Mechanistic Modelling of Radial Polymer Flow in Porous Media

Abstract

Polymer flooding is one of the most successful chemically enhanced oil recovery (EOR) methods, and has primarily been implemented to accelerate oil production by sweep improvement. However, research from the last couple of decades have identified several additional benefits associated with polymer flooding. Firstly, improved polymer properties have extended their use in reservoirs with high temperature and high salinity. Secondly, improved understanding of the viscoelastic flow behavior of flexible polymers have revealed that they may in some cases mobilize capillary trapped oil as well. Despite of the recent progress, extensive research remains to quantify the appropriate flow mechanisms and accurately describe polymer flow in porous media.

Simulations and history match operations performed in this thesis are aimed at improving the modelling of radial polymer flow in porous media. This has been achieved by (1) evaluating the accuracy and robustness of two different history match methods which are used to estimate the in-situ rheology of non-Newtonian fluids in radial flow, (2) investigating potential rate and memory effects (at the Darcy scale) of viscoelastic polymer solutions in radial flow, and (3) quantifying polymer in-situ rheology and polymer injectivity.

The accuracy and robustness of both history match methods which are used to estimate the in-situ rheology of non-Newtonian fluids in radial flow was clearly demonstrated in radial flow experiments where effective (or cumulative) error was below 5 % of the maximum preset transducer pressure range. Thereby, the observed shear-thinning behavior of partially hydrolyzed polyacrylamide (HPAM) at low flux in porous media could not be attributed to insufficiently accurate pressure transducers during in-house flow experiments, as suggested by some researchers.

The estimation of polymer in-situ rheology showed invariance between excluding and including the polymer pressure data outside the near-wellbore region. Thus, it was proposed that the polymer in-situ rheology is mainly defined by the pressure data originating from the near-wellbore region during radial polymer flow. Results showed that not only could the polymer in-situ rheology be (quantitatively) estimated from measurements of stabilized pressure, but could also be (qualitatively) identified from the pressure build-up during radial flow experiments. Consequently, the anchoring data from the pressure build-up during radial polymer flow was proposed as an additional tool for history matching field injectivity tests.

Rate and memory effects (at the Darcy scale) of several HPAM polymers were investigated in flow through Bentheimer sandstone discs. Results showed that no rate effects occurred for mechanically undegraded polymer (Flopaam 3330S). However, rate effects were observed for mechanically degraded polymers (Flopaam 3630S and Flopaam 5115SH) where the onset of shear-thickening increased with volumetric injection rate. While memory effects (at the Darcy scale) were absent for the mechanically undegraded and relatively low molecular weight polymer, Flopaam 3330S, the mechanically degraded and relatively higher molecular weight (18 MDa) polymer, Flopaam 3630S, exhibited memory effects in which apparent viscosity decreased with radial distance. As mechanical degradation is suggested to be confined to the near-wellbore region in radial polymer flow, the memory effect was proposed to originate from the elastic properties of the polymer.

In accordance with recent literature, the in-situ rheology of HPAM was shown to depend on flow geometry. During single and two-phase polymer flow, the shear-thinning behavior of HPAM was much more pronounced, and the extent of shear-thickening significantly reduced in radial compared to linear flow. Furthermore, the onset of shear-thickening during single-phase flow occurred at significantly higher velocities in radial relative to linear flow. However, this behavior was not consistent during two-phase flow as the onset of shear-thickening during linear and radial polymer flow coincided. Moreover, comparative studies of polymer flow in radial versus linear flow geometries during single and two-phase flow revealed that the impact of oil was to reduce apparent in-situ viscosity of HPAM. The low-flux in-situ rheology behavior was addressed and showed Newtonian behavior in linear flow while significant shear-thinning was observed during radial flow. Thus, both flow geometry and presence of oil were suggested to be key factors for estimating polymer in-situ rheology.