I am currently developing a numerical model in an open-pit mine. It is based in the finite element method through the FEFLOW code. The pit is some 250 m depth, with a diameter of 1-2 km in the upper surface and several hundred meters in the bottom.
Here we have a marls pack confining a thin porous aquifer and a fractured, low permeability paleozoic basement where the orebody is also located.
In addition to the main drainage with a set of drainage Wells around the pit, we find an upwelling of water in the bottom of the pit, which is also pumped away in order to allow proper operation.
Currently the model implements a 3rd-kind, Cauchy-type boundary condition, where the reference height is the absolute elevation in the bottom and the conductance is adjusted with the available measurements (some 30-50 l/s, that is 2500-4500 m3/day).
However, somebody does not agree with this decision, arguing that the most adequate would be a 1st-kind, Dirichlet type boundary condition, taking the height of the water accumulated in the bottom as the reference piezometric head.
What do you think. Is it Cauchy, or Dirichlet, or any one of these but implemented in a different way?
Your contributions will be highly appreciated. Many thanks in advance