Limitation of self-organization within a confined aquifer

Preferential flow paths are omnipresent in the subsurface, but very hard to observe or parameterize. Often they are even sub-grid processes requiring effective parametrization. In an ongoing study,Westhoff et al. (2016) show within a lab experiment that the steady state effective hydraulic conductance evolves (under certain circumstances) to the conductance that maximizes power by the flux through the confined aquifer. Here we explore why in one setup the effective conductance did obey this maximum power principle, while the same setup with slightly different boundary conditions did not lead to this.We suggest here, with a detailed numerical setup of the experiment, that the degrees of freedom to create a long enough preferential flowpathwas too limited in the latter case: a foam rubber layer, placed between the sand and the plate covering the sand (to pressurize the sand layer) prevented further development of the preferential flow path. While this preferential flow path was long enough to result in an effective conductance leading to maximum power in the first setup, it needed to be longer in the second setup.