Failure to set up a constant boundary concentration by input a large volume in the first cell
Dear Sir or Madam,
Sorry to disturb you. I set up a simple 1-D contaminant transport model using TOUGHREACT v3.0 following the 8.1 Problem 1 – Aqueous Transport with Adsorption (Linear Kd) and Decay (EOS9) in TOUGHREACT User Guide Version 2.0. I just wanted to set up the constant boundary concentration as C0, so I input a very large volume (≥10E20 m3) for the boundary grid block following the TOUGHREACT v3.0 manual. However, I found it failed and since the output concentration in the column were almost the initial concentration which means no injection went into the column.
Is there anybody can explain to me what was wrong with my model? Thank you!
Best regards,
Lang
4 replies
-
Hi Lang,
Simple fix: The injection water block at the left cannot be set to a large volume if water is injected into it. By making a block extremely large it has the effect of fixing the concentration because any small flow in or out won't affect the concentration, temperature or pressure. By injecting into a large block it is like adding a drop of fresh water into the ocean. So, either fix the concentration and pressure (with a large block volume) so there is a pressure gradient driving flow from left to right, or inject a water of some concentration into a "normal" grid block volume. The injection water composition is given in the chemical.inp file and therefore, by definition it is fixed. I reduced the volume as follows, and it worked fine:
ELEME----1----*----2----*----3----*----4----*----5----*----6----*----7----*----8
F 1 rock1 2.00E-1
F 2 57 1rock1 2.00E-1
F 60 rock1 2.00E-1cheers,
Eric -
Eric Sonnenthal
Dear Dr. Sonnenthal,
Thank you for your prompt and helpful reply.
Best regards,
Lang
-
Eric Sonnenthal
Dear Dr. Sonnenthal,
Sorry to disturb you. Thank you for your helpful reply. I just followed your instruction to reduce the hug grid block volume to a normal one and set up two model based on the 8.1 Problem in TOUGHREACT User Guide Version 2.0: Model 1 was just the example model; Model2 was based on the example model and I only increased the D0 so that τ*D0 = 0.1m2/day=10*Dnum.
When I used the analytical solution (under constant boundary condition) to fit the numerical solution, I found they did not agree for the big diffusion coefficieent (for D0 = 0, it matched well). However, when I used the another analytical solution under Robin conditions to compare with the numerical solution. I found they matched perfectly, and the analytical solution were given by van Genuchten, 1981.
This means the the 8.1Problem--Aqueous Transport with Adsorption (Linear Kd) and Decay (EOS9) in TOUGHREACT User Guide Version 2.0 should be under the Robin condition: -D*∆C/∆x + v*C=v*C0 @x=0, and the boundary concentration is not fixed.
So how can I set up a constant concentration boundary model. Thank you!
Best regards,
Lang
Reference
van Genuchten M. T. Analytical solutions for chemical transport with simultaneous adsorption, zero-order production and first-order decay[J]. Journal of Hydrology, 1981, 49(3-4): 213-233.
-