Evaluation of Polymers for Heavy Oil EOR
Saskatchewan Research Council
The Saskatchewan Research Council (SRC) has investigated various aspects of polymer flooding of heavy oil for numerous years under both exclusive and multiclient consortium contracts. Projects have had a strong focus on field application and have included rheological characterization of various types of polymers (including relatively new hydrophobically associating polymers), adsorption tests, evaluation of inaccessible pore volume, core displacement tests with different oil types, and numerical simulation by history matching of results. This technical report summarizes some of the most important findings accumulated over this period and discusses guidelines for laboratory evaluation of polymer solutions of high viscosity that are typically considered for enhanced oil recovery (EOR) in heavy-oil fields.
Polymer Flooding in Western Canada
SRC analyzed publicly available data for chemical flooding projects in western Canada to evaluate the technical success of these projects and identify factors that correlated with their performance (Renouf 2014). Reservoir parameters that were found to have the most significant effect on the success of polymer floods were vertical permeability and pay thickness. A higher ratio of horizontal producers and injectors also correlated favourably with project success. Renouf (2014) found that any failure of chemical flooding projects was most often not caused by the design or concept but rather by operational problems; this was particularly true for projects involving alkali and/or surfactant solutions. Polymer degradation (mechanical, chemical, and biodegradation) and interactions with clay were found to be commonly occurring problems. Polymer injectivity issues were also fairly common.
Careful chemical design is recognized as being of utmost importance for the success of field EOR processes. American Petroleum Institute’s Recommended Practice (API RP) 63 describes laboratory polymer preparation and testing procedures in great detail. However, since 1990, when these procedures were developed, there have been significant changes in the properties of polymers used for EOR and the types of fields where polymer EOR is applied. Viscosity of polymer solutions used in EOR processes has increased dramatically, owing both to the use of a higher concentration of polymers and to the intrinsic properties of the newly developed polymers (very high molecular weight exceeding 20 million Dalton, inclusion of various hydrophobic monomers, and variations in the structure of the polymer chain). This has necessitated some revisions in the polymer selection and evaluation process.
Polymer Preparation in the Laboratory
Most polyacrylamides are supplied in a form of dry powder and need to be hydrated before they can be used. A lot of variables (e.g., mixing speed, type of mixing blade, and size of beaker) can impact the hydration process by changing the mixing pattern in the beaker and the shear applied to polymer particles that are being hydrated. Too much shear can potentially degrade the polymer, while too little will cause polymer granules to clump and, as a result, form “fish eyes,” lenses of unhydrated polymer, in which the outer layer prevents water penetration and stops the process of hydration.
The SRC tested polymer hydration using several mixing speeds between 250 and 700 rev/min (with a three-blade, propeller-type 60-mm stainless steel stirrer) and prepared a 3000-ppm solution of a high molecular weight hydrolyzed polyacrylamide. The polymer mixed at 700 rev/min was on average 15% less viscous than that mixed at 250 rev/min, indicating that it was being mechanically degraded at high mixing speed. This confirms that high mixing speeds should be used with caution, especially when working with high-molecular-weight polyacrylamide. An optimal mixing speed for most of the tested polymer appears to range between 200 and 400 rev/min (300 rev/min or more for hydrophobically associating polymer).
The vast majority of polymer solutions used in EOR processes are non-Newtonian fluids. Their viscosity is usually measured using a rheometer with a simple geometry (e.g., cup and bob or cone and plate) and approximated by the Carreau model. However, in porous media, the rheological behaviour of a polymer solution often deviates markedly from the model. Two main reasons for this are polymer adsorption and the resulting modification of permeability profile and polymer elasticity. The latter has the greatest effect at high shear rates (e.g., often existing at and near the wellbore of an injection well), when a polymer molecule does not have enough time to relax while flowing from one pore throat to another. This deviation can result in significantly higher injection pressures in the field, as well as potentially increasing mechanical degradation of the polymer. It is therefore important that the apparent viscosity of the selected polymer solution is evaluated in the reservoir rock over the range of expected flow rates. This becomes crucial for hydrophobically associating polymers (HAPs) and traditional high-molecular-weight hydrolyzed polyacrylamides (HPAMs) that typically have more pronounced viscoelastic properties.
The SRC found that polymer adsorption in porous media greatly depended on the polymer and rock type. Adsorption typically increased with increasing polymer molecular weight. For sands from predominantly sandstone formations (that typically have negative surface charge), adsorption was significantly higher for cationic polymers. Adsorption behaviour of hydrophobically associating polymers was also markedly different from that of traditional polyacrylamides. HAPs appeared to adsorb in multiple layers, but the association between such layers was relatively weak and could be broken at high shear rates.
Standard filter ratio testing (in accordance with API RP 63) achieves limited throughput values of the polymer, significantly lower than can be expected around the vertical injectors in the field. The filterability test developed by Seright (2008) overcomes this limitation and allows injectivity of the polymer to be evaluated under more realistic conditions. The size of the filter membranes used in the test can be adjusted to better match the permeability of the reservoir. Whenever possible, a final injectivity verification test should be conducted using a reservoir core representing the lowest-permeability areas of the formation with polymer prepared in fresh reservoir brine to avoid surprises during field operations.
Long-term stability measurements are important to verify the longevity of the polymer solutions in reservoir conditions. Such tests are typically conducted for a period of 1 month to 1 year. It was found that polymer solution can sometimes increase with time because of hydration of the amine moiety; however, this often leads to a rapid chemical degradation once more than 30 to 40% of the polymer monomers are hydrolyzed. Fresh reservoir brine should be used for these tests to ensure that bacteria that are naturally contained in the reservoir do not degrade the polymer. We found that chemical and/or bacterial degradation of a polymer can be significant and varies greatly for different types of polymers and with varying composition of a specific reservoir brine.
Effect of pH
We evaluated the effect of pH level on the viscosity of solutions of two polymers: traditional straight-chain-polyacrylamide and the hydrophobically associating polymer. The solutions’ pH levels were adjusted using both hydrochloric and citric acids. Several trial batches of polymer solutions were prepared and tested at different pH levels. The viscosity of the polymer solutions was reduced by more than one order of magnitude as the pH was brought down to 1.5 to 2. Most of the viscosity was then restored after neutralization of acid with sodium hydroxide.
SRC has conducted many coreflood experiments over the last several years to evaluate polymer flooding. One of the most interesting findings was that microscopic recovery factors were often similar for polymer solutions that had up to an order of magnitude difference in viscosity for corefloods conducted at a relatively slow displacement-front velocity (Fig. 1). Residual oil saturation of the rock appeared not to be affected by the changes in the type or viscosity of the polymer solution. This contradicts some of the previously published results but is consistent with the theory that polymer flooding improves only macroscopic (sweep) efficiency but does not impact microscopic (pore-level) displacement. Levitt et al. (2011) observed similar phenomena in their work. Heterogeneous formations may also possibly benefit from use of polymer solutions of higher viscosity and/or elasticity.
Linear corefloods experiments are an industry standard for the evaluation of various displacement processes. These experiments are typically conducted with cylindrical models of porous media measuring 2.5 to 5 cm in diameter and 15 to 30 cm long. It is usually accepted that these experiments are 1D (i.e., oil displacement is piston like, and macroscopic (sweep) displacement efficiency is close to unity). This assumption, while acceptable for experiments with light oil in lower-permeability cores, fails to take into account the significant instability of the displacement front and the resulting viscous fingering phenomena that occur because of the dramatic difference in viscosities between water and heavy oil, and the much higher permeability in heavy-oil reservoirs. It is important to keep this in mind while using data obtained in linear coreflood experiments with heavy oil for numerical simulation and predictions of process performance in the field.
Polymer flooding can be successful in recovering heavy oils with viscosity up to 10 000 mPa∙s and even higher. Several successful polymer EOR projects are currently ongoing in western Canada. Careful design that avoids or minimizes potential operational issues is a key to technical and economic success in the field. Procedures for polymer evaluation in the laboratory need to be modified and expanded to account for differences among polymer types and viscosity as well as specific conditions and operational realities that exist in various fields.
Levitt, D., Jouenne, S., Bondino, I., Gingras, J.P., and Bourrel, M. 2011. The Interpretation of Polymer Coreflood Results for Heavy Oil. Presented at the SPE Heavy Oil Conference and Exhibition, Kuwait City, Kuwait, 12–14 December. SPE-150566-MS. http://dx.doi.org/10.2118/150566-MS.
Renouf, G. 2014. A Survey of Polymer Flooding in Western Canada. Presented at the SPE Improved Oil Recovery Symposium, Tulsa, 12–16 April. SPE-169062-MS. http://dx.doi.org/10.2118/169062-MS.
Seright, R.S., Seheult, M., and Talashek, T. 2008. Injectivity Characteristics of EOR Polymers. Presented at the 2008 SPE Annual Technical Conference and Exhibition, Denver, 21–24 September. SPE-115142-MS. http://dx.doi.org/10.2118/115142-MS.