Vertical groundwater head gradient in a hydrostatic problem
Miguel Calpe
Dear pflotran group,
I have a simple rectangular domain where I only apply a hydrostatic initial condition. The datum has been chosen to be at the top of the domain.
The dimensions of the rectangular domain are 100000 m x 20000 m x 10000 m.
The grid is structured with a discretization nx x ny x nz = 1000 x 20 x 10.
When I plot
the LIQUID_HEAD, there is a vertical head gradient. Since I have only a hydrostatic
initial condition, I would expect the LIQUID_HEAD to be constant in depth, i.e.
no flow in z-direction.
The problem is fully saturated and the liquid saturation is 1 at all time steps.
I am using the branch glenn/total-liquid-head:
https://bitbucket.org/pflotran/pflotran/branch/glenn/total-liquid-head
Please find attached the input file to reproduce the issue.
Thanks for the help.
Miguel
Hammond, Glenn E
Miguel,
The results below are due to the use of an equation of state to calculate water density. The water density varies with depth. Add the following card to your input deck, and you will see that the LIQUID_HEAD ~= 0.
EOS WATER
DENSITY CONSTANT 1000.
END
The code uses the density of the grid cell to calculate the head, but you cannot do that when the density is variable in space. For this reason, I believe that it would be prudent to remove LIQUID_HEAD as an output option in PFLOTRAN. I have assumed that users understand this limitation.
Glenn
From: pflotra...@googlegroups.com <pflotra...@googlegroups.com>
On Behalf Of Miguel Calpe
Sent: Tuesday, October 17, 2023 4:07 AM
To: pflotran-users <pflotra...@googlegroups.com>
Subject: [pflotran-users: 7327] Vertical groundwater head gradient in a hydrostatic problem
Check twice before you click! This email originated from outside PNNL.
Dear pflotran group,
I have a simple rectangular domain where I only apply a hydrostatic initial condition. The datum has been chosen to be at the top of the domain.
The dimensions of the rectangular domain are 100000 m x 20000 m x 10000 m.
The grid is structured with a discretization nx x ny x nz = 1000 x 20 x 10.
When I plot the LIQUID_HEAD, there is a vertical head gradient. Since I have only a hydrostatic initial condition, I would expect the LIQUID_HEAD to be constant in depth, i.e. no flow in z-direction.
The problem is fully saturated and the liquid saturation is 1 at all time steps.
I am using the branch glenn/total-liquid-head:
https://bitbucket.org/pflotran/pflotran/branch/glenn/total-liquid-head
Please find attached the input file to reproduce the issue.
Thanks for the help.
Miguel
--
You received this message because you are subscribed to the Google Groups "pflotran-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
pflotran-user...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/pflotran-users/e4ecc31e-155d-4ba3-8d62-3d144ecc23a6n%40googlegroups.com.