Improvement in Heavy-Oil Reservoir Evaluation Using Nuclear Magnetic Resonance: Long Lake and Kinosis SAGD Projects, Alberta, Canada
- M. Cheng (Nexen Energy ULC) | A. Kotov (Baker Hughes) | K. Pyke (Nexen Energy ULC) | A. Hanif (Baker Hughes)
- Document ID
- Society of Petrophysicists and Well-Log Analysts
- Publication Date
- June 2015
- Document Type
- Journal Paper
- 239 - 250
- 2015. Society of Petrophysicists & Well Log Analysts
- 2 in the last 30 days
- 251 since 2007
- Show more detail
The Lower Cretaceous McMurray Formation in Alberta is the primary bitumen-producing horizon in the Long Lake and Kinosis SAGD (steam-assisted gravity drainage) development projects. Conventional log analysis methods used to evaluate this heavy-oil reservoir employ a resistivity-based water-saturation (Sw) equation. These methods are challenged by the vertical and lateral variability in formation water resistivity (Rw), which is often unknown to the log analyst. In order to overcome this limitation, log-derived saturations are calibrated to Dean-Stark (DS) extraction results. However, not all wells are cored and this leaves a high level of uncertainty in the log-calculated Sw in these wells. In addition, the core-measured porosities are often found higher than the corresponding wireline density porosity and the validity of core-measured saturation becomes questionable.
The nuclear magnetic resonance (NMR) log, which is insensitive to variations in water salinity, provides an alternative to core-calibrated log saturation. The McMurray oil sands environment is favorable to NMR logging due to low borehole fluid invasion, low gas-oil ratio, high viscosity, low formation temperature and simple sand/shale lithology. Saturation analysis relies on the fact that the NMR T2 signal of viscous organic matter relaxes and decays rapidly before the downhole instrument performs any measurement, resulting in an apparently low NMR porosity. This porosity deficiency is used to estimate bitumen volume in the reservoir.In this paper, we present examples, with comparison of log and core results from a Kinosis well, to illustrate the methodology, which provides consistent saturation determination that is independent of formation water salinity. Probabilistic uncertainty analysis shows the NMR model provides a reliable estimate of bitumen saturations in clean and moderate shaley sands. However, uncertainty increases with shale content due to overlapping of shale and bitumen T2 signals, which is a limitation of the model. The producing reservoir in Long lake and Kinosis is relatively clean and the log-derived shale volume is generally less than 30%. High confidence in the NMR analysis allows for optimization of the coring program and reduction in field development costs.
|File Size||1 MB||Number of Pages||12|