Qualification of New Methods for Measuring In Situ Rheology of Non-Newtonian Fluids in Porous Media

摘要

Pressure drop (Δ ) versus volumetric injection rate ( ) data from linear core floods have typically been used to measure in situ rheology of non-Newtonian fluids in porous media. However, linear flow is characterized by steady-state conditions, in contrast to radial flow where both pressure and shear-forces have non-linear gradients. In this paper, we qualify recently developed methods for measuring in situ rheology in radial flow experiments, and then quantitatively investigate the robustness of these methods against pressure measurement error. Application of the new methods to experimental data also enabled accurate investigation of memory and rate effects during polymer flow through porous media. A radial polymer flow experiment using partially hydrolyzed polyacrylamide (HPAM) was performed on a Bentheimer sandstone disc where pressure ports distributed between a central injector and the perimeter production line enabled a detailed analysis of pressure variation with radial distance. It has been suggested that the observed shear-thinning behavior of HPAM solutions at low flux in porous media could be an experimental artifact due to the use of insufficiently accurate pressure transducers. Consequently, a generic simulation study was conducted where the level of pressure measurement error on in situ polymer rheology was quantitatively investigated. Results clearly demonstrate the robustness of the history match methods to pressure measurement error typical for radial flow experiments, where negligible deviations from the reference rheology was observed. It was not until the error level was increased to five-fold of typical conditions that significant deviation from the reference rheology emerged. Based on results from pore network modelling, Chauveteau (1981) demonstrated that polymer flow in porous media may at some rate be influenced by the prior history. In this paper, polymer memory effects could be evaluated at the Darcy scale by history matching the pressure drop between individual pressure ports and the producer as a function of injection rate (conventional method). Since the number of successive contraction events increases with radial distance, the polymer has a different pre-history at the various pressure ports. Rheology curves obtained from history matching the radial flow experiment were overlapping, which shows that there is no influence of geometry on in-situ rheology for the particular HPAM polymer investigated. In addition, the onset of shear-thickening was independent of volumetric injection rate in radial flow.