Solute transport as function of soil water content

[andrew_m]

I'm attempting to model the transport of a surfactant solute in a soil during wetting. I believe that as the soil dries out, the solute retreats into pores and then when the soil becomes wet again, the solute starts to diffuse back out of the pores in a nonequilibrium process. However, I am unable to recreate this scenario in Hydrus 1D because I am unable to control the solute behavior as a function of soil water content. Is there an approach in Hydrus 1D that would make this possible? Thanks!

[Jirka]

I guess you will have to use some of the nonequilibrium models implemented into HYDRUS, such as the MIM dual-porosity models. J.

[andrew_m]

Thanks so much for replying so quickly!!

Thanks also for the advice. I have tried the MIM approach previously, but the solute seems to transfer from the immobile to the mobile phase at the same rate (Alpha) regardless of Theta. I've attached images that show an example where a very dry soil is wetted from the bottom and the solute, which is initially entirely in the immobile phase, transfers to the mobile phase at a constant rate regardless of Theta. In this example ThImob is 0.02 and the initial Theta is 0.025. Do you think there is a different approach that would be useful?

[andrew_m]

Oops, here is the immobile phase concentration profile for the example

Immobile phase concentration.JPG (47.92 KiB) Viewed 1128 times

[Jirka]

You should study the mathematical description of the model. The mass transfer is driven by a concentration gradient, i.e., is equal to Gamma_s=alpha*(c_m-c_im), see eq. 3.32. While the water content of neither domain is directly involved in the mass transfer of solute between the two domains (which is driven by a concentration gradient), the water content of both domains should clearly affect the overall process, since the increase or decrease of concentration in either phase will depend on water contents. If you remove Gamma_s from a small domain, the concentration will change much faster than if you remove Gamma_s from a large domain, and vice versa for addition. J.

[andrew_m]

Thank you again for all the help. It has been fascinating to learn more about Hydrus.

However, I'm still struggling to model my experiment. I packed a very dry soil, with a previously sorbed chemical, into a vertically-oriented column and pumped water up through the column. I would like to model solute desorption as a nonequilibrium process (ideally chemical nonequilibrium), where the solute desorbs only after the wetting front moves past. Below is an image of the experimental progression.

I believe what I'm asking is: is it possible to have no desorption until theta changes from the initial value? Also, are there any examples of similar systems being modeled?

Thanks

Slide1.jpeg (57.18 KiB) Viewed 1004 times

[Jirka]

Note that for non-equilibrium solute transport model, you need to specify the initial condition for the liquid phase concentration and for the kinetically sorbed concentration (concentration on Type-2 sites). While the former one is an equilibrium concentration and thus the equilibrium sorbed concentration will be assigned automatically instantaneously as well, the latter one is kinetic and thus the release of solute from these sites will depend on the mass transfer coefficient, and indirectly on water content. J.

[andrew_m]

Thanks again.

So, since the water content is always nonzero, the mass transfer coefficient is also always nonzero? Is it possible to change that? Can I set a minimum water content for mass transfer to occur?

I understand that sounds odd and unrealistic, but since my soil has had a long time to dry and equilibrate, I don't believe significant mass transfer is occurring prior to the arrival of the wetting front.

[andrew_m]

Sorry, the second sentence should read "So, since the water content is always nonzero, mass transfer is also always nonzero?"

[Jirka]

No. The mass transfer is nonzero only when there is non-equilibrium, i.e., when the liquid phase is not in equilibrium with the solid phase (see eq. 3.9). When the soil is dry, there is only very little solute mass in the liquid phase (at equilibrium) and the mass transfer can start only once water content increases and present solute dilutes, introducing nonequilibrium.

You can make any changes you want in the code, which is posted on the HYDRUS website.

J.

[andrew_m]

Thanks, that makes sense. At lower water contents, as the solute desorbs the liquid phase solute concentration quickly rises, causing the system to approach equilibrium and the mass transfer rate decreases.

That said, I'm still incapable of reproducing my experimental system. In my system, at t=0, solute-free water enters a dry soil from the bottom and a wetting front propagates upwards. I would like the solute to start desorbing from Type 2 sites only after the wetting front passes by. Is that possible in HYDRUS?

[Jirka]

That's exactly what happens. There is no mass transfer if you have initially an equilibrium. Once water content increases, it dilutes concentration and mass transfer starts. J.