The acoustical data on Spirit River sandstone are from Knight and Nolen-Hoeksema (1990). Both Murphy (1982) and Walls (1982) also studied the Spirit River sandstone, and Walls' sample SR6547 is apparently the same one studied by Knight and Nolen-Hoeksema. Imbibition data are expected to fall closer to the homogeneous saturation curve, which is seen to be true here for both Vp and Vs. Although the drainage data are expected to fall closer to the patchy saturation line, this tendency is only observed here at the higher saturation levels (> 40 %).
Walls (1982) studied the gas permeability of the Spirit River sandstone as a function of both saturation and effective pressure. At room pressure the gas permeability changed from 100 D to 47 D as the liquid saturation changed from zero to 40%. Then, as the effective pressure increased, the permeability of the 40% saturated sample fell more rapidly than that of the other sample, differing by an order of magnitude at 30 MPa.
Since the porosity of this sample is about 5%, the volume fraction of the whole sample occupied by liquid at 40% saturation is about 2%. It is commonly observed that liquids can begin to percolate (i.e., maintain a continuous connected pathway) across a porous sample when their volume fraction is of the order of 2-3%. So we assume that this dramatic departure of the observed drainage data is in some way related to this percolation threshold.