AbstractsEarth & Environmental Science

Pressure Behaviour and Fracture Development during Polymer Injection in a Heavy Oil Saturated Unconsolidated Sand:

by R. Logister




Institution: Delft University of Technology
Department:
Year: 2014
Keywords: Polymer; Pressure; Fracture; Petroleum; Injectivity
Record ID: 1268370
Full text PDF: http://resolver.tudelft.nl/uuid:9bdb62d9-e5fb-44f7-a944-d87f5497eaa2


Abstract

Polymer flooding is an important EOR method in heavy oil reservoirs, where field studies showed a recovery up to 59%. However, accurate assessment of injectivity of viscous polymer solutions can give major challenges. Little research has been devoted to pressure build-up phenomena with polymer injection in porous media. Therefore, the objective of this study is to experimentally investigate the water and polymer injectivities in an unconsolidated sand pack saturated either with brine or with heavy oil, subject to confining axial and radial stresses in the range of those encountered in reservoirs. In particular we are interested in the typical reservoir conditions under which failure/hydraulic fracturing of the sample occurs. We also intend to investigate the effect such a hydraulic fracture will have on polymer injectivity. The injection experiments were conducted in a biaxial pressure vessel containing a porous medium, in which flow rates can be controlled and pressures can be measured. Data from rheological measurements and a single phase water flood experiment carried out in this study were utilised for understanding and comparing the injectivity results for polymer flood experiments. The outcome of experimental analysis showed that the viscous nature of the polymer solution and the viscosity of the oil are the main causes for the loss in injectivity compared to a water flood. A change in confining pressure does not influence the injectivity. However, it does affect the occurrence of fracturing. It was found fracturing occurred at a pressure drop of 4.3 times the minimum confining pressure and caused injectivity to rise. Pressure data obtained from the water flood experiments was fitted with prediction models that describe pressure behaviour during injection in a porous medium. Fitting resulted in a value for permeability of the porous medium and was used in prediction models that describe a polymer flood. Polymer flood data gave a best fit when polymer with a constant average viscosity was assumed to have displaced all saturation fluid. The rheology showed that the polymer solution has shear thickening behaviour, making a non-Newtonian model that only considers shear thinning invalid. A displacement model, which describes the displacement of water by a polymer solution with constant viscosity, made clear that retention effects are highest close to the wellbore and decrease when polymer is further invaded in the porous medium. Finally, the results from this study lead to recommendations to refine future experimental polymer flood of this scale.