Open-to-flow boundary condition implementation.

[anupkgeos]

Hi
I want to solve CO2 infection problem in a brine-saturated reservoir:

1. How can implement open-to-flow boundaries as:
   
   Here primary variables are pgas, zi, and xnacl. Initially, the reservoir is fully saturated with brine(Xnacl=0.1).

[BCs]
  # Left pgas = No flow
  [top_pgas]
    type = DirichletBC
    variable = pgas
    value = 5e6  # 5 MPa
    boundary = 'top'
  []
  [top_zi]
    type = DirichletBC   # or PorousFlowOutflowBC ??
    variable = zi
    value = 0  
    boundary = 'top'
  []
  [top_xnacl]
    type = DirichletBC  # or PorousFlowOutflowBC ??
    variable = xnacl
    value = 0.1 
    boundary = 'top'
  []
[]

For outflow, which one would be more appropriate here: DirichletBC # or PorousFlowOutflowBC ??

Thank you!

[GiudGiud]

Hello

DirichletBC are appropriate for pressure at the top I think.
For flow I would not expect to set the concentration on the boundary. Especially free-flow.
@cpgr may want to pitch in, though did you check if there's an example of this in the module?

Guillaume

[cpgr]

I would try Dirichlet for pressure and temperature at the top and bottom (and I guess a function Dirichlet on the right hand side), and the PorousFlowOutflowBC for the other variables. I'm not sure what they mean by constant saturation though - if gas phase CO2 can reach a boundary and flow out then the saturation will change at that boundary?

[anupkgeos]

In the paper pgas and sgas are used as the primary variable and a constant saturation is applied.
I tried the BCs as you mentioned...simulation runs okay for a few steps but it seems CO2 saturation fronts don't move forward and pressure increases too much and at the end getting pressure must be less than 60 MPa and non-linear solution doesn't converge. Here is my input and mesh file...would you please have a look?

https://github.com/anupkgeos/moose_co2_test/blob/main/5cappa2011a_2comp_simple.i
https://github.com/anupkgeos/moose_co2_test/blob/main/cappa2011_simple.msh

[cpgr]

Sorry about the late reply! I think your injection rate is too high - it is integrated over the boundary. Did you mean for this to be a cylindrical model? At the moment, it is only 2D.