In recent years, there has been a number of experimental and numerical studies on the transport and settlement of proppant in scaled laboratory fractures that are of interest to understand hydraulic stimulation in hydrocarbon reservoirs. The results from these studies are generally taken with skepticism by the industry. Part of this is due to the perception that results must be very sensitive to the boundary conditions. Small changes in the experimental configuration are expected to lead to very different results. In particular, little is known on how the actual roughness of the fracture walls and the way the slurry is drained from the laboratory cell affect the results. Here, we report experimental results on the proppant transport in a scaled vertical rough cell that mimics a vertical fracture induced during stimulation. We compare results with those obtained in a smooth wall cell, which is more common in laboratory tests. We found that a rough wall allows for a significant increase in the amount of proppant deposited in the cell at high pumping rate. However, the effect of roughness seems to be marginal for low pumping rates, suggesting that most previous studies in the literature which use of a reduced pumping rate were not significantly affected by the use of smooth walls. We also designed different flow configurations at the fracture tip (where the slurry drains form the cell) to study their effect on the way in which the proppant is deposited in the fracture. We found that, in general, the deposited dune is fairly independent of the particular boundary conditions. This suggests that most results reported in the literature are also robust against changes in the selected outflow configuration.