussed
7. L59 what further treatment? If further treatment is needed, why bother with in-situ remediation?
8. L62 by using "the contaminated site" the authors assume that the reader is familiar with the site, which is questionable
9. L85 the more appropriate term would be variably saturated zone. e.g., it is more than likely that the clay layer (or the soil just above it) gets saturated at leas occasionally
10. L119for how long?
11. L129 m^3 or m^-3? not clear
12. L129 and elsewhere the term injection is used throughout the manuscript. Is that the right term? my understanding of the system is that it is mostly gravity driven
13. L140 raise power
14. L163 is the use of first order reaction, for reaction that (as the authors claim) depends on the lack of oxygen and the availability of oxygen donor, justified? Can the authrs show that the oxygen levels, pH and ORP are in the desired ranges. One can speculate that the experiment is simply dilluting the contaminants (to be clear, I do believe that the experiment is performing as planned, but the text does not support it well neither by measurements (many of which unreliable, as the authors indcate, nor by the model that lacks the complexity and the supprting measurments
15. L167 a word about model heterogeneity would be in place. The subsurface seem to be highly heterogeneous (Fig. 1) and therefore some heterogeneity, physical (VG parameters, porosity) and chemical (reaction rates) should be considered.
16. the use of a one-dimensional model should be justified. Clay layers tend to induce lateral flow
17. L185 why atmospheric BC? there is no runoff, true, but there is also no rainfall and almost no evaporation is the site is covered by plastic, as mentioned above
18. L186 Where is the lower boundary condition? still in the vadose zone or in groundwater (and why so)?
19. L190 again. those are HYDRUS terms. What are the BCs? For the contaminant, the concentration changes over time - making the BC non-linear and state dependent. How was that taken into account?
20. L198 trial and error is fine, but still - some 15-20 parameters are involved. Any details about the process and its convergence would be nice
21. L217 it is really funny to see equations written in formal mathematical notation, and then others that make use of asterisk ... or others that include unit conversion as part of the equation
22. Fig. 3 clearly the model calibration is problematic beyond 2 m. Is that an issue? The authors mostly bame the measurements, possibly rightfully, but they do not discuss the concequence of a less-calibrated model. Are the model results and the conclusions drawn reliable?
23. L263 an alternative explanation may be that the water bypasses this region, that is beneath the clay. Here the question of 1D vs 3D comes to mnd
24. Fif. 4 for 17 m the model is really off. For 36 m the model captures the general trend, but not the details. Any reason other than easurements?
25. location of tables should be re-thought
26. Table 1 quality of the calibration, per layer, would be nice here
27. Fig. 6 I do not see this figure as being useful by any means
28. Fig 7 horizontal axes is missing
29. L391 other than nitrate (that is related in a way to the degradation of the main contaminant), the value of the other co-contaminants is not clear. For nitrate, it ill be useful if the authors can show that its degradation is actually related, and not just assumed so
Citation: https://doi.org/10.5194/egusphere-2022-1179-RC2 -
AC2: 'Reply on RC2', Ilil Levakov, 09 May 2023
We would like to express our sincere gratitude to the reviewers for their valuable feedback. We believe that their comments and suggestions have significantly improved the quality of the manuscript. We are truly appreciative of their time, effort, and expertise. Please see our response in the attached file.
-
AC2: 'Reply on RC2', Ilil Levakov, 09 May 2023
-
AC3: 'reply to the editor', Ilil Levakov, 07 Jun 2023
Dear prof. Zehe
First, we would like to express our sincere gratitude for your input and expertise in refining the manuscript. Your constructive feedback has contributed to improving the quality and impact of the work.
Our choice of using one-dimensional modeling is due to the lack of evidence or observations for two and three-dimensional processes such as lateral flow. The VMS was installed diagonally in order to capture the vertical flow in undisturbed profiles. Accordingly, all monitoring points are shifted vertically and laterally a few meters apart from each other. Nevertheless, data from these multiple 1D vertical profiles cannot provide information for such multi-dimensional processes. The implementation of a more complicated model must be backed up with more 3D geological and hydrological information, which currently is lacking.
Regarding the bromide correlation, results from 2.6m didn’t match the model prediction. Nevertheless, both peaks at 5.5 and 2.6m depth were completely missing from the observations. We assumed they were not measured due to the large gap in sampling during that period (3/2016). No other significant peaks were observed to provide information for preferential flow. We elaborated on those results in lines 307-316.
Both reviewers showed concern regarding the choice of modeling methods. Specifically, the issue of parameters’ sensitivity and optimization, due to the large number of parameters involved. Therefore, we implemented the Morris method sensitivity analysis (Morris, 1991) using the SAFE Matlab code provided by Pianosi et al. (2015). In this method, the parameters are modified one at a time and sensitivity is estimated as the partial derivative of the change in model output for a given change in a single input parameter (Perzan et al., 2021). Subsequently, the sensitive parameters were calibrated following the method presented by Perzan et al. (2021). As part of the optimization procedure, the uncertainty and optimal values are calculated according to the behavioral simulations (by achieving evaluation goals).
In addition to the model improvements, several adjustments were added to the manuscript according to the reviewers' comments such as the elaboration of different treatment approaches around the world, the addition of soil parameters to the supporting information (oxygen and pH value in the soil), and further specific clarifications.
Ultimately, simulating the water flow and reactive transport in the unsaturated zone presents significant challenges due to the high complexity of the vadose zone and the multiple variables that are required for the calculations. However, the main purpose of the model is to forecast the duration of treatment needed, and thus slight deviations from the desired outcome may be acceptable. The available data sets that were obtained in the current field experiment do not support a 3D model. Thus, implementing such a complicated model would not be beneficial.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1179', Anonymous Referee #1, 19 Dec 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-1179/egusphere-2022-1179-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Ilil Levakov, 09 May 2023
We would like to express our sincere gratitude to the reviewers for their valuable feedback. We believe that their comments and suggestions have significantly improved the quality of the manuscript. We are truly appreciative of their time, effort, and expertise. Please see our response in the attached file.
-
AC1: 'Reply on RC1', Ilil Levakov, 09 May 2023
-
RC2: 'Comment on egusphere-2022-1179', Anonymous Referee #2, 07 Apr 2023
This interesting manuscript follows a relatively large scale experiment in which contaminated water (perchlorate) is pumped, and the shallow subsurface acts as a reactor for the remediation. The presented work include experimenta work (some of it already presented, but its extent is not clear from the presentation), and a numerical model using the HYDRUS platform. Overal the manuscript is easy to follow, presents an interesting approach (even if somewhat questionable), and is of value. Most of the conclusions make sense. However. too many aspects need to be improved before the manuscript can be accepted for publication. Primarily, the authrs need to convince that their model is reliable - I was not convinced and therefore I could not evaluate the model results and the conclusions drawn by these results
1. While the idea is certainly interesting, one should wonder why use the shallow vadose zone, where it is far from being trivial to control pH and oxygen levels, instead using a controlled reactor and let cleane water percolate. After all, the water is pumped anyway. The authors should at the very least to discuss their alternative vs. more cassic pump & treat approach
2. overall the text is "slopy", especially the introduction and materials and methods - see too many examples below - and should be polished
3. acronyms should be defined, even if the authors believe they are common (they are not)
4. L19 and on This part of the abstract lacks clarity. 70 days or additional 200 days?
5. L33 There are so many "conventional methods", that accuracy s needed. The most conventional method is pump and treat, but it is not trivial to call it in-situ method
6. L45 if the final product is water, the chemical equations should contain hydrogen and its source (and the consequence of its use - pH change) should be discussed
7. L59 what further treatment? If further treatment is needed, why bother with in-situ remediation?
8. L62 by using "the contaminated site" the authors assume that the reader is familiar with the site, which is questionable
9. L85 the more appropriate term would be variably saturated zone. e.g., it is more than likely that the clay layer (or the soil just above it) gets saturated at leas occasionally
10. L119for how long?
11. L129 m^3 or m^-3? not clear
12. L129 and elsewhere the term injection is used throughout the manuscript. Is that the right term? my understanding of the system is that it is mostly gravity driven
13. L140 raise power
14. L163 is the use of first order reaction, for reaction that (as the authors claim) depends on the lack of oxygen and the availability of oxygen donor, justified? Can the authrs show that the oxygen levels, pH and ORP are in the desired ranges. One can speculate that the experiment is simply dilluting the contaminants (to be clear, I do believe that the experiment is performing as planned, but the text does not support it well neither by measurements (many of which unreliable, as the authors indcate, nor by the model that lacks the complexity and the supprting measurments
15. L167 a word about model heterogeneity would be in place. The subsurface seem to be highly heterogeneous (Fig. 1) and therefore some heterogeneity, physical (VG parameters, porosity) and chemical (reaction rates) should be considered.
16. the use of a one-dimensional model should be justified. Clay layers tend to induce lateral flow
17. L185 why atmospheric BC? there is no runoff, true, but there is also no rainfall and almost no evaporation is the site is covered by plastic, as mentioned above
18. L186 Where is the lower boundary condition? still in the vadose zone or in groundwater (and why so)?
19. L190 again. those are HYDRUS terms. What are the BCs? For the contaminant, the concentration changes over time - making the BC non-linear and state dependent. How was that taken into account?
20. L198 trial and error is fine, but still - some 15-20 parameters are involved. Any details about the process and its convergence would be nice
21. L217 it is really funny to see equations written in formal mathematical notation, and then others that make use of asterisk ... or others that include unit conversion as part of the equation
22. Fig. 3 clearly the model calibration is problematic beyond 2 m. Is that an issue? The authors mostly bame the measurements, possibly rightfully, but they do not discuss the concequence of a less-calibrated model. Are the model results and the conclusions drawn reliable?
23. L263 an alternative explanation may be that the water bypasses this region, that is beneath the clay. Here the question of 1D vs 3D comes to mnd
24. Fif. 4 for 17 m the model is really off. For 36 m the model captures the general trend, but not the details. Any reason other than easurements?
25. location of tables should be re-thought
26. Table 1 quality of the calibration, per layer, would be nice here
27. Fig. 6 I do not see this figure as being useful by any means
28. Fig 7 horizontal axes is missing
29. L391 other than nitrate (that is related in a way to the degradation of the main contaminant), the value of the other co-contaminants is not clear. For nitrate, it ill be useful if the authors can show that its degradation is actually related, and not just assumed so
Citation: https://doi.org/10.5194/egusphere-2022-1179-RC2 -
AC2: 'Reply on RC2', Ilil Levakov, 09 May 2023
We would like to express our sincere gratitude to the reviewers for their valuable feedback. We believe that their comments and suggestions have significantly improved the quality of the manuscript. We are truly appreciative of their time, effort, and expertise. Please see our response in the attached file.
-
AC2: 'Reply on RC2', Ilil Levakov, 09 May 2023
-
AC3: 'reply to the editor', Ilil Levakov, 07 Jun 2023
Dear prof. Zehe
First, we would like to express our sincere gratitude for your input and expertise in refining the manuscript. Your constructive feedback has contributed to improving the quality and impact of the work.
Our choice of using one-dimensional modeling is due to the lack of evidence or observations for two and three-dimensional processes such as lateral flow. The VMS was installed diagonally in order to capture the vertical flow in undisturbed profiles. Accordingly, all monitoring points are shifted vertically and laterally a few meters apart from each other. Nevertheless, data from these multiple 1D vertical profiles cannot provide information for such multi-dimensional processes. The implementation of a more complicated model must be backed up with more 3D geological and hydrological information, which currently is lacking.
Regarding the bromide correlation, results from 2.6m didn’t match the model prediction. Nevertheless, both peaks at 5.5 and 2.6m depth were completely missing from the observations. We assumed they were not measured due to the large gap in sampling during that period (3/2016). No other significant peaks were observed to provide information for preferential flow. We elaborated on those results in lines 307-316.
Both reviewers showed concern regarding the choice of modeling methods. Specifically, the issue of parameters’ sensitivity and optimization, due to the large number of parameters involved. Therefore, we implemented the Morris method sensitivity analysis (Morris, 1991) using the SAFE Matlab code provided by Pianosi et al. (2015). In this method, the parameters are modified one at a time and sensitivity is estimated as the partial derivative of the change in model output for a given change in a single input parameter (Perzan et al., 2021). Subsequently, the sensitive parameters were calibrated following the method presented by Perzan et al. (2021). As part of the optimization procedure, the uncertainty and optimal values are calculated according to the behavioral simulations (by achieving evaluation goals).
In addition to the model improvements, several adjustments were added to the manuscript according to the reviewers' comments such as the elaboration of different treatment approaches around the world, the addition of soil parameters to the supporting information (oxygen and pH value in the soil), and further specific clarifications.
Ultimately, simulating the water flow and reactive transport in the unsaturated zone presents significant challenges due to the high complexity of the vadose zone and the multiple variables that are required for the calculations. However, the main purpose of the model is to forecast the duration of treatment needed, and thus slight deviations from the desired outcome may be acceptable. The available data sets that were obtained in the current field experiment do not support a 3D model. Thus, implementing such a complicated model would not be beneficial.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 434 | 193 | 48 | 675 | 91 | 30 | 30 |
- HTML: 434
- PDF: 193
- XML: 48
- Total: 675
- Supplement: 91
- BibTeX: 30
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 245 | 36 |
| Germany | 2 | 76 | 11 |
| Israel | 3 | 59 | 8 |
| China | 4 | 53 | 7 |
| France | 5 | 27 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 245
Ilil Levakov
Zeev Ronen
Tuvia Turkeltaub
Ofer Dahan
This preprint has been withdrawn.
- Preprint
(1195 KB) - Metadata XML
-
Supplement
(570 KB) - BibTeX
- EndNote