Grid orientation effects on mineral dissolution

Hello everyone,


I made the model with homogeneous mineral distribution and  set the acid source at center position.

At first, I run the model without permeability enhance due to the mineral dissolution on flow(KCPL=0). The simulation showed the concentric dissolution, which seems appropriate results. 

However, when we considered the model with permeability enhance due to the mineral dissolution (KCPL=1、Ipptyp = 1: Simplified Carman-Kozeny), it showed the predominant dissolution on grid direction. 

The result is obviously inappropriate.

I used the TOUGHREACT version3.0 ECO2N 


I have no idea to solve the problem,  I appreciate if someone help me.

I have already tried 9- point finite difference methods. It alleviates the problem but  it is not complete solution.

Thanks a lot for your time and help!



3replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • Ozawa,

    it seems you are running TOUGHREACT under Petrasim. Have you tried to use a Voronoi grid ("Polygonal" in Petrasim) instead of a cartesian one?

    You can also try to force Petrasim to build a hexagonal regular grid by assigning related node coordinates through the "Additional refinement" window.



  • Alfredo


    Thanks for kindly reply

    Now, I'm trying to use a Voronoi grid.

    I'm a bit struggling to run the simulation. (it  needs long time )

    But,  Intermediate output shows the homogeneous dissolution independent of direction so far.

    I will show you the result,  after the simulation finish.



  • Alfredo


    I tried the simulation with Voronoi grid.

    It showed homogeneous  distribution  until the middle, but finally it showed inhomogeneous dissolution.



    I also tried a circular grid to match the flow direction with the grid direction.

    but it also showed inhomogeneous dissolution.

    Small numerical errors make a big difference later time.

    In practically, geological setting is not homogeneous condition, I wonder if small numerical errors are acceptable. 



Like Follow
  • 1 yr agoLast active
  • 3Replies
  • 123Views
  • 2 Following