[branch fix/kai-update-to-3.0] Develop a useful setup for the fracture exercise

Currently we have two different options for boundary conditions. Either we use top and bottom Dirichlet and control the system via the boundary saturations. Or we use a Neumann condition at the bottom and inject water from the bottom center.

The domain is rather big, the fractures not much connected. Therefore, it is not much resembling to a situation in shale and overburden.

Ideas how to improve that? Using a fracture network only in the bottom part? Or adding a number of well-connected fractures in the bottom part?