4D Seismic Inversion via Optimal Regularisation Weight-Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 4D seismic inversion techniques for reservoir evolution analysis are prone to instability due to noise and require extensive interpretation and computational time, limiting their practicality for real-time reservoir management.
Innovation Solution
The process involves deriving an optimal regularisation weight-map from multiple seismic surveys to improve the accuracy of time-lapsed seismic images, using a cost function that balances data fitting and constraint imposition, thereby reducing the need for prior knowledge and computational time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If cross-correlation is used with a large window size, then the coverage of 4D effects is improved, but the correlation accuracy deteriorates
Solution Approach 1:
The patent divides the seismic trace into multiple overlapping windows of appropriate size, allowing each window to maintain good correlation accuracy while the collection of windows collectively covers all 4D effects. This segmentation approach resolves the contradiction by breaking down the large window into smaller manageable segments that can be processed independently and then combined.
Solution Approach 2:
The patent introduces the time dimension by performing correlation at multiple time lags and combining results from multiple windows. This transforms the single large-window correlation problem into a multi-dimensional processing approach where accuracy is maintained through temporal sampling while comprehensive coverage is achieved through aggregation across time and space.
2Measurement precision
If cross-correlation is used with a small window size, then the correlation accuracy is improved, but the coverage of 4D effects deteriorates
Solution Approach 1:
The patent merges results from multiple small windows that are positioned at different locations and time lags. By combining the correlation results from these individual small windows, the method achieves both the accuracy benefit of small windows and the comprehensive coverage that would require a large window, effectively resolving the contradiction through aggregation.
Solution Approach 2:
The patent performs preliminary correlation operations on multiple small windows with different time lags before final aggregation. This preliminary action allows each small window to be optimized for accuracy while the collective set of preliminary operations ensures comprehensive coverage of all 4D effects through strategic positioning and temporal sampling.
3Measurement precision
If inversion is performed to derive velocity changes, then the characterization of reservoir evolution is improved, but the computational time and complexity increase
Solution Approach 1:
The patent extracts velocity change information directly from the correlation results and time-lag measurements without performing full inversion. By taking out only the essential velocity change parameters needed for reservoir characterization and deriving them through simplified calculations rather than complete inversion, the method maintains characterization accuracy while dramatically reducing computational time and complexity.
Solution Approach 2:
The patent uses simplified correlation-based velocity estimation as a computationally inexpensive alternative to full inversion. This disposable, lower-cost approach provides sufficient velocity change information for reservoir monitoring purposes, eliminating the need for computationally intensive inversion while maintaining practical utility and reducing processing time from weeks to hours.
4Reliability
If full inversion with regularisation is performed, then the stability of the solution is improved, but the computational time increases
Solution Approach 1:
The patent extracts stable velocity change information directly from the correlation framework without performing full regularized inversion. By extracting the essential velocity parameters from the correlation measurements and applying minimal processing, the method achieves solution stability through the inherent robustness of correlation while avoiding the computational burden of full inversion with regularisation.
Data Source
AI summary
Disclosed is a process for characterising the evolution of a reservoir over a time lapse comprising the steps of: providing a base and a plurality of monitor surveys of the reservoir, each having a set of seismic traces at different times; deriving an optimal regularization weight-map from a combination of at least three surveys; and using the optimal regularization weights to invert and obtain an improved time lapse seismic image between pairs of seismic surveys.