Node placement in volume of original model

[LonnekevB]

Insert a description of your pull request (PR) here, and check off the boxes below when they are done.

---

Contributor checklist

• 🎉 This PR closes Node placement in volume of original model #368.
• 📜 I have broken down my PR into the following tasks:
  • Task 1
  • Task 2
• 🤖 I have added tests, or extended existing tests, to cover any new features or bugs fixed in this PR.
• 📖 I have considered adding a new entry in CHANGELOG.md.
• 📚 I have considered updating the documentation.

[wouterjdb]

I tested the code - but had to make some minor changes such that the random point generation actually happens in the space of the original model.

This is a picture of the Norne model (top layer). Now stretching all the way to the west - the area that doesn't have any wells.

The full Norne model, with many nodes per layer, would now look like:

And for reference the real Norne model (initial pressure is quite off by the way, @edubarrosTNO I hope this is not the case in the cases you have been testing?):

[edubarrosTNO]

And for reference the real Norne model (initial pressure is quite off by the way, @edubarrosTNO I hope this is not the case in the cases you have been testing?):

I will have to check that, but what I can say now is that those cases that I have been running are certainly using much fewer additional nodes than what we see in these pictures.

Are there any current hypotheses for the initial pressure being off? To me it looks like the pressure is in the right range in most of the model, except for those red spots in one of the "tentacles". It would be useful to check the EQUIL regions of the full grid and FlowNet models, and see if there is any overlapping with the red spots in pressure in the FlowNet grid. If that is the case, we will have to revise the EQUIL settings of that region

[LonnekevB]

@wouterjdb I realized we were using convex and concave hull interchangeably in the Mitchells code. But for the grid cell bounding boxes it is a concave hull, but for the perforations, we actually use a convex hull.
I think the naming is consistent now and once the tests have passed ready to merge!

[olelod]

I see that during the time I did something else, all of my comments about convex/concave became outdated...

LGTM! Nice work.

It all became a bit messy since there were commits addressing my requested changes before I submitted the review

[olelod]

I just have one comment about a docstring mentioning the hull_factor = 1.0. Other than that everything looks good to go! Nice job.