Injector Fluid w/ Changing Composition
Hi all,
I am looking to model reservoir response to changing injection water composition and am seeking advice on how to do so. Specifically, I am wondering if it would be best, and/or feasible, to
1. Create identical generators, with altered generation times in Record GENER.1.1 and separate injection water compositions in chemical.inp, or
2. Break the simulation period up, input the savechem file as inchem, and adjust a single injected water composition in the chemical.inp file for each period (if inchem and chemical.inp can be used simultaneously?).
Any suggestions or additional insights on how to tackle this problem are greatly appreciated!
Thanks,
John
1 reply
-
Hi John,
Definitely, option 2 will work, because the savechem does not save the injection water composition, just the compositions in the grid blocks.
Regarding option 1, that probably will not work, since the injection composition is associated with the grid block being injected into. However, there are certainly some tricks that would likely work. If you inject into a series of extra grid blocks (unconnected to each other) that are either at the wellhead or attached to the downhole reservoir element, you could inject sequentially into each one of those blocks which are then connected to the reservoir or well element. In that way the injection rate, enthalpy, and times could also be different for each different injection composition. A downstream weighting might be needed at the interface so that there is no back flow into the other elements that are not being injected into. That might be minimal, though, if there is a lot of injection relative to the volume of those elements. If you try it, let us all know if it works!
cheers,
Eric