PhreeqcUsers Discussion Forum

Beginners => PHREEQC basics => Topic started by: ttgeochem on January 12, 2018, 12:31:05 AM

Title: Conceptual Understanding of PHREEQC-RM
Post by: ttgeochem on January 12, 2018, 12:31:05 AM
Hi,

I'm learning about PHREEQC-RM and reading all the documentation, with the hopes of implementing it in a 3D hydrodynamical model I have that incorporates hydro data (current, depths, etc.) and meteorological data (wind speeds, etc.).

I'm just a bit confused because everything seems to be defined in terms of a single column of N cells. For example an advection simulation has a column of N cells and most examples are in 1D. How does it translate to a rectangular grid or 3D?

How does ADVECTION or TRANSPORT get velocities to implement the movement? Or am I misunderstanding something? To add, how would I make PHREEQC-RM use the velocities from my model? The TRANSPORT documentation says "The velocity of water in each cell is determined by the length of the cell divided by the time step", what if the velocity changes dynamically?

Finally, I'm conceptually having some difficulty understanding the process. Say I have a 5x5 grid of water (25 cells) on the computer screen and I have 3D velocities and each cell has a location (x,y,z). Is it possible to transfer this to PHREEQC to simulate mixing, dispersion, and movement of chemicals and not just within each of the individual 25 cells but as chemicals move from one cell to the next (up (North) or below (South) or side-to-side (West/East) or deeper into the water)?

I think I have a understanding of the rest of the coding processes and such and my input file is mostly ready, I'm just not sure I'm fully understanding how PHREEQC 'physically' works.

Cheers.
Title: Re: Conceptual Understanding of PHREEQC-RM
Post by: dlparkhurst on January 12, 2018, 02:22:11 AM
PhreeqcRM is intended to be used as a geochemical module. The transport capabilities (ADVECTION and TRANSPORT) would not, in general be used.

The intended strategy is to calculate flow and conservative transport of component with a flow and multicomponent transport model. The transported concentrations would then be sent to PhreeqcRM to calculate geochemical reactions (reactions like EQUILIBRIUM_PHASES, KINETICS, etc), and the reacted concentrations are then sent back for another round of flow and transport. The concentrations of minerals, kinetic reactans, surface, etc are kept in PhreeqcRM, and solution concentrations are passed back and forth. So, no, PhreeqcRM is not intended to simulate "mixing, dispersion, and movement of chemicals".

You can look at the code for PHAST to see a 3D model that uses PhreeqcRM. The 3D set of cells is numbered in the "natural" sequence inherited from HST3D, but you can use any numbering (1 through n). There is a method that allows creating a mapping between the 3D natural numbering and a set of active cells (1 through nchem, where nchem <= n) for which chemistry needs to be calculated. The mapping allows eliminating inactive regions of the domain, and possibly taking advantage symmetry, where the chemical evolution is the same in two or more parts of the domain.

Title: Re: Conceptual Understanding of PHREEQC-RM
Post by: ttgeochem on January 13, 2018, 01:01:00 AM
Thanks, David, that cleared up my confusion. Three more questions, for now, PHREEQC-RM still requires a PHREEQC input file, right? I suppose this would be used for the first timestep only, as the intermediate outputs would be kept inside PHREEQC-RM (except reacted concentrations)?

I read somewhere on this forum that "PHREEQC works best with solutions that have about 1 L of water" and that solutions should be scaled accordingly. Is this the case with PHREEQC-RM as well? So then despite a simulation in our RT-simulator having several litres of water per gridcell, when brought into PHREEQC-RM, it should be normalized to 1.0 litres and then expanded when brought back to the RT-simulator?

Finally, you mention only solution concentrations are moved between the RT-simulator and PHREEQC-RM: does this include precipitates, etc.?

Cheers.
Title: Re: Conceptual Understanding of PHREEQC-RM
Post by: dlparkhurst on January 13, 2018, 01:55:15 AM
The PHREEQC input file is used to define the initial concentrations in the cells and inflowing boundary condition concentrations, but also the types of reactions and initial conditions for all of the reactants in each cell, that is the EXCHANERs, SURFACEs, KINETICs, SOLID_SOLUTIONs, EQUILIBRIUM_PHASES that are in each cell. Basically, solutions and reactants defined in the input file are distributed as needed to set the initial conditions and reactions in each of the cells.

Results calculated by PhreeqcRM are extracted through SELECTED_OUTPUT definitions and the PhreeqcRM methods that retrieve SELECTED_OUTPUT. SELECTED_OUTPUT and USER_PUNCH all you to retrieve cell solution concentrations, amounts of reactants, and many other properties.

PhreeqcRM uses one liter of porous medium by default, although you can set the representative volume to another volume. The volume of water is then defined by the porosity and the saturation, which are quantities that can be set at each iteration if necessary by PhreeqcRM methods.

You pass the intensive property concentration to PhreeqcRM, not the extensive properties of liters of solution and moles of solutes.

PhreeqcRM was designed assuming the amounts of precipitate do not affect the flow system and would could be kept in PhreeqcRM, although it is possible to extract the amounts of precipitate (SELECTED_OUTPUT and USER_PUNCH) at each iteration and use that information to refine the porosity, dispersivity to affect the flow and transport calculation.