PHAST: too fast migration

[daliagrig]

Dear PHREEQC/PHAST developers and users,

I'm trying to model transport of unreactive tracer through the porous media with hydraulic conductivity of 5e-10 m/s, porosity of 0.1, tortuosity of 0.0383, dispersion 0.01 m and 0.001 m (l and t), specific storage of 4.3e-7 1/m. The dimensions of the sample are 105 mm in z direction and 50 mm in x direction. Top boundary condition is specified head of 0.105 m with fixed solution with tracer concentration 1 mmol/L. Bottom boundary condition is specified head 0 m. Under such conditions interstitial velocity is 5e-9 m/s. Darcy velocity is 5e-10 m/s or about 1.6 cm/y. Therefore, I expect that the tracer should reach the bottom (10.5 cm) in about 10 years. However, looking at the results with ModelViewer already after 109 days about half of the region has tracer concentration of inflowing solution (i.e. 1 mmol/L). (To avoid the influence of diffusion, diffusivity is decreased to 1e-12 m2/s.)

Could you suggest the reason why the migration is so fast? (The model files are attached)

Thank you.

[dlparkhurst]

The Darcy flux is 5e-10 m/s, but the pore water velocity is 5e-9 m/s (Darcy flux divided by porosity).  5e-9*3.15e7 (s per y) = 0.16 m. So, advection has moved a parcel of water through the entire column in less than a year.

Note that tortuosity only affects the diffusion part of the calculation, which will be negligible given the advection and dispersion in your column.

Also, if your system is 1D, you should define both X and Y directions to have only 2 cells, and set the chemistry dimensions to just Z (GRID). The calculation will run much faster.