Karst Cave Reservoir Focus Identification via Velocity Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional seismic exploration methods for Karst cave reservoirs face inaccuracies in migration velocity modeling, leading to non-convergent 'beads' imaging, which hampers reservoir prediction and drilling success due to high calculation consumption and susceptibility to local extremum convergence issues.
Innovation Solution
An automatic focus identification method and system that performs continuous pre-stack time migration with velocity replacement, utilizing sparse statistical data from seismic traces and attributes to optimize migration velocity and seismic traces, improving prediction accuracy and drilling success rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If least-squares migration imaging method is used to obtain convergent beads imaging, then imaging convergence is improved, but calculation consumption increases significantly
Solution Approach 1:
The patent segments the velocity model into multiple layers and performs velocity spectrum picking for each layer independently. This divides the complex full-waveform inversion problem into smaller, more manageable sub-problems, reducing overall calculation consumption while maintaining imaging convergence quality.
Solution Approach 2:
The patent performs preliminary velocity spectrum picking to obtain an initial velocity model before conducting detailed migration imaging. This preliminary action provides a better starting point for subsequent imaging processes, reducing the computational burden of achieving convergence from scratch.
2Measurement precision
If least-squares migration imaging method is used, then imaging convergence is improved, but the method becomes highly dependent on initial model and prone to local extremum
Solution Approach 1:
The patent employs dynamic velocity adjustment through spectrum picking, where the velocity model is continuously refined and updated based on imaging results. This dynamic approach allows the system to adapt to different geological conditions and avoid getting trapped in local extrema, improving convergence reliability.
Solution Approach 2:
The patent implements a feedback mechanism where migration imaging results are used to update and refine the velocity model iteratively. This feedback loop ensures that the initial model dependencies are reduced over time, as each iteration improves upon the previous model, leading to more stable and reliable convergence.
3Productivity
If conventional velocity spectrum picking with large grid is used, then processing speed is maintained, but migration velocity model accuracy decreases
Solution Approach 1:
The patent segments the velocity model into multiple finer layers, allowing for more precise velocity picking in each layer. This segmentation enables accurate velocity modeling without requiring excessively large computational grids, thus maintaining processing speed while improving accuracy.
Solution Approach 2:
The patent changes the parameter resolution by performing velocity spectrum picking at multiple scales and resolutions. This allows the system to capture both large-scale velocity trends and local variations, achieving high accuracy velocity models while maintaining computational efficiency through adaptive parameter selection.
Data Source
AI summary
Disclosed are an automatic focus identification method and a system for a Karst cave reservoir. According to the method, velocity replacement is performed based on a pre-stack time migration velocity model to obtain a plurality of replacement migration velocity volumes; pre-stack time migration is performed based on the plurality of replacement migration velocity volumes to obtain a plurality of post-stack seismic data volumes; a three dimensional cube window is selected based on a seismic response characteristic of the Karst cave reservoir; sparse statistical data of a seismic trace or sparse statistical data of a seismic trace attribute of each of the post-stack seismic data volumes in the three dimensional cube window are obtained respectively; and an optimized migration velocity and an optimized seismic trace are obtained through optimization determination.