Abstract
This study aims at enhancing reservoir characterization through simultaneous history matching of time-lapse seismic and electromagnetic (EM) data. We propose an efficient ensemble-based history-matching approach by exploiting the complementary nature of these data types. The developed workflow consists of two main steps. Firstly, water saturation and porosity as the rock cross-properties relating to both seismic velocities and formation conductivity are estimated by joint inversion of seismic and EM data using an iterative ensemble smoother with an adaptive non-distance-based Kalman gain localization. Prescribed rock-physics models are used to quantify the corresponding relationship of rock properties to seismic velocities and formation conductivity, based on which synthetic EM and full-waveform seismic data are simulated. Secondly, instead of integrating the inverted saturation field directly, front positions are identified from the inverted saturation field and used as the observed data for history matching. Model parameters are conditioned to the observed fronts using the iterative ensemble smoother with a feature-oriented distance parameterization. The approach is verified with synthetic data sets. Results show that joint inversion of seismic and EM data leads to better estimates of rock properties when compared with those from the separate inversion. Moreover, by integrating the extracted fronts from the improved estimation of the saturation field, which synthesizes the information from both types of data, the proposed approach significantly reduces the redundancy and number of data while still capturing the essential information. The novelty of the proposed approach consists in combining the feature-based history matching with ensemble-based geophysical inversion to achieve an efficient joint integration of time-lapse seismic and EM data.
Original language | English (US) |
---|---|
Title of host publication | International Petroleum Technology Conference |
Publisher | International Petroleum Technology Conference |
ISBN (Print) | 9781613996751 |
DOIs | |
State | Published - Jan 11 2020 |
Bibliographical note
KAUST Repository Item: Exported on 2020-10-01Acknowledgements: Support for the authors is provided by the research project