0

EOS7c and NCG injection

Dear all,

I am using EOS7c module, to model the injection of a NCG (either CO2 or N2) in a depleted CH4 reservoir. As soon as the mass fraction of the non-condensible gas X_NCG_G approaches 1 (i.e. X_METHAN_G become small, smaller than let's say 1E-5)  the simulation stops after some iterations failure, providing the message 

TOUGH STATUS: Failed in EOS. IGOOD=           2  


Is this a limitation of the module or am I messing up something?
I have set-up the simulation without diffusion (NB = 6 in MULTI). 

The temproary fix I did is injecting a small quantity of methane (<1% of the injected NCG) where the NCG is injected. This way the iterations procede without failure, but maybe there is a more elegant solution. 

Any additional suggestion/comment on overcoming the problem is welcome, thanks.

3 replies

null
    • Staff Scientist
    • Christine_Doughty
    • 2 yrs ago
    • Reported - view

    I would need to see more of your output to know for sure, but I have experienced this sort of thing also.  In the original EOS7c, it was a strict requirement that some CH4 always be present, but that was a long time ago, so I doubt you have that version of the code.  Turning on diffusion might help.  Your temporary fix of co-injecting a little CH4 is something that I have also used successfully - in fact that is my go-to action when component disappearance causes problems.  You could also check that your grid resolution does not change abruptly anywhere - that can cause difficulties regarding phase change in general.  

      • lurpi.1
      • 2 yrs ago
      • Reported - view

      Christine Doughty thanks for the reply. I am using TOUGH3, version 1.0.0 from 2018, so I think it is not an issue version.
      I admit that my grid may not be the best, I will run again some small tests and if I consistently need to stabilize the NCG injection with CH4 injection I will report it here. 
      Thanks again

    • kenny
    • 2 yrs ago
    • Reported - view

    EOS7C use an iterative approach for solving gas mutual solubility. This approach sometime has the convergence problem and causes  failure of the iteration.  We are working on a non-iteration method for next version of TOUGH codes.

Content aside

  • 2 yrs agoLast active
  • 3Replies
  • 89Views
  • 4 Following