Modeling and inversion of seismic, electromagnetic and production data
Modification of techniques from theoretical physics for adaptation in geophysical modelling and inversion.
The methods for inversion and imaging of seismic and electromagnetic data are typically developed by using integral equation (Green's function) methods similar to those that have been used to develop rock physics models and similar to those that have been developed in acoustics, optics and quantum scattering theory. The integral equations that describe the scattering of seismic or electromagnetic waves by reservoir rock units and other geological structures (e.g., faults) are usually linearized, so that the seismic data can be relatively easily inverted with respect to the elastic parameters of the target structures in the underground. Attempts are presently made to use quasi-linearized approximations, inspired by the T-matrix approach to rock physics.
The integral equations are based on a decomposition of the inhomogeneous medium into a relatively simple background medium and a perturbation. For calculations of seismic wavefields in the background medium, asymptotic ray theory is normally used. The development and use of asymptotic methods have recently been enhanced by our internal studiess and collaboration with international experts. The fact that we are using the same kind of methods to model the propagation of low-frequency electromagnetic pulses in the diffusion regime as well as the behaviour of seismic waves in laterally inhomogeneous media is both interesting and promising when it comes to our analytical and applied work on joint inversion.
We are currently developing methods for simultaneous processing (joint inversion) of seismic and electromagnetic data, particularly in connection with reservoir monitoring and dynamic reservoir characterization related to petroleum production as well as CO2 storage. The advantage of performing a quantitative integration of seismic and electromagnetic data is that these data types are complementary, as the electromagnetic data are very sensitive to fluid saturations (oil vs. water, etc.), whereas seismic data have better resolution. By performing a simultaneous analysis seismic and electromagnetic data, we expect to get the best information from these two data types. A part of our research activity in the development of seismic and electromagnetic methods is within the framework of current projects at CIPR, where we participate in projects involving production data (reservoir simulators).
The development of rock physics theories is important for the integration of different data types, usually associated with different length scales. By using rock physics models, we can relate seismic and electromagnetic data to a common set of parameters (e.g., porosity and saturation), and it is actually this rock physics link that makes it possible to perform a joint inversion of different data types. Also, the rock physics models provide a link to geology in this context.