HR: 0800h
AN: GP41B-0882 INVITED     [Abstracts]
TI: Joint Inversion of Marine Seismic and CSEM Data for Fluid Saturation Prediction
AU: * Hoversten, G M
EM: gmhoversten@lbl.gov
AF: Lawrence Berkeley National Laboratory, One Cyclotron Rd Mail Stop 90-1116, Berkeley, CA 94720 United States
AU: Gasperikova, E
EM: EGasperikova@lbl.gov
AF: Lawrence Berkeley National Laboratory, One Cyclotron Rd Mail Stop 90-1116, Berkeley, CA 94720 United States
AU: Chen, J
EM: jchen@lbl.gov
AF: Lawrence Berkeley National Laboratory, One Cyclotron Rd Mail Stop 90-1116, Berkeley, CA 94720 United States
AU: Newman, G
EM: gnewman@lbl.gov
AF: Lawrence Berkeley National Laboratory, One Cyclotron Rd Mail Stop 90-1116, Berkeley, CA 94720 United States
AB: Recent developments in the application of controlled source marine electromagnetic (CSEM) data in petroleum exploration have brought this technology to the attention of many in exploration and production within the oil and gas industry. These developments are founded on more than two decades of research carried out in academia and at U.S. national laboratories. The commercial availability of CSEM data now makes it possible to consider integrating this new data with existing seismic data in ways that will add considerable value. In particular, the sensitivity of CSEM data to water saturation (Sw), when combined with the spatial and reservoir parameter sensitivity (porosity, Sw, gas saturation [Sg], and oil saturation [So]) of seismic data, can provide enhanced prediction of fluid saturations within existing or prospective reservoirs. There are many ways in which CSEM and seismic data can be combined to estimate reservoir parameters. The possibilities range from what we term cooperative inversion, in which both data sets are used without any formal linkage in the inversion of either, to fully coupled joint inversion, in which both data sets are inverted simultaneously to directly estimate reservoir parameters. Hoversten et al. (2003) present an example of the former, in which crosswell EM and seismic travel-time tomography are used to estimate reservoir parameters using time-lapse changes in shear velocity, electrical conductivity, and acoustic velocity to sequentially strip off the effects of pressure and water saturation before estimating oil and CO2 saturations. Direct reservoir parameter estimation by joint inversion was demonstrated by Hoversten et al. (2004) and Chen et al. (2004), where marine CSEM and AVA data were used in a formal joint inverse to estimate reservoir Sw, So, Sg, and porosity (φ). The formal joint inversion is currently being extended to replace the 1D CSEM solution with full 3D. While the development and testing of more computationally demanding approaches is underway there is interest in an approach that can be deployed quickly. One method for combining seismic and CSEM data is a relatively straightforward extension of what is currently done using seismic data alone (Bachrach and Dutta, 2004). The use of Bayesian inversion, which couples a rock-physics model with estimates of geophysical parameters, can be extended to include electrical conductivity. In this paper, we demonstrate the use of AVA inversion to estimate acoustic- (Vp), shear-velocity (Vs), and density (ρ) coupled with 3D CSEM (Newman and Boggs, 2004) inversion to estimate electrical conductivity (σ) in a Bayesian inverse for reservoir fluid saturations and φ. This approach is compared to the formal joint inversion described by Hoversten et al. (2004, 2005). The two-stage process has the advantage that it can be done using existing industry software, with only the addition of the electromagnetic inversions to estimate electrical conductivity. The estimated water saturation and porosity compare well to both log data and those derived from a formal joint inversion of marine AVA and electromagnetic data. However, the two-stage estimates of oil and gas saturation do not compare favorably to those obtained using a formal joint inversion of both data sets simultaneously.
DE: 0520 Data analysis: algorithms and implementation
DE: 0644 Numerical methods
DE: 0925 Magnetic and electrical methods (5109)
DE: 0935 Seismic methods (3025, 7294)
DE: 3006 Marine electromagnetics
SC: Geomagnetism and Paleomagnetism [GP]
MN: Fall Meeting 2005