equinor
Distribute volume per cell
After #356 is merged, a possible follow-up could be to distribute the volume from the original model per cell instead of per tube. This would lead to smaller volumes close to wells (where tubes come together) and larger volumes further away from a well.
This was tested with the #356 implementation but need further works as it led to many holes due to the limited resolution of the original source model:
Depending on how many cells that are actually missing here, it could be an idea to do some kind of moving average along the flow tubes here, to "fill in the blanks".
I repeat and extend a comment I made in #356: an alternative could perhaps be to first assign volumes to flow tubes, and then to distribute that volume over the flow tube cells based on the distance to either (1) its end points or (2) to wells. This would also prevent the effect seen in the above figure that volumes are locally reduced at places where 2 flow tubes at different depths 'cross' each other.
