TOUGH2 Sample Problem
Hello,
This is Mo. I just started learning TOUGH2 and iTOUGH2. I am working on Problem No.4 (Five-Spot Geothermal Production /Injection). I am trying to modify the problem to my own problem. I have a question.
1) For the computational grid generation, I was wondering how they created the grid blocks. (36 blocks). What preprocessor program they used?) I looked at the grids carefully and could figure out the details of each block and the dimensions and I can replicate manually but I want to know how I can generate those using an automated algorithm? Should I modify the (meshm.f) subroutine? I appreciate if someone explains this to me.
Thanks a lot for your time and help!
Mo
11 replies
-
Mo,
Frankly, I do not know how Karsten Pruess develop this particular primary mesh (Karsten enjoys his retirement, but I'll ask him next time I see him). I suspect he did it by hand. I'm not aware of a pre-processor that would generate this specific mesh, but "general" mesh generators (such as AMESH and WinGridder) could be used to create a mesh that represents the symmetric five-spot geometry (I'm sure, however, that the resulting mesh would look very different from that used in Problem No. 4).
In case you decide to write such a pre-processor (either as part of meshm.f, or as a stand-alone program), please consider making it available to the TOUGH2 community by submitting it to the Free Software web site.
Stefan
-
Stefan,
Thanks for your response. I ended up using his mesh for my problem. However I could understand most of his mesh values and figured it out. But I did not understand one part of the mesh and I think it is general case and I appreciate if you let me know the answer.
For the CONNE parts:
I do not understand the D1 and D2 distances. In the TOUGH2 manual it says the D1 is distance (m) from first and second element respectively to their common interface. What does it mean?
For example: if I have two cubic grids next to each other and each side let say is 5 m. The D1 is 2.5 m ? and D2 is 2.5 m?
I really appreciate if you explain to me this. I am a beginer..
Thanks
Mo
-
Mo,
Not sure I understand the issue (the example you give is correct). Please see Figure B.1 in the manual, which shows D1 and D2 (Dm and Dn, respectively).
Note that in fracture-matrix systems (and other special cases, e.g., connections between well and rock elements), you should set the nodal distance from the fracture (well) element to exactly zero, so that the matrix (rock) permeability is used for the corresponding connections, overwriting "upstream weighting" (otherwise, flux from the fracture to the matrix would be overestimated). This is automatically done when running MINC - check the MINC file in Problem No. 4!
Stefan
-
Thanks Stefan for the response.
-
Hello Stefan,
This is Mo again. I hope you are fine. I have another question. I am trying to model a five spot tracer test. I saw your name on one of Pruess papers (MODELING OF PHASE-PARTITIONING TRACERS IN FRACTURED RESERVOIRS, 2000) .
I have a stupid question. I am using EOS1 which has water and tracer components. I am trying to define sink and sources in the input file. I am having hard time to define the tracer injection data. Let say I want to simulate upto 150 days and the tracer injection was for 8 hours of 100 ppm. Could you please let me know how I can define the tracer component injection data in TOUGH?
Let say the injection rate was held steady at 16 kg/s.
I just want to see how I can define this injection time in the table format for TOUGH2.
I deeply appreciate your time.
Thanks
Mo
-
Mo,
Please check the description of the GENER block in the manual, with LTAB>0 for specifying time-dependent generation rates. I would love to learn which variable or concept or description in the manual causes difficulties. Without that knowledge, all I could do is repeat what the manual says.
Stefan
-
Can I send you my input file so you take a look ?
I am trying to replicate the paper outputs without diffusion. I am attaching the paper so you know what I am talking about.
Thanks
-
Mo,
I would really prefer you making the effort to describe the problem, rather than me just solving it for you. This way, we can improve the description of the GENER block in the manual to the benefit of everybody reading this Forum. It is not even clear to me whether you have a problem with the input file or experienced difficulties running the problem (i.e., whether you got any error messages, and if so, what they are).
Thanks for your understanding,
Stefan
-
Stefan,
Thanks. I guess I found out how to fix the issue. From now on, I will be clearer if I run into any issue. Thank you again