Reaction Model for Anoxic Columns

[mcossio23]

Hello,

We are trying to develop a reactive model of saturated anoxic columns. These columns remove SO4, NO3 and Se(VI) via reduction; therefore, we are putting together a simple RM to try and model the leachate from these anoxic columns.

The column is 0.5 m, and has a surface area of 0.0074 and a porosity of 0.41. The flow rate is 1.2 L/day. Our current model is setting up a reactive transport for a 0.5 m column divided in 10 cells (0.05). The breakthrough curve for our Br tracer is currently matching observed results so the next step is to incorporate kinetics.

The columns are subjected to a constant influent of ~ 1200 mg/L SO4, 21 mg/L NO3 and 0.96 mg/L Se.

We are simplifying the model by only account reduction kinetics (no precipitation/dissolution or adsorption reactions) and we are using the Monod / Michaelis-Menten Kinetic with some theoretical values for k_max and k_half.

Current results from the model show that Se is indeed being removed from solution via reduction quite rapidly; however, NO3 and SO4 appear to be unaffected. I am wondering if this is due to our rate of recharge being higher that the reduction rate or whether the code was implemented properly.

This is my first time putting a reactive model together so any help would be greatly appreciated.

Cheers,
Manny

[dlparkhurst]

First, let's talk about redox. By using one of the default databases, you are implicitly assuming an equilibrium system. Therefore, redox elements will be reduced in the order of thermodynamic preference. I have attached a redox titration of your initial solution by addition of CH2O.

Looking at the graph, you can see that initially NO3 is reduced to N2, then Se(6) is reduced to Se(-2), and finally N(0) and S(6) are reduced to N(-3) and S(-2), respectively. That will be the sequence in your column assuming equilibrium.

Now, your reactions need to be changed. KINETICS ignores charge in your -formulas, so you are removing the elements, but not the charge from solution, which makes for unwanted pH and redox reactions. Assuming equilibrium redox, you might add CH2O (assuming organic matter is your reductant) for each reaction. (If you want, you can correct the stoichiometry for the different electron acceptors considering each mole of CH2O provides 4 electrons.) However, if you use CH2O for each kinetic reaction then it is clear from the graph that initially, all of the CH2O added will go to reducing NO3, and not Se or S. Alternatively, you could remove a charge balanced quantity for each reaction, but I am not sure what that would be (NaNO3?).

But now you say, "That's not what I want", and my response is two-fold (1) to treat each redox state separately is doable, but much more work, and (2) if you want to treat the redox states, separately, you could simply include one electron acceptor at a time for your fitting without having to split all the redox states so that you can treat them all individually and simultaneously.

I think I'll stop here, with a couple more comments. Example 9 shows an example of splitting redox states for Fe, and attached is a database that splits common redox states, including N and S, but not Se.