Inversion-Based Reflector Dip Estimation in Seismic Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing seismic data processing methods for hydrocarbon exploration face challenges in accurately estimating reflector dip angles due to noise and computational inefficiencies, particularly in post-stack images, leading to unreliable and time-consuming processes.
Innovation Solution
An automated method using conjugate-gradient least-squares inversion to iteratively flatten reflector dips in post-stack image traces, providing a robust and efficient estimation of reflector dip angles, which are then used to construct a more accurate subsurface velocity model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional slant-stack-type reflector dip estimation methods are used on post-stack images, then the process is automated, but the results suffer from conflicting dips and wild swings due to noise
Solution Approach 1:
The patent segments the post-stack image into multiple windows, each containing a subset of traces. By processing each window independently and combining results, the method reduces the impact of noise on the overall dip estimation while maintaining automation. This segmentation allows localized dip calculations that are less susceptible to global noise patterns.
Solution Approach 2:
The patent performs preliminary flattening of reflectors using an initial velocity model before conducting dip estimation. This preliminary action stabilizes the input data by removing coherent migration noise and aligning reflectors, thereby improving the reliability of subsequent automated dip estimation without sacrificing automation extent.
2Reliability
If manual picking of interfaces is used to estimate dip angles, then dip estimation stability is improved, but the process becomes time-consuming and inaccurate for complex structures
Solution Approach 1:
The patent implements an automated system that performs dip estimation without requiring manual intervention. The algorithm automatically picks interfaces, calculates dip angles, and handles complex interlayer structures through iterative optimization, thereby eliminating the time-consuming nature of manual picking while maintaining stability through robust computational methods.
Solution Approach 2:
The patent replaces the manual mechanical process of interface picking with an automated computational algorithm. The system uses computer-implemented methods including trace extraction, flattening calculations, and dip angle computation to substitute human operators, significantly reducing processing time while maintaining or improving accuracy for complex structures.
3Productivity
If cross-correlation algorithms are used for automatic picking, then computational efficiency is improved, but the solutions become sensitive to window size and signal-to-noise ratio
Solution Approach 1:
The patent dynamically adjusts the window size and processing parameters based on the local characteristics of the seismic data. By adapting the analysis window to the signal-to-noise ratio and reflector complexity in different regions, the method maintains computational efficiency while improving measurement precision. The algorithm iteratively refines depth shift estimates, allowing it to overcome the static limitations of fixed-window cross-correlation methods.
4Device complexity
If post-stack images are used for dip estimation, then the process is simplified compared to pre-stack gathers, but the images contain strong coherent migration noise and low signal-to-noise ratio
Solution Approach 1:
The patent extracts a specific window of traces from the post-stack image centered about a reference trace. By isolating a localized region with defined boundaries and characteristics, the method separates the signal of interest from the broader noisy context, enabling more accurate dip estimation despite the presence of coherent migration noise in the full image.
Solution Approach 2:
The patent introduces an intermediate flattening step that uses an initial velocity model to pre-process the post-stack image traces. This intermediary processing stage reduces the impact of coherent migration noise before the actual dip estimation occurs, acting as a mediator that preserves the simplicity of using post-stack images while mitigating their harmful noise characteristics.
Data Source
AI summary
Method for estimating reflector dips in a window of post stack image traces (51) of seismic data for use in velocity tomography (57). The method iteratively (56) flattens (55) the image traces against a specified reference trace through the application of conjugate-gradient least-squares inversion (53). Different from other dip estimation methods which emphasize on strong-amplitude reflectors, the inventive method automatically inverts for the reflector dip for every grid point in the image window.