摘要

Carbonated water injection (CWI) is a CO2-augmented waterflooding technique for improving oil recovery and CO2 storage. The oil recovery and CO2 storage benefits of CWI as compared to plain (conventional) waterflood in secondary and tertiary recovery modes were investigated experimentally and numerically through a series of coreflood experiments and detailed compositional simulation. A reservoir core, stock tank crude oil and seawater were used in the tests at temperature and pressure typical of real reservoir conditions. We examined the behaviour of the dissolved CO2 by observing the CO2 front propagation as this affects the effectiveness of CWI as an oil recovery method. We also evaluated the capabilities and limitations of a commercial compositional flow simulator in modelling the CWI process.
The coreflood tert results show that CWI yields higher oil recovery than the conventional water flooding in both secondary and tertiary recovery methods. CWI shows a high potential as a CO2 storage injection strategy, as demonstrated by the relatively high percentage i.e., about 45-51%, of the total volume of CO2 injected (in carbonated water) stored at the end of the secondary and tertiary carbonated waterfloods. The experimental observation also revealed that in the secondary CWI process, the CO2 moves ahead of the carbonated water front which was not totally depleted of its CO2 content indicating good delivery of CO2 during the displacement This is evidence of dispersion and diffusion of CO2 from the injected carbonated water into the oil at the front.
The existing commercial numerical simulators cannot predict this behaviour resulting in a poor match between experimental and simulated recoveries. The oil recovery from CWI predicted by the simulator was higher than that reported experimentally. The instantaneous equilibrium and complete mixing assumptions inherent in the commercial simulator used (and generally in other compositional simulators as well) are not appropriate for modelling the local non-equilibrium process as demonstrated at the core scale and this might also be at the larger reservoir scale. Using the commercially available reservoir simulators would lead to inaccurate evaluation of CWI. Thus a new simulation approach is required.

  • 出版日期2011-4