the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Nov 2025
Multi-component reactive transport in near-saturated deformable porous media
Abstract. This study develops a hydro–mechanical–chemical (HMC) framework for simulating reactive solute migration in near-saturated, deformable porous media. The model couples the pore-water mass balance, force equilibrium, and advection–dispersion equations, and further incorporates a flexible geochemical reaction module to address both single-reaction and multi-component, multi-mineral systems. Numerical results indicate that deformation, mechanical loading, saturation, and mineral reactions jointly control the distribution and evolution of the solute. Compression and stronger mechanical loads accelerate solute transport in the early stage but later hinder migration as the pore structure tightens. Moreover, reduced saturation promotes concentration build-up by enhancing advective transport and limiting the ability of the aqueous phase to dilute accumulated solutes. The framework improves the predictive capability for long-term plume behaviour and mineral alteration in reactive porous systems where mechanical, hydraulic, and geochemical processes interact.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2103 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2103 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2025-5497', Giacomo Medici, 19 Dec 2025
- AC1: 'Reply on CC1', Dong-Sheng Jeng, 24 Jan 2026
-
CC2: 'Scope and validity of the near-saturated formulation need to be pinned down', Hannah Menke, 28 Dec 2025
Right now, near-saturated risks sounding like a convenience assumption rather than a controlled approximation. You explore S_r 0.80–1.0, but (as written) saturation appears prescribed rather than evolving, which is a big modeling decision for landfill liners and many barrier systems. Please add a short model scope paragraph early (end of Introduction or start of Methods) stating explicitly: (1) saturation is constant in time (if that’s the case), (2) when this is reasonable (e.g., quasi-saturated low gas mobility, slow infiltration variability relative to consolidation time, etc.), (3) when it is not (drying fronts, strong unsaturated flow, gas phase transport). Consider adding a brief comparison to Richards/unsaturated THMC literature (even if you don’t implement it). And be explicit whether near-saturated here means “two-phase but simplified compressibility” vs “single-phase with effective compressibility.”
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC2 -
CC8: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
Publisher’s note: this comment is a copy of AC3 and its content was therefore removed on 26 January 2026.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC8 - AC3: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
-
CC8: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
-
CC3: 'Validation is currently insufficient for a framework paper', Hannah Menke, 28 Dec 2025
You have two verification benchmarks, but the paper itself acknowledges no lab/field quantitative validation. While this isnt a fatal flaw, you do need to strengthen credibility with additional numerical evidence. Consider adding (1) Grid/time-step sensitivity for one representative case (even a coarse/medium/fine study + one plot at the monitoring point). (2) Splitting / sub-stepping sensitivity: show that results don’t materially change when the chemistry sub-step size changes within a window (or quantify the trade-off). Your own text highlights synchronisation issues, so you should demonstrate control. (3) For benchmark #2, quantify mismatch: e.g., L2 error vs time, and show it decreases with refinement or explain the irreducible discrepancy (boundary formulation differences are mentioned but not demonstrated).
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC3 - AC2: 'Reply on CC3', Dong-Sheng Jeng, 24 Jan 2026
-
CC4: 'More information needed for reproducibility', Hannah Menke, 28 Dec 2025
At minimum, include: (1) The PHREEQC database used, the key input blocks (KINETICS/EQUILIBRIUM_PHASES/etc.), and where they are provided (supplement/Git repo). (2) OpenFOAM discretisation choices for transport (schemes, limiters), linear solvers/preconditioners, coupling tolerances, and convergence criteria (you mention a prescribed tolerance but not what it is). (3) Use a public repository plus archived case files. Without this, the ChemWindow contribution especially will be hard to credit.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC4 - AC4: 'Reply on CC4', Dong-Sheng Jeng, 24 Jan 2026
-
CC5: 'More information needed for reproducibility', Hannah Menke, 28 Dec 2025
At minimum, include: (1) The PHREEQC database used, the key input blocks (KINETICS/EQUILIBRIUM_PHASES/etc.), and where they are provided (supplement/Git repo). (2) OpenFOAM discretisation choices for transport (schemes, limiters), linear solvers/preconditioners, coupling tolerances, and convergence criteria (you mention a prescribed tolerance but not what it is). (3) Use a public repository plus archived case files. Without this, the ChemWindow contribution especially will be hard to credit.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC5 - AC5: 'Reply on CC5', Dong-Sheng Jeng, 24 Jan 2026
-
CC6: 'Internal consistency of the transport and chemistry modelling choices', Hannah Menke, 28 Dec 2025
I have some concerns about hybrid decisions. (1) You use element-based ADE transport, with a weighted molecular diffusion coefficient built from prescribed species fractions, while noting that fully coupled PhreeqcRM could update speciation and thereby D_m dynamically, but you don't do that in this study. (2) You also include a linear adsorption K_d term in the ADE, while stating that more detailed sorption could be handled by PHREEQC reaction terms. Please add a why this choice paragraph: is the goal robustness and speed? If so, say that explicitly and acknowledge the trade-off. Specify whether K_d is used in the case studies and, if yes, whether PHREEQC includes any overlapping sorption reactions.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC6 - AC6: 'Reply on CC6', Dong-Sheng Jeng, 24 Jan 2026
-
CC7: 'The case studies are plausible, but are they physics appropriate?', Hannah Menke, 28 Dec 2025
You model landfill-liner-like consolidation with chemistry. There are some questions with (1) the mechanical model (elastic vs elasto-plastic/creep typical in clays), (2) the lack of reaction-induced porosity/permeability evolution (I know you already acknowledge this which is good, but it reduces interpretability of long-term plume changes), (3) boundary conditions (drainage, injection as fluxes at multiple points) and whether they represent a realistic exposure. Consider adding a short justification for the constitutive law used and explicitly call the case studies “demonstration problems” rather than predictive landfill designs. Add one sentence in the Conclusions about what would change when property feedbacks are included (directional expectation is fine, but label it as expectation).
Also some minor comments below:
(1) In benchmark #2 you say early-time deviations come from transport formulation and boundary representations; please add one sentence indicating which boundary treatment differs (Dirichlet/Neumann/mixed; inlet formulation), and whether refinement improves it.
(2) Consistently distinguish “hydraulic conductivity K” vs “permeability k” (the symbol list emphasises K).
(3) Injection strengths are given as fluxes at J1–J3; ensure it’s always explicit whether that is imposed as boundary flux (mol/(m²·s)) and over what area.
(4) Computational performance section: helpful, but reviewers will ask for scaling beyond 4 MPI ranks and/or mesh size dependence. Even a small strong-scaling plot (1,2,4,8,16) on one case would help.
(5) Consider moving the limitations to earlier in the discussion.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC7 - AC7: 'Reply on CC7', Dong-Sheng Jeng, 24 Jan 2026
-
RC1: 'Reproducibility', Anonymous Referee #1, 06 Jan 2026
E.g. validation : Line 280 - information regarding mesh, parameters, boundary conditions and similar are not reported. This circmstance makes the reproducibility of the results a major concern.
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC1 - AC8: 'Reply on RC1', Dong-Sheng Jeng, 24 Jan 2026
-
RC2: 'Use of saturation', Anonymous Referee #1, 06 Jan 2026
please better clarify how saturation can remain constant upon loading and compaction.
why wasn't the water content considered instead?
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC2 - AC9: 'Reply on RC2', Dong-Sheng Jeng, 24 Jan 2026
-
RC3: 'Comment on egusphere-2025-5497', Anonymous Referee #2, 23 Jan 2026
The work addresses a well-recognized gap in reactive transport modelling: the integration of mechanical deformation with multicomponent geochemistry. The manuscript is generally well written and logically organized, and it fits well within the scope of coupled processes in hydrology and earth sciences.
- It would be helpful to more explicitly articulate what new capabilities are enabled here beyond coupling OpenFOAM with PhreeqcRM. For example, a short comparison paragraph contrasting this framework with TOUGHREACT or COMSOL approaches would clarify this.
- The assumption of constant saturation (which may not be true in vadose zones) is central to several conclusions. I wonder what the physical interpretation of “near-saturated” is and how sensitive the results are to small variations in saturation.
- The manuscript excludes reaction-induced changes in porosity and permeability. The authors should more clearly discuss how this assumption may bias long-term predictions.
- The treatment of molecular diffusion uses a weighted-average diffusion coefficient. I wonder how these weights are selected in practice.
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC3 - AC10: 'Reply on RC3', Dong-Sheng Jeng, 24 Jan 2026
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2025-5497', Giacomo Medici, 19 Dec 2025
General comments
Good research on contaminant transport in porous media that needs some improvement and further detail. See my specific comments, they will improve the potential new version of the manuscript.
Specific comments
Line 23….“has been the most widely applied framework in the field”. Last words not backed up by references. Add these papers that incorporate discussion on evidence of the Fick’s law in the field:
- Agbotui, P.Y., Firouzbehi, F., Medici G. 2025. Review of effective porosity in sandstone aquifers: insights for representation of contaminant transport. Sustainability 17, no. 14 (2025): 6469.
- Parker, B.L., Cherry, J.A., Wanner, P. 2022. Determining effective diffusion coefficients of chlorohydrocarbons in natural clays: unique results from highly resolved controlled release field experiments. Journal of Contaminant Hydrology 250, 104075.
Line 30. This sentence does not work as it is. Before “purely physical transport” and then you involve geochemistry. Please, revise the structure.
Line 76. You need to disclose the general aim of the research at the end of the introduction.
Line 76. You need to describe the specific objectives of your research by using numbers (e.g., i, ii, and iii) at the end of the introduction.
Line 199. Insert reference to Parkhurst for PHREEQC.
Line 250. This one looks to me a figure, not a table or an equation. Am I correct?
Line 700. More detail on the acknowledgement. Funding bodies are unclear.
Figures and tables
Figure 3. Make the image larger.
Figure 5. Same here, make the figure larger.
Two figures 11. This is not ok.
Figure 9. Make the figure larger also here.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC1 - AC1: 'Reply on CC1', Dong-Sheng Jeng, 24 Jan 2026
-
CC2: 'Scope and validity of the near-saturated formulation need to be pinned down', Hannah Menke, 28 Dec 2025
Right now, near-saturated risks sounding like a convenience assumption rather than a controlled approximation. You explore S_r 0.80–1.0, but (as written) saturation appears prescribed rather than evolving, which is a big modeling decision for landfill liners and many barrier systems. Please add a short model scope paragraph early (end of Introduction or start of Methods) stating explicitly: (1) saturation is constant in time (if that’s the case), (2) when this is reasonable (e.g., quasi-saturated low gas mobility, slow infiltration variability relative to consolidation time, etc.), (3) when it is not (drying fronts, strong unsaturated flow, gas phase transport). Consider adding a brief comparison to Richards/unsaturated THMC literature (even if you don’t implement it). And be explicit whether near-saturated here means “two-phase but simplified compressibility” vs “single-phase with effective compressibility.”
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC2 -
CC8: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
Publisher’s note: this comment is a copy of AC3 and its content was therefore removed on 26 January 2026.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC8 - AC3: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
-
CC8: 'Reply on CC2', Dong-Sheng Jeng, 24 Jan 2026
-
CC3: 'Validation is currently insufficient for a framework paper', Hannah Menke, 28 Dec 2025
You have two verification benchmarks, but the paper itself acknowledges no lab/field quantitative validation. While this isnt a fatal flaw, you do need to strengthen credibility with additional numerical evidence. Consider adding (1) Grid/time-step sensitivity for one representative case (even a coarse/medium/fine study + one plot at the monitoring point). (2) Splitting / sub-stepping sensitivity: show that results don’t materially change when the chemistry sub-step size changes within a window (or quantify the trade-off). Your own text highlights synchronisation issues, so you should demonstrate control. (3) For benchmark #2, quantify mismatch: e.g., L2 error vs time, and show it decreases with refinement or explain the irreducible discrepancy (boundary formulation differences are mentioned but not demonstrated).
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC3 - AC2: 'Reply on CC3', Dong-Sheng Jeng, 24 Jan 2026
-
CC4: 'More information needed for reproducibility', Hannah Menke, 28 Dec 2025
At minimum, include: (1) The PHREEQC database used, the key input blocks (KINETICS/EQUILIBRIUM_PHASES/etc.), and where they are provided (supplement/Git repo). (2) OpenFOAM discretisation choices for transport (schemes, limiters), linear solvers/preconditioners, coupling tolerances, and convergence criteria (you mention a prescribed tolerance but not what it is). (3) Use a public repository plus archived case files. Without this, the ChemWindow contribution especially will be hard to credit.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC4 - AC4: 'Reply on CC4', Dong-Sheng Jeng, 24 Jan 2026
-
CC5: 'More information needed for reproducibility', Hannah Menke, 28 Dec 2025
At minimum, include: (1) The PHREEQC database used, the key input blocks (KINETICS/EQUILIBRIUM_PHASES/etc.), and where they are provided (supplement/Git repo). (2) OpenFOAM discretisation choices for transport (schemes, limiters), linear solvers/preconditioners, coupling tolerances, and convergence criteria (you mention a prescribed tolerance but not what it is). (3) Use a public repository plus archived case files. Without this, the ChemWindow contribution especially will be hard to credit.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC5 - AC5: 'Reply on CC5', Dong-Sheng Jeng, 24 Jan 2026
-
CC6: 'Internal consistency of the transport and chemistry modelling choices', Hannah Menke, 28 Dec 2025
I have some concerns about hybrid decisions. (1) You use element-based ADE transport, with a weighted molecular diffusion coefficient built from prescribed species fractions, while noting that fully coupled PhreeqcRM could update speciation and thereby D_m dynamically, but you don't do that in this study. (2) You also include a linear adsorption K_d term in the ADE, while stating that more detailed sorption could be handled by PHREEQC reaction terms. Please add a why this choice paragraph: is the goal robustness and speed? If so, say that explicitly and acknowledge the trade-off. Specify whether K_d is used in the case studies and, if yes, whether PHREEQC includes any overlapping sorption reactions.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC6 - AC6: 'Reply on CC6', Dong-Sheng Jeng, 24 Jan 2026
-
CC7: 'The case studies are plausible, but are they physics appropriate?', Hannah Menke, 28 Dec 2025
You model landfill-liner-like consolidation with chemistry. There are some questions with (1) the mechanical model (elastic vs elasto-plastic/creep typical in clays), (2) the lack of reaction-induced porosity/permeability evolution (I know you already acknowledge this which is good, but it reduces interpretability of long-term plume changes), (3) boundary conditions (drainage, injection as fluxes at multiple points) and whether they represent a realistic exposure. Consider adding a short justification for the constitutive law used and explicitly call the case studies “demonstration problems” rather than predictive landfill designs. Add one sentence in the Conclusions about what would change when property feedbacks are included (directional expectation is fine, but label it as expectation).
Also some minor comments below:
(1) In benchmark #2 you say early-time deviations come from transport formulation and boundary representations; please add one sentence indicating which boundary treatment differs (Dirichlet/Neumann/mixed; inlet formulation), and whether refinement improves it.
(2) Consistently distinguish “hydraulic conductivity K” vs “permeability k” (the symbol list emphasises K).
(3) Injection strengths are given as fluxes at J1–J3; ensure it’s always explicit whether that is imposed as boundary flux (mol/(m²·s)) and over what area.
(4) Computational performance section: helpful, but reviewers will ask for scaling beyond 4 MPI ranks and/or mesh size dependence. Even a small strong-scaling plot (1,2,4,8,16) on one case would help.
(5) Consider moving the limitations to earlier in the discussion.
Citation: https://doi.org/10.5194/egusphere-2025-5497-CC7 - AC7: 'Reply on CC7', Dong-Sheng Jeng, 24 Jan 2026
-
RC1: 'Reproducibility', Anonymous Referee #1, 06 Jan 2026
E.g. validation : Line 280 - information regarding mesh, parameters, boundary conditions and similar are not reported. This circmstance makes the reproducibility of the results a major concern.
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC1 - AC8: 'Reply on RC1', Dong-Sheng Jeng, 24 Jan 2026
-
RC2: 'Use of saturation', Anonymous Referee #1, 06 Jan 2026
please better clarify how saturation can remain constant upon loading and compaction.
why wasn't the water content considered instead?
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC2 - AC9: 'Reply on RC2', Dong-Sheng Jeng, 24 Jan 2026
-
RC3: 'Comment on egusphere-2025-5497', Anonymous Referee #2, 23 Jan 2026
The work addresses a well-recognized gap in reactive transport modelling: the integration of mechanical deformation with multicomponent geochemistry. The manuscript is generally well written and logically organized, and it fits well within the scope of coupled processes in hydrology and earth sciences.
- It would be helpful to more explicitly articulate what new capabilities are enabled here beyond coupling OpenFOAM with PhreeqcRM. For example, a short comparison paragraph contrasting this framework with TOUGHREACT or COMSOL approaches would clarify this.
- The assumption of constant saturation (which may not be true in vadose zones) is central to several conclusions. I wonder what the physical interpretation of “near-saturated” is and how sensitive the results are to small variations in saturation.
- The manuscript excludes reaction-induced changes in porosity and permeability. The authors should more clearly discuss how this assumption may bias long-term predictions.
- The treatment of molecular diffusion uses a weighted-average diffusion coefficient. I wonder how these weights are selected in practice.
Citation: https://doi.org/10.5194/egusphere-2025-5497-RC3 - AC10: 'Reply on RC3', Dong-Sheng Jeng, 24 Jan 2026
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 522 | 272 | 81 | 875 | 130 | 255 |
- HTML: 522
- PDF: 272
- XML: 81
- Total: 875
- BibTeX: 130
- EndNote: 255
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Bolin Wang
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2103 KB) - Metadata XML
General comments
Good research on contaminant transport in porous media that needs some improvement and further detail. See my specific comments, they will improve the potential new version of the manuscript.
Specific comments
Line 23….“has been the most widely applied framework in the field”. Last words not backed up by references. Add these papers that incorporate discussion on evidence of the Fick’s law in the field:
- Agbotui, P.Y., Firouzbehi, F., Medici G. 2025. Review of effective porosity in sandstone aquifers: insights for representation of contaminant transport. Sustainability 17, no. 14 (2025): 6469.
- Parker, B.L., Cherry, J.A., Wanner, P. 2022. Determining effective diffusion coefficients of chlorohydrocarbons in natural clays: unique results from highly resolved controlled release field experiments. Journal of Contaminant Hydrology 250, 104075.
Line 30. This sentence does not work as it is. Before “purely physical transport” and then you involve geochemistry. Please, revise the structure.
Line 76. You need to disclose the general aim of the research at the end of the introduction.
Line 76. You need to describe the specific objectives of your research by using numbers (e.g., i, ii, and iii) at the end of the introduction.
Line 199. Insert reference to Parkhurst for PHREEQC.
Line 250. This one looks to me a figure, not a table or an equation. Am I correct?
Line 700. More detail on the acknowledgement. Funding bodies are unclear.
Figures and tables
Figure 3. Make the image larger.
Figure 5. Same here, make the figure larger.
Two figures 11. This is not ok.
Figure 9. Make the figure larger also here.