An Experimental Study of Non-Newtonian Polymer Rheology Effects on Oil Recovery and Injectivity
- Richard W. Gleasure (U. of Toronto)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- November 1990
- Document Type
- Journal Paper
- 481 - 486
- 1990. Society of Petroleum Engineers
- 2.4.3 Sand/Solids Control, 5.4.7 Chemical Flooding Methods (e.g., Polymer, Solvent, Nitrogen, Immiscible CO2, Surfactant, Vapex), 5.3.1 Flow in Porous Media, 1.6.9 Coring, Fishing, 5.2 Reservoir Fluid Dynamics, 5.4.1 Waterflooding
- 2 in the last 30 days
- 526 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
Pseudoplastic non-Newtonian polymer solutions were examined for theirenhanced oil recovery performance. Detailed results are reported for xanthangum (XAN), Kelzan XCD, and a viscoelastic polyethylene oxide (PEO), PolyoxOF-50. Increases in the power-law coefficient resulted in improved displacementefficiency. Effects were also observed in the injectivity-index parameterresults.
Most polymers proposed for polymer flooding, regardless of their chemicalorigin, display non-Newtonian fluid behavior to some degree. The apparentviscosity of the polymer solution is a function of the shear rate to which itis subjected. At shear rates encountered at reservoir conditions (0.1 to 100seconds), polymers typically exhibit shear thinning or pseudoplastic behavioraccording to tangential shear measurements. Certain synthetic polymers,particularly the PEO type and, to a lesser extent, partially hydrolyzedpolyacrylamides (PHPA), show dilatant- or shear-thickening-type responsesduring flow through porous media, despite demonstrating pseudoplasticity in atangential shear viscometer. Numerous researchers have explained this behaviorin terms of a viscoelastic contribution to the overall non-Newtonian propertiesof these polymers. The displacing-fluid viscosity has been shown to be of majorimportance to the outcome of an immiscible displacement process as quantifiedby the overall displacement efficiency, the product of displacementefficiencies at the microscopic and macroscopic levels. The microscopicdisplacement efficiency attained in a flood is reflected in the residual oilsaturation. Sor, achieved. Correlations for predicting Sor for various porousmedia have shown that increasing the displacing-fluid viscosity can improvemicroscopic displacement efficiency. Macroscopic displacement efficiency islargely a function of the mobility contrast between displaced- and displacingfluid phases and is commonly expressed in terms of the well-known mobilityratio, M, where
When M is less than 1, the displacement takes place in a pseudo-piston-likemanner with a stable displacement front. Under these circumstances, sweepefficiency with respect to the volume of displacing fluid injected is high.Therefore, increasing the displacing-fluid viscosity obviously benefits themacroscopic displacement efficiency. For immiscible polymer flooding, themacroscopic velocity, and hence shear-rate, profiles that develop withinjection/production well patterns influence polymer viscosity and overalldisplacement efficiency because of the non-Newtonian polymer rheology. In thevicinity of injection or production wellbores, fluid velocities are highest,but rapidly decline as the distance from the wellbore increases. In idealhomogeneous reservoirs, velocities also tend to be more elevated along thediagonals connecting injector to producer. For pressure-constrained injectivityoperations, the velocity profiles also will be influenced by the transient flowcharacteristics encountered during the flood. The displacing-fluid injectionrates will vary, depending on the effective mobility of the displacing fluid ata given stage in the flood. Hence, different operating practices for a givenflood-injection flow rates or pressures might yield different performanceoutcomes. The goal of this study was to investigate the effects ofnon-Newtonian displacing-fluid rheology on the oil-displacement process.Oil-displacement experiments involving two polymer types were carried out underconditions representing pressure-constrained injectivity operations.Specifically, the effect of shear rate on the oil-displacement process wasexamined by varying pressure gradients in linear-geometry corefloods. Thepolymer solutions were also characterized in terms of shear stress andinterfacial tension (IFT).
The two polymers examined in this study were a XAN and a PEO. The PEOpolymer was Polyox OF-50 manufactured by Union Carbide Corp. The XAN polymerwas a Kelzan XCD sample supplied by the Kelco Rotary Co. Polymer solutions wereprepared according to recommended procedures provided by the polymermanufacturers. To simulate a realistic aqueous solvent, a brine with acomposition similar to that of the Canadian Pembina field polymer pilot wasused in this study. Polymer solutions were characterized rheologically by useof a Couette viscometer, the Haake Rotovisco RV100/CV100. This viscometer,fitted with a Mooney-Ewart ME 45 coaxial cylinder sensor, is capable ofmeasuring apparent viscosities of dilute polymer solutions over a shear-raterange of 0 to 300 seconds-1 with an upper/lower bound on shear stress of 8000and 10 mPa, respectively. The rheological data obtained in this manner werecharacterized with the well-known de Waele (power-law) function forviscosity:
The displacement experiments were performed on an unconsolidated porousmedium in the form of Grade F75 Ottawa quartz sand. The average sand porositywith respect to the brine was 38.6%. The oil used in this study was a mineraloil with a viscosity of 30 mPa.s and a specific gravity of 0.855-correspondingto 0.85 g/cm3 at 25 degrees C. IFT's of the displacing-fluid/oil systems weredetermined from measurements carried out at 25 degrees C with a Fischer de Nuoytensiometer. Two core holders of Lucite pipe equipped with detachablestainless-steel end sections were used in the linear coreflood experiments.Core Holder 1 was 38.3 cm long and 4.5 cm in diameter, and had a PV of 235 mL.Core Holder 2 was 38.6 cm long and 4.7 cm in diameter, with a PV of 258.5 mL.The equipment layout consisted of a pressurized vessel connected to the linearcore holder from which fluid effluent samples were collected during thedisplacement experiments. A typical coreflood proceeded as follows. First, avibrating shaker was used to wet pack the core holder with sand. The coreholder was then sealed, and the core permeability to brine, k, was determinedby flow-rate/pressure-drop measurements. The core was then saturated with oilto produce an interstitial water saturation, Siw, at which time the corepermeability to oil, ko(Siw), and the oil mobility, Lambda o(Siw), at thissaturation were calculated. The core was then displaced with a polymer solutionat a fixed imposed pressure gradient. Effluent samples were taken periodicallywere taken at the core outlet to determine flow rates and phase compositions.Floods were carried out with polymer solutions at concentrations of 2,500,1,500, and 500 ppm. For comparative purposes. a base-case waterflood also wascarried out.
|File Size||459 KB||Number of Pages||6|