Re: Phreeqc simulations

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

> The question is, exactly what is happening during the reaction
simulation?
Are small quantities of Fe(OH)3 and CH2O being allowed to react to
equillibrium at each time step, with the quantity of reactants dictated by
the rate expressions in the RATES block?

Basically, yes, although the quantities or may not be small, depending on
the rate expression.
The reaction increases the concentrations of Fe, O, H, and C in the
dissolved state, in accordance with the stoichiometry of -formula and the
moles calculated by the rate expression.

> I assume that this must be the
case, since the KINETICS block does not contain information on the amount
of
HCO3- liberation or H+ consumption associated with the overall reaction:

Fe(OH)3 + 0.25 CH2O + 1.75H+  =>  Fe2+ + HCO3- + 2.5H2O

> Although this approach is of course quite acceptable, I'm wondering
whether
it is possible to use Phreeqc in a more generic way, so that the
consumption
of reactants and production of end-products associated with the reductive
dissolution reaction are accounted for explicitly, without having to assume
global redox equilibrium.

It is possible, but I don't think advantageous. In the above equation, how
confident are you that the products are not FeOH+, CO2, CO3-2, FeHCO3-,
FeCO3(aq), etc? How do you know that the reaction does not switch to
methanogenesis? If you want to do things explicitly, then you start having
to write kinetic reactions for a lot of things. I am still a little
concerned with the definition that adds both Fe(OH)3 and CH2O, because it
ignores the point where methanogenesis would start, so your kinetics would
have to consider all these possibilities.

So you can define [Fe3] and [Fe2] as separate elements, you can use "[]" to
allow numbers in the element name. You would have to write all the aqueous
species for Fe3 and Fe2, as well as all the phases. Then you could define
kinetic reactions that remove Fe3 and replace it with Fe2. To do everything
explicitly, you would have to do this for every redox state of every redox
element and then determine appropriate reaction kinetics for all the
transformations. It is possible, but you should probably use some other
model to do it that is designed with some of these kinetics already
established. PHREEQC might have a hard time if the resulting kinetic
expressions formed a stiff set of equations.

> The reason I ask is that I do not commonly think
about biogeochemical reaction systems in terms of global equilibrium;
rather, instinct leads me to consider everything as a
kinetically-controlled
reaction - with the exception of aqueous phase speciation and surface
complexation reactions. Hence, I would like to learn to use Phreeqc in this
more general way, which I assume is possible if the input file is
constructed properly. Thanks very much in advance for your input.

Note that only part of the system is in equilibrium. In this example, the
organic substrate is not in equilibrium with the rest of the system. I
think the key is to isolate the real nonequilibrium part and treat it
kinetically, for the rest, equilibrium is often an adequate approximation.
In general, you do not see O2 and Fe(+2) in the same water; if you look at
sediment pore waters, the redox couples are not exactly in equilibrium, but
they are not far off. And it makes sense because the bacteria that drive
the reactions can only do what is thermodynamically feasible.

David


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