Investigations into the variability of bitumen contained in Athabasca Tar Sands have shown that the bitumen is of a widely varying composition. Bitumen from the bottom of the deposit is low in distillate and high in asphaltenes. The efficiency of the hot water process for separating bitumen from tar sand is greater for tar sand from the bottom of the deposit than for material from the top of the deposit. In addition, the hot water process preferentially floats certain bitumen components. These three factors cause a higher concentration of asphaltenes in the bitumen fed to upgrading than its natural concentration in the tar sand. As a consequence of these relationships and the" sensitivity of upgrading technology to feedstock characteristics, proper design of upgrading facilities requires extensive piloting of the preceding bitumen extraction processes.
A considerable number of authors have addressed themselves to the many aspects of the Athabasca tar sands, to their extent(1), geology(2,3), principles of the hot water extraction (4,5), mining schemes (6,7), bitumen upgrading processes(8,9,10), and to the engineering task of putting it all together into an optimum configuration(11) with one 60,000 bpd plant in operation and one 125,000 bpd plant under construction one could well imagine that the necessary technology and the understanding of underlying principles is available for anyone who wishes to develop a mining lease in the tar sands. It is the object of this paper to show that despite these advancements it remains of vital importance to any prospective tar sand operator to enter into an extensive and indepth research program so that unpleasant surprises may be avoided. The source of these surprises is two fold:
One, the variable nature of tar sands,
TWO, the interaction between tar sand and the hot water process for extraction of bitumen.
The variations in tar sand have been described by a number of authors from a geology point of view(12,13). Blair (14) in his classical report relates that the specific gravity and the viscosity of the bitumen component of tar sands varies with geographic location: sp.gr. 1.027 at McMurray to 1. 005 in the Ells River district. In general, bitumen in the north is lighter than that in the south, and viscosity of the unaltered bitumen is much greater in the southern part of the area than in the north. Thus the viscosity of bitumen from Abasand is 600,000 poises at 50 ºF, compared with that from Ells River and Bitumount, which is only 6,000 to 9,000 poises at this temperature. From these slight variations and others(15) reported since the Blair report it could be concluded that bitumen in tar sand is reasonably uniform with only small variations between localities with viscosity as the only exception. Nothing is further from the truth. For reasons other than those associated with a curiosity about tar sands, we set out to determine the variability of bitumen in a vertical direction rather than lateral.