Steam injection into heavy oils has been well characterized over the last 40 years, and while steam has been injected into light oils almost as long, the mechanisms and effectiveness of this process are much less understood. When this lack of understanding is coupled with the complexities of flow in low permeability fractured reservoirs, even less is known. This study examines thermal recovery in low permeability, fractured reservoirs using thermal compositional simulation. A diatomaceous reservoir provided input for the rock model. The oil phase is represented by three pseudo-components that characterize a relatively light crude oil. Both the areal and vertical recovery efficiencies are improved for steam injection compared to water injection. The incremental recovery depends on the distribution of permeability and is greatest for homogeneous distributions.
In regard to recovery mechanisms, thermal expansion of the hydrocarbon fluids accounts for over half of the incremental recovery early in the steam drive; after roughly 0.4 pore volume of steam injection (cold water basis), the incremental recovery is split equally among thermal expansion, vaporization, and oil viscosity reduction. Late time behavior is dominated by vaporization as the distillate bank breaks through to the producer. As a consequence of steam injection three separate fluid banks form: a cold water bank, a combined hot water and distillate bank, and the steam front. Hence, displacement mechanisms are substantially different from the heavy-oil situation where oil viscosity reduction and, frequently, gravity drainage are dominant. Although the study uses a diatomite rock and fluid model, this work clearly extends to include all low permeability fractured reservoirs that have low primary and waterflood recoveries. For such reservoirs, significant additional recovery is obtained by implementing a steam drive.
Introduction
In the 1960's, one of the first light-oil steam flood (LOSF) field trials was initiated at the Brea Field near Los Angeles. Brea consists of steeply dipping sands with an average permeability around 70 md and a typical porosity of 22%.1 The development of further LOSF projects was not as rapid as that of heavy oil. Heavy oils benefit tremendously from steam injection (compared to water injection) because heavy-oil viscosity decreases many fold upon heating. For some heavy-oil reservoirs, oil can only be produced with thermal recovery techniques, such as steamflooding. Unlike heavy oils, light oil reservoirs can be developed with water drives, and because water drives are seen as less risky with less initial investment than steam injection, LOSF were developed more sparsely.2 Nevertheless, there are a number of applications where LOSF seems to be a good alternative to waterflooding. In highly-dipping reservoirs, thermally-enhanced gravity drainage improves the recovery process; likewise, in confined unconsolidated sandstones residual oil saturations can be very low, around 5%. Application of steam floods in low-permeability hydraulically-fractured reservoirs, such as diatomite, is also promising.3,4,5,6
Significant detail has been paid to steamflooding recovery mechanisms. Wu7 presented a critical review. He and others8 have identified the following, thorough, but not exhaustive, list:viscosity reduction;distillation (vaporization);distillate (in-situ solvent) drive;steam (gas) drive;thermal expansion;relative permeability and capillary pressure variation; andgravity segregation.
Figure 1 displays schematically how the most significant mechanisms vary with oil reservoirs containing viscous oil. The objective of steamflooding is to increase oil production by reducing oil viscosity; thus, allowing oil drainage at significantly increased rates. Conversely, for low viscosity oils, the primary objective is to reduce remaining oil saturation below that obtainable by waterflooding. This is accomplished through vaporization of oil. In fractured systems or highly heterogeneous reservoirs, thermal conduction allows heat to sweep areas of the reservoir not contacted by steam. In this case, thermal expansion is an important recovery mechanism.