sand column evaporation simulation using EOS4 of itough_V7.1.0
I am interested in this sand column evaporation case, which is shown in this video https://www.youtube.com/watch?v=-U0vivPLf3s, simulated by STURA. I want to use ITOUGH to simulate this process, so I am using EOS4 of ITOUGH_V7.1.0 on Ubuntu Linux system, and come up with several problems. The input file and the results are attached.
As it is an evaporation case with vapor flow dominant, so that I want to close the liquid flow at the surface (i.e., OUT_0). I use the second relative permeability function (IRP=2), with the parameter value of 30, then it will get a very small relative permeability (small liquid saturation) of OUT_0, and a very small permeability. However, the flow rate through connection between AE__1 and OUT_0 is still too large. Can anyone (1) tell me why I can not use IRP=2 to control the permeability of OUT_0, and (2) help me to colse liquid flow through connection between AE__1 and OUT_0?
The problem with your model is not really related to your choice of the relative permeability function (even though I would not use the one you chose). Your choice is actually irrelevant because the gradient you specified points into, not out of element OUT_0; because of upstream weighting (see MOP(11)), this function is never used!
My key question is: How do you want to implement evaporation? There are several options; they were previously discussed in this Forum. Why did you pick EOS4 (I can guess the answer, but you would need to invoke vapor diffusion).
Once you have decided on a conceptual model, pick an appropriate EOS module and implement the conditions at the evaporation boundary and the atmosphere as they exist in the real system (no "tricks" are needed).
Additional suggestions can be found at http://tough.lbl.gov/assets/docs/support/atmo-boundary-condition.pdf. The paper by Ghezzehei et al. (Vadose Zone J., 3, 806–818, doi:10.2136/vzj2004.0806, 2004) may also be helpful.