Innovative Wave Equation Migration
Goal: The goal is to develop novel seismic-migration algorithms using wave theory for imaging complex structures that are difficult to image using the industry standard ray-based Kirchhoff migration algorithms.
Objectives: The objective of this research is to significantly enhance the capability to accurately image subsalt structures in regions like the deepwater Gulf of Mexico.
Performers: Los Alamos National Laboratory (LANL) – project management and research products
- Several new and improved wave-equation migration methods have been developed
- Migration examples using 2-D and 3-D synthetic data and field-acquired data have demonstrated that the new methods can produce clear images showing detail structures of complex regions.
- More than 40 participants from 20 oil/gas companies regularly attended the project meetings.
It is very difficult for the widely used ray-tracing-based Kirchhoff migration to accurately image complex structures such as those beneath salt bodies. Wave-equation migration is a promising tool for accurately imaging these complex structures and is becoming more feasible for practical applications because of increasing computer speed. Wave-equation migration based on recursive one-way wave downward continuation can be implemented in the frequency-space and frequency-wavenumber domains (dual-domain). The dual-domain migration methods are efficient and have several advantages over pure frequency-space finite-difference schemes. For instance, they have no or significantly less grid dispersion that introduces artifacts into images, they can handle high frequencies more easily (yielding higher-resolution images), and they have few or no artifacts caused by operator-splitting that introduces artificial anisotropy for three-dimensional imaging. Researchers at LANL have developed a suite of new wave-equation migration methods that are more accurate than the well-known dual-domain migration algorithm termed the split-step Fourier method. One of these new methods, termed globally optimized Fourier finite-difference, is likely the most accurate and efficient dual-domain wave-equation migration method for imaging complex structures having strong lateral velocity variations and steep interfaces. Project researchers will develop innovative wave-equation methods that will maintain or improve the accuracy of the most accurate dual-domain method while increasing the computational efficiency as much as possible. In addition, they will develop new amplitude-preserving wave-equation migration methods for reliable seismic reservoir characterization.
In May of 2002, the NETL entered into a field office work agreement with the LANL to develop innovative wave-equation methods. This project is intended to improve the accuracy of the dual-domain method while increasing the computational efficiency. New true-amplitude wave-equation migration methods will also be developed for reliable seismic reservoir characterization.
Better seismic images of reservoirs need to be obtained if we are to continue to explore and produce from ever more complex geologic structures. These reservoirs are difficult to image properly using conventional methods because they do not adequately account for the wave phenomena that can occur in complex structures. This work will build on previous efforts to increase the accuracy and efficiency of the information needed to properly characterize our natural gas resource.
Current Status and Remaining Tasks: Currently, LANL is conducting research to improve computational efficiency of the wave-equation migration methods while maintaining accuracy. Once the migration algorithms have been refined, LANL will investigate how to extract material physical properties across an interface from prestack wave-equation migration images. This will add important information for reliable reservoir characterization.
Project Start: May 1, 2002
Project End: September 30, 2005
DOE Contribution: $439,000
Performer Contribution: $0
NETL – John D. Rogers (firstname.lastname@example.org or 304-285-4880)
LANL – James N. Albright (email@example.com or 505-667-4318)