RE: Nitrogen redox

["David L Parkhurst" <dlpark@xxxxxxxx>]

> Problem:  Product water is injected into an aquifer and begins to mix,
simulated by progressive amounts of product water added to 1 liter of
groundwater.  What are the effects on manganese concentrations assuming
birnessite solubility and redox are the controlling factors?

Questions:
> 1.  If I assume N2 and NO2 are stable, as in this file, manganese
concentrations are held to 1 nM or less.

As far as I can tell, in this file, N2, NO3, and NH3 are all allowed to
react to equilibrium. This means that most of the nitrogen probably ends up
as N2. Using llnl.dat is a little overkill, I would stick with wateq4f.dat
or phreeqc.dat (augmented with birnessite), with fewer species and minerals
to deal with. I think I explained before that you may want to redefine
N2(aq) to have a much smaller K to effectively remove it from the
calculation. I've included the K's to do this in the following input file.

> However, assuming only nitrate and ammonia are stable results in >5
orders of magnitude increase in manganese concentration.  Why does this
occur and which state is more
realistic?

I don't know exactly, you need to look at the results of the simulation.
Many of the redox reactions are caused simply my mixing the two waters, so
look at the difference between the mixture and the mixture plus birnessite.
Then you need to consider

MnO2 + 2e- + 2H2O -> Mn+2 + 4OH-

At high pe, very little MnO2 dissolves, at lower pe more MnO2 dissolves and
raises the pH. Note that MnO2 is reduced, so something must be oxidized in
the process. That something is probably N(-3), so enough MnO2 reacts to
oxidize the amount of N(-3) available. It makes some difference whether it
is oxidized to N2 or NO2 or NO3, but not that much.

> 2.  Is the problem posed correctly in this file?

No. Your mixing fractions make 2 liters * 2 *3*4*5, etc, by which you are
making several thousand liters of water, which is why the program did not
converge. The following file simply mixes the waters in different
fractions.

David

solution_species
      2NO3- + 12H+ + 10 e- = N2 + 6H2O
# original log K
      log_k 207.08
# This log K makes N2 negligible
#     log_k 0

      NO3- + 2H+ + 2e- = NO2- + H2O
# original log K
      log_k 28.57
# This log K makes NO2 negligible
#     log_k 0
TITLE West Basin Simulation 1
SOLUTION 2  Barrier Product Water, December 2001 (Mg: 2001 avg)
        units   mg/L
        pH      7.7
        pe      7
        temp    25
        redox   N(5)/N(-3)
Na    15
K     1.3
Ca    19
Mg    0.07
Alkalinity  69
Cl    22
S(6)  3.3
N(5)  0.5
N(-3) 3.2
END
SOLUTION 3  Groundwater 17B, 2/26/02
units  mg/L
pH    8.25
pe    11
temp  22.6
redox O(0)/O(-2)
O(0)  2.64
Fe    0.028
Mn    0.08
Alkalinity   81
Ca    28
Mg    11
Na    49
K     3.3
Cl    53.3
S(6)  69.5
N(5)  0.31
END
EQUILIBRIUM_PHASES 2
Birnessite  0.0  1.0
END
MIX
      2     0
      3     1
END
MIX
      2     0
      3     1
USE equilibrium_phases 2
END
MIX
      2     .25
      3     .75
END
MIX
      2     .25
      3     .75
USE equilibrium_phases 2
END
MIX
      2     .5
      3     .5
END
MIX
      2     .5
      3     .5
USE equilibrium_phases 2
END
MIX
      2     .75
      3     .25
END
MIX
      2     .75
      3     .25
USE equilibrium_phases 2
END
MIX
      2     0
      3     1
END
MIX
      2     0
      3     1
USE equilibrium_phases 2
END

David Parkhurst (dlpark@xxxxxxxx)
U.S. Geological Survey
Box 25046, MS 413
Denver Federal Center
Denver, CO 80225

Project web page: https://wwwbrr.cr.usgs.gov/projects/GWC_coupled