uranium bioremediation in situ experiment

[Juliannie_Cerda]

Hello,

I am a graduate student working on a geochemical modeling project involving a laboratory column experiment simulating bioremediation of uranium-contaminated groundwater. I am using PHREEQC with the minteq.v4.dat database as the groundwater is also contaminated with high levels of arsenic and zinc.

Problem context:
The experiment consists of introducing anaerobic conditions (via molasses addition) to promote reducing conditions over a 7-day period. The goal is to evaluate changes in aqueous chemistry between influent and effluent.

Observed trends include:

* pH increase (6.3 → ~7.5)
* Alkalinity increase (115 → 355 mg/L as HCO₃⁻)
* Significant sulfate reduction (163 → 45 mg/L)
* Strong removal of Zn and As
* Nearly complete removal of uranium (0.25 → 0.0005 mg/L)

Modeling approach:
I am currently using a lumped REACTION approach to represent:

Nitrate and sulfate reduction
Alkalinity generation
Minor Ca/Mg increases
Reducing conditions (via pe adjustment)

I have also tested inclusion of:

equilibrium phases (UO₂(am), ZnS(am))
surface complexation (HFO)
carbonate system controls

Issue:
Despite achieving strongly reducing conditions (pe ≈ -7 to -8) and reasonable pH (~7–7.5), uranium remains largely in solution in the model and does not approach the observed effluent concentration.

The system appears to be dominated by uranyl–carbonate complexation, preventing precipitation of UO₂(am), even under reducing conditions.

Questions:

1. In PHREEQC equilibrium modeling, how can U(VI) reduction to U(IV) be effectively represented without using full kinetic formulations?
2. Is it appropriate to explicitly define uranium redox transformations or phases (beyond UO₂(am)) when using *minteq.v4.dat*?
3. Would a kinetic approach (e.g., KINETICS block for U reduction) be necessary to reproduce observed uranium removal?
4. Are there recommended strategies to limit carbonate stabilization effects in such systems without artificially constraining alkalinity?
5. Is a calibrated net-reaction approach (direct removal terms) acceptable practice in this type of modeling scenario?

Any guidance on best practices for representing uranium bioreduction in PHREEQC would be greatly appreciated.

Thank you for your time and assistance.

Best regards,
Juliannie

[dlparkhurst]

If you react CH2O, representing molasses while allowing uraninite to precipitate, you will remove uranium.

Code: [Select]

USER_PRINT
10 PRINT "0.25 mg/L equals", 0.25/GFW("U"), "mmol/L"

SOLUTION
-units mg/L
pH 6.3
pe 10 O2(g) -5
U 0.25
END
USE solution 1
REACTION
CH2O 1
0.000001 in 10 steps
EQUILIBRIUM_PHASES 1
    Uraninite 0 0
USER_GRAPH 1
    -headings               rxn U(6) U(4)
    -axis_titles            "CH2O added, moles" "log Mol/kgw" ""
    -initial_solutions      false
    -connect_simulations    true
    -plot_concentration_vs  x
  -start
10 GRAPH_X rxn
20 GRAPH_Y log10(TOT("U(6)")), log10(TOT("U(4)"))
  -end
    -active                 true
END

The uranium reaction is probably not the pH-determining reaction, so, you must consider what other reactions are occurring and how you want to model them. An inverse model for the major ions could help identify pertinent reactions, but don't spend more than a day on inverse modeling.