Robust inversion of controlled source electromagnetic (CSEM) data

Marine controlled source electromagnetic (CSEM) surveys are widely used in exploration as a supplement to seismic data, since CSEM data discriminate well between brine and hydrocarbons. In recent years, the potential of using CSEM data for monitoring hydrocarbon reservoirs during production, as a supplement to time-lapse seismic, has also been investigated.
Most often, a pixel-based representation of the unknown electric conductivity function is used for inversion purposes. A novel, model-based (i.e., as opposed to pixel-based) inversion methodology has recently been developed. In its present form, however, this inversion methodology assumes a known overburden geoelectric model while focusing on identifying fluid movements within a reservoir with known location, and is therefore more suitable for monitoring than for exploration purposes. This project will further develop the novel, model-based methodology for inversion of CSEM data to handle an unknown overburden geoelectric model in addition to a (potentially existing) unknown reservoir.

Publications:

S. Tveit, S.A. Bakr, M. Lien, T. Mannseth. Identification of subsurface structures using electromagnetic data and shape priors. Journal of Computational Physics.

S. Tveit, S.A. Bakr, M. Lien, T. Mannseth. Ensemble-based, Bayesian Inversion of CSEM
Data Using Structural Prior Information, in proceedings of the 76th EAGE Conference and Exhibition 2014, Amsterdam, the Netherlands.

S. Tveit, S.A. Bakr, M. Lien, T. Mannseth. Ensemble-based Bayesian inversion of CSEM data for subsurface structure identification, submitted to Geophysical Journal International.