Potential for Polymer Flooding Reservoirs with Viscous Oils

This report examines the potential of polymer flooding to recover viscous oils, especially in reservoirs that preclude the application of thermal methods. A reconsideration of EOR screening criteria revealed that higher oil prices, modest polymer prices, increased use of horizontal wells, and controlled injection above the formation parting pressure all help considerably to extend the applicability of polymer flooding in reservoirs with viscous oils. Fractional flow calculations demonstrated that the high mobile oil saturation, degree of heterogeneity, and relatively free potential for crossflow in our target North Slope reservoirs also promote the potential for polymer flooding. For existing EOR polymers, viscosity increases roughly with the square of polymer concentration—a fact that aids the economics for polymer flooding of viscous oils. A simple benefit analysis suggested that reduced injectivity may be a greater limitation for polymer flooding of viscous oils than the cost of chemicals. For practical conditions during polymer floods, the vertical sweep efficiency using shear-thinning fluids is not expected to be dramatically different from that for Newtonian or shear-thickening fluids. The overall viscosity (resistance factor) of the polymer solution is of far greater relevance than the rheology. We extended the fractional flow calculations to examine the effectiveness of polymer flooding when a waterflood (with 1-cp water) was implemented prior to the polymer flood. We showed that polymer flooding can be effective in viscous (1,000-cp) oil reservoirs even if waterflooding has been underway for some time. Introduction The objective of this work is to examine whether polymer flooding can provide a feasible means to recover viscous oils from reservoirs where thermal methods may not be applied. First, the screening criteria for application of polymer flooding are reconsidered. Next, fractional flow calculations are used to illustrate improvements in displacement efficiency that can be achieved by polymer flooding viscous oils. A simple benefit analysis is provided to compare polymer flooding with waterflooding. Permeability reduction and its variation with permeability are considered. The impact of polymer solution rheology (especially shear thinning) on vertical sweep is discussed. Finally, the effect of waterflooding before the polymer flood is considered. Reconsideration of Screening Criteria for Polymer Flooding Alaska’s North Slope contains a very large unconventional oil resource—over 20 billion barrels of heavy/viscous oil (Stryker et al. 1995, Thomas et al. 2007). Conventional wisdom argues that thermal recovery methods are most appropriate for recovering viscous oils (Taber et al. 1997a, 1997b). However, for viscous oil reservoirs on Alaska’s North Slope, a number of factors complicate this thinking. The formations that hold vast viscous oil reserves, Ugnu, West Sak, and Schrader Bluff, are relatively close to permafrost. Steam generation is prohibitive here; with severe cold weather on the surface, heat losses while pumping steam down through 700 to 2,200 ft of permafrost, and heat losses when contacting the cold formation. There are also environmental considerations (air and water quality issues, disturbance of wildlife species, Thomas et al. 2007). Kumar et al. (2008) examined waterflood performance using unfavorable mobility ratios. They concluded that viscous fingers dominate high-viscosity-ratio floods, that mobile water can significantly reduce oil recovery, and that reservoir heterogeneity and thief zones accentuate poor displacement performance. Their paper strongly suggested that any improvement in mobility ratio (e.g., polymer flooding) can noticeably improve reservoir sweep and recovery efficiency. Beliveau (2009) reviewed waterfloods in viscous oil reservoirs and concluded that estimated ultimate recoveries could reach 20-40% OOIP under appropriate circumstances. He noted that normally, 50% or more of the oil would be recovered at high water cuts (>90%). Earlier screening criteria indicated that polymer flooding should be applied in reservoirs with oil viscosities between 10 and 150 cp (Taber et al. 1997a, 1997b). Two key factors were responsible for this recommended range. First, considering oil prices (~$20/bbl) and polymer prices (~$2/lb for moderate molecular weight, Mw, polyacrylamide or HPAM polymers) at the time, 150 cp was viewed as the most viscous oil that could be recovered economically using polymer flooding. (For oil viscosities below 10 cp, the mobility ratio during waterflooding was generally viewed as sufficiently favorable that use of polymer would generally not be needed to achieve an efficient reservoir sweep.) Second, for oil viscosities above 150 cp, the viscosity requirements to achieve a favorable mobility ratio were feared to reduce polymer solution injectivity to prohibitively low values (i.e., slow fluid throughput in the reservoir to the point that oil production rate would be uneconomically low). Several important changes have occurred since the previous screening criteria were proposed. First, oil prices increased to ~$70/bbl, while polymer prices remained relatively low ($0.90 to $2/lb for HPAM). Second, viscosification abilities for commercial polymers have increased, Thermal methods can’t be used for some viscous oils—because of thin zones, ambient cold, environmental constraints, permafrost, etc. Is polymer flooding viable for viscous oils? Old (1997) screening criteria for polymer flooding: •~150-cp oil was the upper limit because of (1) polymer costs and (2) injectivity losses. Changes since the old screening criteria: •Higher oil prices (~$70 versus ~$20/bbl). •Modest polymer prices ($1.50 versus $2/lb). •Greater use of horizontal wells. •Controlled injection above the parting pressure. partly from achieving higher polymer molecular weights and partly from incorporating specialty monomers (e.g., with associating groups, Buchgraber et al. 2009) within the polymers. Conventional wisdom from earlier polymer floods was that it was highly desirable to achieve a mobility ratio of unity or less (Maitin 1992). However, with current high oil prices, operators are wondering whether improved sweep from polymer injection might be economically attractive even if a unit mobility ratio is not achieved. In wells that are not fractured, injection of viscous polymer solutions will necessarily decrease injectivity. In order to maintain the waterflood injection rates, the selected polymer-injection wells must allow higher injection pressures. Another important change since the time when earlier screening criteria for polymer flooding were developed (Taber et al. 1997a, 1997b) has been the dramatic increase in the use of horizontal wells. Use of horizontal wells significantly reduces the injectivity restrictions associated with vertical wells, and injector/producer pairs of horizontal wells can improve areal sweep and lessen polymer use requirements (Taber and Seright 1992). Open fractures (either natural or induced) also have a substantial impact on polymer flooding. Waterflooding occurs mostly under induced fracturing conditions (Van den Hoek et al. 2009). Particularly, in low-mobility reservoirs, large fractures may be induced during the field life. Because polymer solutions are more viscous than water, injection above the formation parting pressure will be even more likely during a polymer flood than during a waterflood. The viscoelastic nature (apparent shear-thickening or “pseudo-dilatancy”) for synthetic EOR polymers (e.g., HPAM) makes injection above the formation parting pressure even more likely (Seright 1983, Wang et al. 2008a, Seright et al. 2009c). Under the proper circumstances, injection above the parting pressure can significantly (1) increase polymer solution injectivity and fluid throughput for the reservoir pattern, (2) reduce the risk of mechanical degradation for polyacrylamide solutions, and (3) increase pattern sweep efficiency (Trantham et al. 1980, Wang et al. 2008a, Seright et al. 2009c). Using both field data and theoretical analyses, these facts have been demonstrated at the Daqing Oilfield in China, where the world’s largest polymer flood is in operation (Wang et al. 2008a). Khodaverdian et al. (2009) examined fracture growth during polymer injection into unconsolidated sand formations. During analysis of polymer flooding in this work, we assume that injectivity limitations will require either use of horizontal wells or that polymer injection must occur above the fracture or formation parting pressure. Consequently, linear flow will occur for most of our intended applications. Polymer Flooding Considerations Many factors are important during polymer flooding (Sorbie 1991, Wang et al. 2008b). During design of a polymer flood, critical reservoir factors that traditionally receive consideration are the reservoir lithology, stratigraphy, important heterogeneities (such as fractures), distribution of remaining oil, well pattern, and well distance. Critical polymer properties include costeffectiveness (e.g., cost per unit of viscosity), resistance to degradation (mechanical/shear, oxidative, thermal, microbial), tolerance of reservoir salinity and hardness, retention by rock, inaccessible pore volume, permeability dependence of performance, rheology, and compatibility with other chemicals that might be used. Issues long recognized as important for polymer bank design include bank size (volume), polymer concentration and salinity (affecting bank viscosity and mobility), and whether (and how) to grade polymer concentrations in the chase water. For brevity, only a few of these factors will be addressed in this report. Rheology. Rheology in porous media is commonly raised as an issue during discussion of polymer flooding (AlSofi and Blunt 2009, Lee et al. 2009, Seright et al. 2009c). However, concern about polymer rheology must be tempered in view of injectivity realities. Achieving economic injectivities and fluid throughputs with polymer solutions (especially when displacing viscous oil) requires the use of either horizontal wells or fractured vertical wells (Seright 2009b, Seright et al. 2009c). With vertical fractures in vertical wells, fluid flows linearly away from the fracture. For horizontal wells, flow will be radial for a short time but soon become linear. For both horizontal wells and fractured vertical wells, the fluid flux will be quite low as the polymer enters the porous rock. For our target North Slope fields, we estimated flux values to be 0.01 to 0.2 ft/d for a vertically fractured injector, 0.2 ft/d as the fluid first enters the formation from a horizontal well, and 0.01 ft/d for most of the distance between two parallel horizontal wells. For this range of flux values, HPAM polymer solutions show Newtonian or near-Newtonian behavior (Seright et al. 2009c, Seright et al. 2010). Fig. 1 illustrates this point for a HPAM polymer (6-8 million Daltons) in cores with permeabilities ranging from 17.5 md to 5,120 md. The x-axis plots a parameter that allows flux values (u, in ft/d) to be correlated for different permeabilities (k) and porositites (). The large arrow provides a point of comparison for different permeabilities. Over the practical range of permeabilities and velocities anticipated for our target fractured or horizontal wells (0.01 to 0.2 ft/d), HPAM polymer solutions (in representative North Slope brines at the target reservoir temperature) exhibit nearly constant resistance factors (effective viscosity in porous rock relative to water) (Seright 2009b, Seright et al. 2009c, Seright et al. 2010). These observations simplify our analysis of the potential for polymer flooding using HPAM. (Discussion of a strongly shear-thinning fluid, such as xanthan, will be included later.)