GEOS icon indicating copy to clipboard operation
GEOS copied to clipboard

Negative pressure change after stopping injection

Open liushilongpku opened this issue 10 months ago • 1 comments

Hello everyone, I found that the well and the surrounding grid produced Delta_P<0 after the injection stopped when I use GEOS process, that is, the pressure between the grid selected as the well and some surrounding grids after the injection stopped was less than its initial value.(See Figure 1.)

图片 Figure 1 : Negative pressure change after stopping injection

This occurs in the following situations: When CO2 is injected into an initial high water cut or high oil bearing formation under a constant pressure boundary, the well is set up by applying a sourceTerm to a specified grid in place of the well.

Other tests:

  • During the injection process, the change of mesh pressure is greater than 0.
  • This may not be relevant to the constitutive model, at least for both the black oil model and the saltwater model.
  • This does not occur when water or oil is injected into formations with high water and oil content.
  • Under non-constant pressure boundary conditions, all grid pressure changes are positive, but the pressure increase between the grid and the surrounding grid after the well has stopped injection is still smaller than that of the more distant grid.(See Figure 2.)
  • Changing the volume of gas injection does not change the positive and negative relationship between the pressure and the initial value after stopping injection, but only affects the absolute value of the pressure change, and this effect is basically linear.

图片 Figure 2 :The increase in pressure in the center is smaller than the increase around it, even though the values are positive. The injection is stopped at 2.0e8.

Is this a correct physical phenomenon or a problem within the program?

My platform is :

  • Machine: [Ubuntu20.04 and Ubuntu22.04]
  • GEOSX Version : [up to now.]

Negative pressure change after stopping injection.docx

gas.zip

liushilongpku avatar Apr 09 '24 15:04 liushilongpku

Hi @liushilongpku Can you please set up a flow only simulation (without poromechanics) using the same data. Just to check if the mechanical response is causing some of the effect.

dkachuma avatar Jul 03 '24 19:07 dkachuma