Full text loading...
-
A Scientific Approach for Generating a Rheology Model to Simulate Polymer Injectivity
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, IOR 2019 – 20th European Symposium on Improved Oil Recovery, Apr 2019, Volume 2019, p.1 - 12
Abstract
Modelling polymer injection at near-wellbore conditions is challenging, as it is strongly affected by Non-Newtonian polymer rheology among other parameters. During polymer injection, viscosities vary significantly near the wellbore where flow velocities and therefore shear rates are high. Current commercial reservoir simulators have limited capabilities in capturing this behaviour. Modification of properties on each grid around the injector including reducing fluid viscosity, increasing permeability along with building extremely fine grid is often performed in the simulation. However, this results in limited prediction capability and will be inefficient for full field simulation where multiple injectors with different properties and rates must be considered. This paper presents both; a workflow to generate an appropriate rheology model using viscometer and core flood data, and polymer injectivity simulation.
Viscosity vs. shear rate and viscosity vs. velocity data has been generated from rheometer and core floods at different velocities respectively. Data is then plotted together after converting core flood velocities into shear rate. A correction factor is established by matching viscosity at high shear rate regimes. Based on this, a rheology model for highly shear thinning biopolymer Schizophyllan was generated using Carreau-Yasuda correlation. The rheology model was then used to simulate and match the bottomhole pressure (BHP) response of a recently conducted single well test in 2017 using conceptual radial and actual Cartesian grid model. The matching was achieved with and without grid refinement for the Cartesian model while correcting the skin factor for the grid size (Behr et al.). Matching exercise required numerical tuning due to highly shear thinning behaviour. Additionally, the same rheology model was validated by matching the multi-well pilot injector BHP for a longer period without any near wellbore modification. In contrast, earlier matching attempts had required multiple modifications in either viscosity or permeability at different time periods with progressing flood.
The newly generated rheology model accounts for both viscometer and core flood data and represents the polymer behaviour much closer to reservoir performance. The results from the single- and multi-well polymer injection simulation showed a decent history match without any near wellbore grid property modification.
The workflow to generate a rheology model and deriving a shear correction factor is relatively novel for this biopolymer. The advantage of such a rheology model becomes more distinct for the simulation and history matching of a full field scale polymer injection with multiple injectors, as the overall process can be simplified.