The introductory discussion assumes that an advantage of CO2 injection for EOR is the potential to offset the costs of sequestration through enhanced hydrocarbon recovery. In this section, important aspects related to phase and flow behavior of CO2 injection during enhanced recovery are discussed.
4.2.1 Immiscible versus miscible recovery
The displacement efficiency associated with the injection of a fluid into a porous medium is determined by multiphase flow characteristics and the phase behavior among injection gas, crude oil, and reservoir brine. In the case of the flow of immiscible fluids, two important factors to consider are interfacial tension (IFT) among phases and rock wettability. Under immiscible conditions, the presence of two, or more, distinct phases is apparent and evidenced by the existence of a sharp interface. In the case of immiscible displacement, IFT plays a role in the competition among the phases to occupy pore space. In conventional waterflooding and other immiscible displacement fluid systems, large values of IFT are associated with large values of residual oil saturation.
Ultimately, oil recovery performance for immiscible CO2 injection is limited by volumetric sweep, displacement efficiency of the injected fluid, and the finite solubility of CO2 in the oil phase that, generally, reduces oil viscosity and oil density. These factors are, in turn, determined by reservoir structural heterogeneities, gravity effects, viscous fingering, rock wettability, crude oil phase behavior, and so on. In most cases, a significant fraction of the oil contacted by the displacing phase is trapped or isolated within unreachable pore spaces because capillary forces immobilize the oil. Thus, the recovery process halts and reduces the oil relative permeability to near zero. The ultimate oil recovery after immiscible fluid injection ranges between 20 and 40 % OOIP, on average (Stalkup, 1984).
4.2.2 Development of miscibility
When two miscible phases flow together, they become indistinguishable and the fluids mix in all proportions. Because there is no interface, there is no interfacial tension between the oil and the solvent. Also, the single mixed phase is not affected by relative permeability or wettability of the rock to oil or solvent. In most cases of miscible displacement, microscopic recovery efficiency approaches 100 %, in the absence of water. Hence, the residual oil saturation is quite small, or zero, where the injected phase contacts the oil.
Depending on the mixing mechanism of the miscible system composed by injected solvent and oil, the miscible displacement process will be first- contact miscible (FCM) or multiple-contact miscible (MCM). In an FCM displacement, the injected fluid forms a single phase with the reservoir oil upon first contact, whereas in MCM miscibility is developed in situ through changes in composition to both the injected and the displaced fluids as they advance through the reservoir. Generally, CO2 develops miscibility with oil by the MCM mechanism.
A pseudoternary diagram is useful to conceptualize the FCM recovery processes. One vertex represents the heaviest fraction of the oil, C7+, another represents the intermediate components, C2–6, and the third represents gas, c1. On Fig. 4.1, the straight line that spans oil and injected fluid composition points lies outside the two-phase dome that defines multiphase hydrocarbon behavior at reservoir pressure and temperature. Any combination of injectant and oil results in a single-phase fluid. When an FCM solvent is injected for secondary recovery in a batch volume or slug in the reservoir, oil and solvent immediately mix and form a single phase in the contact zone and oil is banked downstream of the contact zone, as shown in Fig. 4.2.
The miscible bank features lower viscosity than the pure oil as a consequence
100 % C1
100 % C2–6 Two-phase
region
100 % C7+
Solvent
Oil
4.1 Example of a pseudoternary diagram for a FCM process.
Solvent Oil + Solvent Oil
(primary slug) (miscible zone) (oil bank)
4.2 Miscible displacement.
of the dispersion and molecular diffusion of the solvent in the oil phase.
Any trapped oil contacted by the miscible bank is solubilized and therefore displaced. As the concentration profile advances through the reservoir, an increasingly larger volume of the original solvent slug is dispersed in the oil. At the same time, because the viscosity of the injected solvent is usually significantly lower than the oil viscosity, and in the absence of IFT affecting phase relative permeability, the mobility ratio for miscible processes is unfavorable and there is a tendency to viscous fingering.
Another potential problem in FCM displacements is asphaltene deposition due to the effect of the solvent on the heaviest fractions of the oil. These may cause pore-plugging, reduce the effective rock permeability, and alter the injectivity and productivity of wells.
MCM displacement processes develop miscibility through in situ composition changes prompted by mass transfer and multiple contacts (i.e., mixing) between the solvent and the oil phase. The interplay of solvent composition with the original fluid plays a major role in the type of displacement that results. The ultimate goal of an MCM process is to follow a composition path that brings the mixture to the miscible region of the pseudoternary diagram. This mixing is assumed to take place under equilibrium conditions, up to the miscible region, by condensation of the C2–C6 components from the solvent to the oil. This process is called a condensing-gas drive, and Fig. 4.3 illustrates a typical case.
If the injected solvent is mostly composed of methane or natural gas, and the oil composition is high on the intermediate-molecular-weight components,
C1
C2–6 C7+
Injection gas Gas 1
Gas 2
Mix 1Mix 2
Oil Liq 1 Liq 2 Limiting tie
line
4.3 Pseudoternary diagram for a case of condensing-gas drive miscibility process.
then the gas is subsequently enriched on several equilibrium contacts until its composition is brought to the miscible limiting tie line value. This process is a vaporizing-gas drive as illustrated in Fig. 4.4. It applies mostly to injection fluids such as natural gas, flue gas or nitrogen, provided that miscible conditions, mainly pressure, are possible in the reservoir.
4.2.3 Carbon dioxide (CO2) displacement mechanisms The parameter summarizing the combination of phase behavior and flow is the MMP or minimum miscibility pressure. The MMP is conventionally defined as the pressure needed to recover 90 % of the oil originally in place from a one-dimensional laboratory slim tube with the injection of 1.2 pore volumes of CO2, where pore volumes are computed at test pressure and temperature. Practically, MMP is the pressure necessary to assure the mutual solubility of oil and CO2 and thereby achieve significant recovery.
MMP varies with oil composition and density and generally increases as oil becomes more dense.
Enhanced recovery mechanisms with CO2 include oil-phase swelling, viscosity reduction, and gas–oil displacement when CO2 is below MMP with the oil. Above MMP, the miscible CO2 injection process is believed to be similar to the vaporizing-gas drive mass transfer process that takes the light fractions out of the crude oil and enriches the gas with them. However, co2 is capable of extracting higher molecular weight components from the oil than methane, and this makes the required miscibility pressures for
C1
C2–6
C7+
Injection gas Gas 1
Gas 2
Mix 1 Mix 2
Oil Liq 1 Liq 2
Limiting tie
line
4.4 Pseudoternary diagram for a case of vaporizing-gas drive miscibility process.
the co2 significantly lower than those required for injection gases such as methane, flue gas, natural gas, nitrogen, etc. Also, for a given pressure and temperature, the two-phase region defined by the binodal curve in the pseudoternary diagram is much smaller for the CO2 system than for the methane system at reservoir conditions, Fig. 4.5. These constitute the main advantages of the miscible displacement with CO2 over other gases.
In spite of many favorable factors, miscible displacement with CO2 is affected by the unfavorable mobility ratio that is usually present in solvent–
oil systems. Reservoir heterogeneities compound the problem, leading to channeling of injectant through zones of large permeability and poor volumetric sweep. Viscous fingering of injectant through oil also occurs due to the high mobility ratio. Because the densities of oil and CO2 are similar at a wide range of reservoir conditions (Stalkup, 1984), gravity segregation between the fluids is relatively unimportant in these systems.
The injected solvent volume is typically limited for economic reasons. It is common practice to inject water as a secondary slug to push the injectant deep into the reservoir. Water injection is also useful to achieve mobility control when multiple slugs of CO2 are injected. This is the principle of the so-called water-alternating-gas (WAG) process where volumes of water and co2 are injected in series. The ratio of injected water and gas volume, the so- called WAG ratio, is a major design parameter. During WAG, the combined mobility of the two phases is less than that of the gas injected alone, and so the mobility ratio of the process is improvedand the sweep efficiency of the gas injectant is increased. The injection of water, however, tends to increase the amount of residual oil. Because water reaches some of the oil-filled pore
C1 or CO2
C2–6
C7+
C1 phase boundary
CO2 phase boundary
Limiting tie line
4.5 Representative comparison of the binodal curves for CO2 and methane.
space first, water blocks the advance of CO2 and the subsequent recovery of oil. This effect has been found to be a strong function of the rock wettability and more detrimental in water-wet rocks(Green and Willhite, 1998). At the same time, because CO2 is partially soluble in water, additional water- trapped oil phase can be recovered from occluded pore volumes through mass transfer. Green and Willhite (1998) mention the work from Campbell and Orr, in which they explained the mechanism as CO2 that solubilizes and then diffuses in the water, and then solubilizes in the oil. Subsequently, CO2 swells the oil and the increase of volume pushes the water phase out of the pore throat, allowing for direct contact with more CO2. A balance of these two counteracting mechanisms can yield a reasonable balance of improved mobility, reduced oil phase trapping, and increased CO2 capture by the oil phase.
4.2.4 Carbon dioxide (CO2) injection in low-permeability fractured systems
When there are significant resources still trapped in the matrix block of a fractured reservoir, gas injection is often performed to activate a gravity drainage recovery mechanism. The difference in density between the gas phase in the fracture and the oil phase in the matrix causes oil production until gravitational and capillary forces become equal. When the matrix also presents low permeability and high capillary pressure, injection of a dry gas with the consequent mass transfer between the gas in the fracture and the oil/gas system in the matrix becomes the main recovery mechanism (Kazemi and Jamialahmadi, 2009).
Enhanced oil and gas recovery from fractured low-permeability reservoir rock is challenging and, while studied, significant challenges remain in improving recovery factor. Geological and structural properties of the reservoir have a profound impact on both the microscopic and macroscopic diffusion processes that take place on a CO2 miscible displacement. Low permeability compounds the challenge of CO2 accessibility to oil. Additional factors limit process effectiveness, such as asphaltene deposition, water slug occlusion of pores, and unfavorable rock wettability, especially under immiscible conditions (Vega et al., 2008). Near the well, such factors are collectively referred to as ‘formation damage’ or a well ‘skin’ factor.
On the other hand, fractures offer relatively low resistance to flow, thereby increasing the effective permeability of the reservoir. Injected CO2 is expected intuitively to flow primarily through the low-flow resistance network of fractures rather than the high-flow resistance matrix. Exchange of co2 between the fractures and the matrix is thus key to both CO2 storage as well as enhanced recovery. Pressure, chemical composition, and gravitational gradients are driving forces relevant to CO2 exchange.
According to the equation for steady-state injectivity defined by Dake (1978),
I q
h p k
r r S
= = 2
(ln( / ) + )e w
D p
m [4.1]
where q is the volumetric flow rate at the bottom of the well, h is the formation thickness, Dp is the pressure drop between the reservoir and the well, k is formation permeability, m is the injected phase viscosity, r the radius, and S is the mechanical skin accounting for formation damage near the well.
Subscripts ‘e’ and ‘w’ refer to equivalent drainage radius of the well and the wellbore radius, respectively. Note that I is directly proportional to permeability and inversely proportional to the viscosity of the injected phase.
However, an attractive feature of CO2 injection is the relatively low viscosity of the co2 phase. At 15 MPa and 47 °C, CO2 phase viscosity is only 0.077 mPa-s (Lohrenz et al., 1964) as compared to 0.68 mPa-s for water. Due to its relatively low viscosity, volumetric injection rates of CO2 can be large in both permeable and low-permeability formations (Kovscek, 2002).
As suggested by Equation 4.1, even generally low-permeability formations can accept large volumes of CO2 and the storage rate is enhanced by regions of high permeability. A well intersecting many fractures thereby encounters a formation with an effective permeability that is substantially greater than the matrix permeability. Interestingly, heterogeneous, high-permeability paths are generally viewed in a negative fashion for CO2-based EOR. Efficiency of oil recovery is reduced by high-permeability paths and gravity segregation that promotes incomplete reservoir sweep. This is another point where conventional EOR differs from simultaneous sequestration and EOR.
co2 injectivity is an important factor that needs to be considered while designing a sequestration and/or EOR project. It is affected by factors pertinent to the reservoir characteristics and the fluids. Some of these factors include reactions between the CO2 and some of the rock minerals and reservoir brine salts. In some cases, reservoir permeability will increase by dissolving minerals from the walls of flow channels or will decrease permeability by releasing particles that migrate and plug pores and throats in the rock. Also, co2 might extract some of the medium to heavy components and asphaltene components of the oil could precipitate and reduce CO2 permeability in the reservoir. WAG cycles also reduce the relative permeability of both water and co2 and injectivity will progressively decrease unless there are other mitigating factors. Diverse results in field trials have been obtained in a range of injection conditions, some with lower than anticipated resulting injectivity for both CO2 and water (Stalkup, 1984).
Vega et al. (2008) suggest that primary depletion of low-permeability oil- filled rock creates continuous gas pathways from the fracture to the matrix because, where the reservoir has been pressure depleted and gas released
from solution, significant remaining gas saturation has been created in the matrix. These gas pathways act as a rough course the CO2 may follow and that aids the ability to deliver injectant to the matrix. They argue that in many cases injected CO2 follows the pre-established gas pathways, the CO2 flow slows, and CO2 mixes with the contacted oil mainly through molecular diffusion, although convective dispersion is also expected to occur.