Salt precipitation dependency on rate (and grid size)
This is about testing salt precipitation in a 1D radial model with a single injection well at its center using ECO2N where capillary forces included and a pressure release well put on its boundary.
I have apparently experienced some dependency on grid size and injection rate when it comes to salt accumulation in near well area. The grid size effect seen as different distributions of salt within grid blocks away from the well, while accumulated precipitation seems to be similar over a certain distance from the well.
It is therefore easier to see a volumetric logic in it, but I am not sure if that will explain all.
But dependency on injection rate is more difficult to get grip on, since the rate of precipitation versus the viscous force (gas movement) seems to be acting differently. While at a high injection rate, models with various grid sizes distribute salt unevenly along the flow path, at a low flow rate, each subsequent grid block deposit salt to its maximum capacity. It could be seen as a result of high retainment time, But I am not quite sure if that's all.
I would like to have your comments and possible experiences on that.
The 1D radial model had it's first 100ft divided equally in 33,100 and 200 cells as coarse, base and fine models. Beyond 100ft, "dr" for each subsequent cell, was doubled until the boundary that lied10s of kilometer away from the well. It models a hypersaline aquifer with 0.2 mass fraction of salt without initial CO2 dissolved. Reservoir P/T: 138bar and 28'C.
In my experience, precipitation of salt accompanying CO2 injection is one of the most grid-sensitive problems there is. With a coarse enough grid you may not see precipitation at all, as you are averaging over such a big volume that enough aqueous phase exists to keep salt dissolved. I highly recommend reading Karsten Pruess and Nadja Muller's paper on this topic (Formation dry-out from CO2 injection into saline aquifers: 1. Effects of solids precipitation and their mitigation, Water Resources Res, 45, W03403, doi: 10.1029/2008WR007102, 2009), for insights into the physical processes and modeling them with TOUGH, which should prove useful for your work.