Discretisation around nodes where Hydraulic head BCs are used to simulate a well

Hi, sometimes we use a set of Hydraulic head BCs to simulate a well, because we want to reduce the head in the well to a certain level and does not know the corresponding pumping rate .

However, when using a Multilayer well BC, I know that correct discretization is crucial for achieving the correct drawdown in the well itself, with a given pumping rate.

Does the same discretization issue apply when we are using hydraulic head BCs, i.e. is the rate with which water is removed from or added to the model only correct with the correct discretization ?.

Christian

I've personally not tested this, but with better discretization you're likely going to get a better estimate of flow, based on the white papers for BHEs... I typically logarithmically scale discretization around all wells based on the that have any sort of boundary condition or require an accurate observation. Symmetry is important, at least on a finer scale.

I'm not sure how good this is, but you can also apply a mesh refinement gradation around supermesh points which will likely help if it's too problematic to modify the supermesh manually around the wells. I'm not sure what kind of algorithm the Triangle Mesh Generator applies for refinement gradation, but it's probably not ideal.

Hope this helps,

Adam

Indeed there is a relationship the other way round, too. While the head depends on discretization when using a Well BC, the flow will depend on discretization when using a Head BC.