specific surface area and minerals volume fraction with PhreeqcRT?

[cyprien]

Hi,

I am trying to couple PhreeqcRT and my transport model developed with OpenFOAM. The main objective is to have a realistic geochemistry model (kinetics, adsorption) at the pore-scale with what we call a micro-continuum model. Basically, I deal with an Eulerian grid and in cells that contain an aggregate of fluid and solid I solve Darcy and in cells that contain fluids only I solve Navier-Stokes.
 
I followed the guidelines for the coupling and so far it works well (I still have some problems with MPI, but that is not the priority today).

There are some points however that are not straightforward for me. The porous structure evolves with time and I’d like to input to PhreeqcRT at every transport time-step the specific surface area computed in OpenFOAM. Is there a simple way of doing it?

Also, how can I recover in my transport code the value of the mass fraction of every mineral computed with PhreeqcRT so I can update the cell porosity and therefore move the fluid/solid interface?

Thank you in advance for suggestions.

Cyprien

[dlparkhurst]

PhreeeqcRM may not be ideal for your particular simulations. I think it is possible to do, but not as simple or elegant as I would hope.

For SURFACEs, it is possible to link the surface area (and number of sites) to the amount of a mineral that is present. That may or may not help.

For kinetics or surface, you can redefine any item, like surface area or kinetic parameters with SURFACE_MODIFY and KINETICS_MODIFY. If you are using the parallel capabilities, it may be difficult to know which thread has which cells. I think  _MODIFY will be ignored for threads that do not have the cell in question, but I am not sure.

I think you would need to use several of the Basic functions to obtain mass fraction. The number of moles of each reactant can be retrieved relatively easily with SELECTED_OUTPUT. USER_PUNCH can be used for more complicated processing. SYS can be used to get lists of all of the equilibrium phases, kinetic reactants, surfaces, and others. The lists can be processed, for example EQUI("Calcite") would return the number of moles of calcite in the cell. PHASE_FORMULA will provide the chemical formula for a mineral. GFW("CaCO3") gives the gram formula weight of calcite. SOLN_VOL gives the volume of water. Normally, PhreeqcRM uses a cell volume of 1 liter.  You can retrieve the SELECTED_OUTPUT and USER_PUNCH information through the methods of PhreeqcRM.

So, while I think it is possible, it may require a fair amount of work to set it up generally.

[cyprien]

Hi,

Thank you for your response and advice. That sounds much more complex than what I thought at the beginning. I'll try to do my best.

Cheers,

[dlparkhurst]

Equilibrium_phases would not depend on a surface area.

To me, mass fraction does not seem any more related to surface area than moles of reactants.

Usuallly, for KINETICS, m (moles of reactant) and m0 (initial moles of reactant) are used to estimate surface area, where surface area is some function like a*m*(m/m0)^k, but you could make the RATES definition a more complicated function dependent on the moles of other minerals and kinetic reactions present.

Surface area for SURFACE is a fixed quantity by default, or proportional to the amount of a kinetic reactant or equilibrium_phase.  If the proportionality approach does not work for you, then you will need to redefine the surface area at each time step, probably with SURFACE_MODIFY definitions.

A new surface area varable could be stored for each cell, like porosity or saturation. That might work for KINETICS with a new surface area function, but SURFACEs are not really set up to take advantage of it, even if it were implemented.