The use of divertants, usually foaming surfactants with nitrogen, are being researched as means of improving steamflood performance. It is shown that recycling of produced bitumen to the injection well is an effective method of reducing mobility in the steam zone. In addition greater recovery of produced heat is possible for thermodynamic reasons. Bitumen recycle may allow economical steamflooding of oil sands reservoirs; independant of gravity drainage. A numerical example is presented, and practical operating considerations are discussed.
The use of diverting agents to improve the sweep efficiency of steamfloods is an active field of research. Foaming surfactants, usually in combination with small amounts of nitrogen, are the most common materials under study; recent reports indicate that mobility reductions of, a factor of 10 to 20 are possible under laboratory conditions. (1,2)
In attempting to steamflood the Alberta oil sands the channelling problem is especially severe, 1eading to rapid breakthrough and low oil/steam ratios (OSR's). Most approaches to economical recovery are based on gravity drainage initiated with horizontal fractures low in the pay, or thin bottom water zones (3,5). The use of diverting agents is another possible strategy, however. and one which does not depend on a specific geometry for the initial communication path.
The contention of this paper is that reinjection of a part of the produced bitumen appears to be an effective means of reducing steam mobility and hence improving steamflood performance in oil sands. Recycling is discontinued after a satisfactory volume of the reservoir has been heated; a conventional steamflood is then used to recover the highly mobile "inventory" of recycled bitumen.
The next section deals with the mechanisms by which mobility reducing agents improve the performance of steamfloods in oil sands. Following this. the effective mobilities of various steam-bitumen mixtures are calculated. The seemingly dubious economics of bitumen injection are then discussed, followed by simulation results which compare a hypothetical Athabasca steamflood with and without bitumen recycle. Finally, some of the production engineering aspects of field application of a bitumen recycle process are outlined.
The effect of a mobility control agent in an oil sands steamflood is somewhat more subtle than simply diverting steam into unswept regions of the reservoir. The cold reservoir does not allow significant flow under any reasonable pressure gradient (unless mobile watersaturations are present); therefore a completely effective divertant, for example cement filling the entire warm zone, would simply cause a loss of injectivity.
Several authors have described the general mechanism of steam front propagation in immobile oil sands for specific situations. Clossman and Smith (6) analyze a horizontal steam front rising vertically from a horizontal fracture. Butler et al (7) calculate the drainage of bitumen down an inclined steam front; and Edmunds (4) applies Butler's calculations to the steam drag effect, where the bitumendrainage is due to an applied pressure gradient rather than gravity.
In each of the above cases the propagation of the steam front depends on the interaction of conduction and convection (figure 1).