Shall we try to keep the representative volume close to 1 in PHREEQRM?

[Ivan K.]

Dear colleagues,
already in several topics on this forum I've seen that PHREEQC works well when the volume of solution is set to a value not more than 1 or 2 orders different from 1. I'm incorporating PHREEQCRM into a tranport code where the cells can have quite arbitrary volumes. So, when using PhreeqcRM::SetRepresentativeVolume at the moment I set the actual cell volumes as arguments to this function. Is it the right way, or shall I try to make the actual volume of solution in every cell close to 1 litre?

Actually my question is stipulated by some problem with pe potential. Somehow pe in my calculations depends on the representative volume I set. At the same time pH is OK. I completely turned off transport, and it seems to be solely coming from the chemistry calculations. If someone has an idea what might be the cause of this problem, please, write.
In these calculations I simply initialize a solution of NaCl with pH=7 and pe=4. And on the first time step pe changes to some value depending on the RV, while it must stay constant as far as nothing has changed in the solution.

[dlparkhurst]

I think it is better to use a representative volume that is around 1, unless you have very small porosities. You just need to make sure the reactants are scaled properly to the representative volume and porosity. InitialPhreeqc2Module scales according to the values of SetUnitsX. Results from PhreeqcRM (GetConcentrations) are normally /L, so the representative volume for chemistry is not important for the transport calculation.
 
If pH is correct, I suspect the variation in pe is not important. It is possible in a poorly poised redox system that pe can very over 10 units without making any difference greater than 1e-14 in the concentration of an aqueous species. In your case, the pe probably adjusts to produce a very small concentration of either O2(aq) or H2(aq). If there is a redox couple with at least a small amount of each redox state, then the pe should be better poised.

In your system, if you want to see exactly the same pe, you could add the following to a file or string that is run through both the initial Phreeqc instance and the worker instances.

SOLUTION_SPECIES
H2O + 0.01e- = H2O-0.01
log_k -9
END