Enhanced Oil Recovery: Mechanisms, Screening, Pattern Flooding and Sweep, Displacement Efficiency, and Chemical, Miscible, Thermal, Microbial and Low-Salinity Methods

After primary recovery and waterflooding, typically well over half the original oil remains, for two reasons. Some of it was never contacted by the injected water, which bypassed it through high-permeability layers, fingered through it because the water was more mobile than the oil, or never reached it between wells — a failure of macroscopic (sweep) efficiency. The rest was contacted but trapped as isolated ganglia in the pores by capillary forces — the residual oil saturation, a failure of microscopic displacement efficiency. EOR is the injection of something the reservoir did not originally contain — chemicals, miscible gases, heat, microbes, or water of altered chemistry — to attack one or both.

The mobility ratio is M = λ_displacing/λ_displaced = (k_rw/μ_w)/(k_ro/μ_o) for a waterflood, with the water relative permeability taken at residual oil (behind the front) and the oil relative permeability at connate water (ahead of it). M ≤ 1 is favourable: the front is stable and sweep is good. M > 1 is unfavourable: the more mobile water fingers through the oil, breakthrough is early and sweep poor. The capillary number N_c = vμ/σ, the ratio of viscous to capillary forces (v the Darcy velocity, μ the displacing-fluid viscosity, σ the interfacial tension), is about 10⁻⁷ to 10⁻⁶ in a waterflood; residual oil only begins to fall when N_c is raised by several orders of magnitude, which in practice means cutting the interfacial tension by a factor of a thousand or more with surfactants or eliminating it altogether by miscibility.

Injection and production wells are laid out in repeating patterns. In a five-spot, each producer is surrounded by four injectors at the corners of a square and the ratio of injectors to producers is 1:1; in the normal seven-spot it is 2:1 and in the normal nine-spot 3:1 (the inverted patterns reverse the roles); line drives place rows of injectors opposite rows of producers, directly or staggered. Peripheral and crestal injection suit dipping reservoirs. The overall recovery efficiency of a displacement is the product E = E_D × E_V, and the volumetric sweep is itself E_V = E_A × E_I: areal sweep E_A (the fraction of the pattern area contacted, which falls as M rises and is read from correlations against M), vertical sweep E_I (the fraction of the thickness contacted, which falls with layering and gravity override), and microscopic displacement efficiency E_D = (S_oi − S_or)/S_oi within the contacted rock, where S_oi is the oil saturation at the start of the flood.

Permeability heterogeneity is the main enemy of vertical sweep. The Dykstra-Parsons coefficient V = (k₅₀ − k₈₄.₁)/k₅₀, from a log-probability plot of the permeability distribution (k₅₀ the median and k₈₄.₁ the value exceeded by 84.1% of the samples, one standard deviation below the median), runs from 0 for a homogeneous reservoir toward 1 for an extremely heterogeneous one; the Lorenz coefficient measures the same thing from a plot of cumulative flow capacity against cumulative storage capacity. High-permeability streaks take the injected fluid and water out early, which is why conformance control (gels, polymer, selective completion) is part of EOR.

Screening matches a process to a reservoir before any detailed study, using the properties that make each process work or fail. Thermal methods suit heavy, viscous oils (whose viscosity heat reduces dramatically), in shallow, thick, permeable sands — depth brings heat loss from the wellbore and high steam pressure. Miscible gas methods suit light oils at reservoir pressures high enough to reach the minimum miscibility pressure (MMP), which usually means depth. Chemical methods suit moderate temperature and salinity (surfactants and polymers degrade in hot, hard brine), sandstones more than clay- or anhydrite-rich carbonates (adsorption and precipitation), and oils of low to moderate viscosity. Microbial methods need temperatures and salinities microbes survive. Screening narrows the options; laboratory core floods, pilots and simulation decide.

Chemical flooding. Polymer flooding (partially hydrolysed polyacrylamide, or the biopolymer xanthan) raises water viscosity and, for polyacrylamide, lowers water permeability, reducing M and improving sweep. Surfactant (micellar) flooding injects a slug that lowers oil-water interfacial tension to ultra-low values, raising N_c and mobilising residual oil into an oil bank; it is followed by a polymer drive for mobility control. Alkaline flooding generates surfactant in situ by reacting with acidic components of the crude, and ASP (alkali-surfactant-polymer) combines all three. Losses by adsorption on rock, and incompatibility with divalent ions, are the chief economic risks. Miscible flooding injects a solvent that mixes with the oil in all proportions, eliminating the interface and so the capillary trapping: first-contact miscibility with solvents such as LPG slugs, and multiple-contact (dynamic) miscibility with CO₂, rich or lean hydrocarbon gas or nitrogen, developed by vaporising oil components into the gas or condensing gas components into the oil above the MMP. CO₂ also swells the oil and lowers its viscosity. Gas is far more mobile than oil, so water-alternating-gas (WAG) injection is used to control mobility and gravity override.

Heat lowers the viscosity of heavy oil by orders of magnitude, and that is the main mechanism of every thermal method, with steam distillation of light ends and thermal expansion adding to it. Cyclic steam stimulation ("huff and puff") injects steam into a well for days to weeks, shuts it in to soak so that the heat spreads, then produces the heated oil from the same well, repeating the cycle as response declines; it is cheap and quick but recovers a modest fraction. Steam flooding injects continuously at injectors and drives heated oil to producers, recovering much more, with steam override and heat losses to cap and base rock as its limits. Hot-water flooding is simpler but carries far less heat per unit mass than steam, whose latent heat dominates. In-situ combustion (fire flooding) injects air and burns a small fraction of the oil — the heavy coke-like residue — in a combustion front that moves through the reservoir, generating heat, steam and gases in place: forward combustion moves with the air, wet combustion adds water to carry heat forward from the burned zone, and reverse combustion moves against the air flow. It has the least surface heat loss and the hardest control.

Microbial EOR (MEOR) injects microbes and nutrients, or stimulates the indigenous population, so that their metabolic products help recovery: biosurfactants lower interfacial tension, gases (CO₂, methane) swell the oil and repressurise, acids and solvents dissolve carbonate and reduce viscosity, and biomass and biopolymers plug high-permeability thief zones to divert flow. It is cheap but slow and hard to control, and limited to temperatures, salinities and pressures that the organisms tolerate. Low-salinity water flooding injects water of much lower salinity (and often altered ionic composition) than the formation brine. Laboratory and field results show extra recovery where the rock contains clays and the oil contains polar components; the proposed mechanisms include wettability alteration toward more water-wet through multi-component ion exchange and expansion of the electrical double layer, and fines migration that diverts flow. It is attractive because it needs no chemicals, but the response is reservoir-specific.