Effect of Polymers on the Imbibition Process: A Laboratory Study

Abstract

Summary Polymer flooding of fractured reservoirs is common. In water-wet fracturedreservoirs, the primary recovery mechanism may be imbibition. This paperpresents results of an experimental study investigating the effect of polymerson the imbibition process. Two sets of experiments, static and dynamic, wereperformed. The static experiments showed that the amounts of oil ultimatelyrecovered by water and polymer-solution imbibition are practically equal. Therate of oil recovery by the polymer solutions, however, is always less than therate with water. The dynamic experiments consisted of flooding oil-saturatedfractured cores through the fractures. The oil-recover behavior in theseexperiments depended not only on the rate of injected fluid imbibition from thefracture into the matrix blocks, but also on the operating injection rate andthe efficiency of the injected fluid in displacing the oil in the fracture. Introduction Excessive water channeling through the high-permeability fracture system ofthe Spraberry field prompted Atlantic Refining Co. researchers to explorewater-imbibition displacement as a production mechanism for recovering oil. Since then, theoretical work and field experience have supported this idea andhave found that water imbibition is not only important, but could be the mainmechanism for recovering oil from fractured reservoirs. In water-imbibition displacement, water flows by capillary forces from thefractures into the oil-saturated, water-wet matrix blocks, causing oil to flowcountercurrently from the matrix blocks into the fractures. This oil then isdisplaced by oncoming water through the fracture system to the productionwells. Polymer flooding is known to offer a better alternative over existingwaterflooding in conventional reservoirs if the mobility ratios are unfavorableor if significant permeability variations exist. This knowledge also makespolymer flooding an attractive process to apply in fractured reservoirs. Polymer flooding would offer better mobility control in the displacement of oilin these reservoirs. Through polymer retention, polymer flooding also couldredirect the flow of injected polymer solution polymer flooding also couldredirect the flow of injected polymer solution from one fracture to another andfrom the fractures into the matrix blocks. Lastly, at any injection rate, polymer flooding would generate more viscous forces than waterfloodingperformed at the same rate. Problem Problem Polymer flooding in fractured reservoirs, or in anyheterogeneous Polymer flooding in fractured reservoirs, or in any heterogeneousreservoir, offers some advantages but also raises some senous questions thathave not been investigated:(1)do polymer macromolecules retard the imbibition process;(2)if so, will the polymer concentration, polymer molecule size, andpolymer-solution salt content play a role in this effect; and(3)if polymerretards the play a role in this effect; and(3)if polymer retards theimbibition process, will the extra viscous forces generated and the bettermobility control of the polymer flooding make up for the imbibitionretardation? The purpose of this study was to provide some answers to these questions, Two sets of experiments were performed: static and dynamic. The staticimbibition experiments involved oil displacement from the matrix blocks infractured reservoirs by imbibition into the fracture system. The dynamicimbibition experiments involved both displacement from the core by imbibitionand displacement of oil by the oncoming injected fluid through the fracturesystem. A literature search found no papers dealing with the effect of polymers onthe imbibition process. Only an example of the data collected polymers on theimbibition process. Only an example of the data collected is illustrated in thefigures. The complete data set is available elsewhere. Experimental Fluids and Cores The four different brines (0, 1,000, 5,000, and 10,000 ppm NACl) used inthis work were all synthetic and prepared in the laboratory. Four oil typeswere used. The first three were synthetic and prepared in the laboratory. Type1 is a 100% n-decane with a viscosity of 0.95 cp. Type 2 is prepared by mixing25 vol% n-decane with 75 vol% mineral oil with a viscosity of 6.5 cp. Type 3 isa 100% mineral oil with a viscosity of 18.3 cp. Type 4 is a natural crude oilwith a viscosity of 143.9 cp. All viscosities were measured at roomtemperature. Partially hydrolyzed polyacrylamide (PHPA) polymers of three differentmolecular weights were used. Polymer L is a low-molecular-weight (2 × 10), Mis a medium-molecular-weight (5 × 10), and H is a high-molecular-weight (11 ×10) polymer. These polymers were used to prepare solutions differing inmolecular weight, concentration, and salt content. Each solution was given acode name accordingly. For example, Solution L460,0 is Polymer L (low molecularweight) of 460-ppm polymer concentration and 0-ppm salt content. All theexperimental fluids underwent physical property measurements. Detailed tablesand graphs of these property measurements. Detailed tables and graphs of thesemeasurements are available elsewhere. 10 The interfacial tension (IFT)measurements showed that the polymer molecular weight or concentration haslittle to no effect on the values measured between the polymer solutions andcertain oils. The salt content of the solu-tions, on the other hand, was foundto have some effect on the measured values. The measured IFT's increasedslightly with the solutions' salt content. Berea sandstone cores with porosities ranging between 0.22 and 0.23 wereused in these experiments. The permeability ranged between 400 and 840 md. Thegeneral dimensions of the cores were 2 in. in diameter and 51/2 in. in length. Ref. 10 gives exact core dimensions and properties. The cores were fired tostabilize any clay materials present in the rock pore space and to achievestrong water-wet conditions. Treating the cores so that they were stronglywater-wet allowed comparison of tests performed with different cores withoutcore wettability being a strong factor. To avoid plugging, the polymers used in this study had molecular sizes lessthan 30 % of the average pore-throat size of the porous medium. The average pore-throat size of the porous medium was determined to be 13um. The size of the polymer random coil can be expressed in terms ofstatistical parameters such as the root-mean-square distance between its ends,(r 2) 1/2, and its radius of gyration and the root-mean-square distance of theelements of the chain from its center of gravity, (S 2) 1/2. Both thesestatistical parameters can be estimated with the polymer viscosity measurementsand the Polymer's molecular weight and configuration. Polymer's molecularweight and configuration. The hydrodynamic radius of gyration of Polymers L, M, and H in deionized water was found to be 0.237, 0.344, and 0.504 um, respectively. The statistical parameter, end-to-end distance of Polymers L, M, and H molecular coils, was found to be 1.419, 2.066. Polymers L, M, and Hmolecular coils, was found to be 1.419, 2.066. and 3.024 Am, respectively. These values suggest that all Polymers L, M, and H can be applied in thiswork's porous media without plugging problems.