Two conceptual approaches about unsaturated model

[Jeonghwan Hwang]

Hello, I have a question for concept about unsaturated transport model.

I consider two ways to conduct the model.
One is fixed water volume and change the total porous volume (1), and the other is fixed total volume and change the water volume (2).
I don't consider the gas phases, so I just consider about the solid-water interaction.
The model assumed that the mineral surface is sufficiently covered with water even at low water contents condition.

I think that the concept 2 is better approach because when I fix the total volume, I just modify the water volume only.
When I use the concept 1, I should modify the amounts of every minerals at each step because the unit of mineral in equilibrium_phases is 'mol'.

Are there any additional considerations I should consider when selecting these two methods?
Or recommend a more suitable concept, thank you.
In addition, I wondered that the concept 1 might be more unstable in terms of model stability.

Any advice would be appreciated.
Sincerely,

Jeonghwan Hwang

[dlparkhurst]

Concept 2 makes more sense to me. It is the way most unsaturated zone models work. You can look at the USGS model VS2DRT by Rick Healy and others. An interface is available and you can look at the code. It uses PhreeqcRM coupled to Rick's 2D heat and solute transport model.

Unfortunately we lost Rick in a freak bicycle accident this year. He will be missed.

[Jeonghwan Hwang]

First of all, rest in peace, Rick Healy.

And, thanks for your recommendation on the modeling concept.

I'm going to refer to Rick's paper (VS2DRTI: Simulating Heat and Reactive SoluteTransport in Variably Saturated Porous Media, Groundwater) and try to use PHREEQCRM.

Thank you
Sincerely,

Jeonghwan Hwang

[djacques]

Dear
an alternative for running unsaturated (transient) flow and transport models using PHREEQC is the HYDRUS1D(2D) codes.
See - https://www.pc-progress.com/en/Default.aspx?hydrus-1d for a free version (with older interface)
or https://www.pc-progress.com/en/Default.aspx?hydrus for the most recent user interface (not free).

These models also use the second approach,

Diederik