Stochastic Seismic Velocity Modeling for Depth Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seismic modeling techniques face uncertainty in subsurface imaging due to incorrect assumptions about seismic velocities and anisotropy, limiting the generation of accurate 3D models and depth uncertainty analysis, which hinders hydrocarbon deposit identification and reserve estimation.
Innovation Solution
A stochastic modeling method that generates 3D model realizations of seismic velocity and anisotropic parameters within predefined bounds using random numbers, testing for detectability to produce images with flat migrated gathers, and converting these models into the depth domain for improved uncertainty analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stochastic modeling generates a large number of 3D model realizations, then the accuracy of depth uncertainty analysis is improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the 3D modeling process into multiple independent realizations that can be generated and tested separately. Each realization represents a possible subsurface model within the uncertainty bounds, allowing parallel processing and systematic evaluation of multiple scenarios without requiring a single monolithic computational approach.
Solution Approach 2:
The patent introduces a probabilistic dimension to the traditional deterministic 3D modeling approach. By generating multiple realizations within uncertainty bounds and evaluating their detectability, the method adds a fourth dimension of analysis that quantifies uncertainty while maintaining computational efficiency through systematic sampling and filtering.
2Reliability
If the number of 3D model realizations is increased to improve uncertainty analysis, then the reliability of hydrocarbon deposit identification is improved, but the time required for processing increases
Solution Approach 1:
The patent performs preliminary actions by generating multiple 3D model realizations within uncertainty bounds before final interpretation. These realizations are pre-tested for detectability and converted to the depth domain in advance, allowing the most reliable models to be identified and used for hydrocarbon deposit identification without requiring time-consuming iterative processing during the final analysis stage.
Solution Approach 2:
The patent implements feedback through the detectability testing process, where each generated realization is evaluated and filtered based on whether it produces detectable signals with flat migrated gathers. This feedback mechanism allows inefficient models to be discarded early, reducing the total processing time required while maintaining high reliability in the final hydrocarbon identification.
3Ease of operation
If assumptions about seismic velocities and anisotropy are simplified, then the ease of modeling is improved, but the measurement precision of subsurface properties deteriorates
Solution Approach 1:
The patent changes the approach from using fixed, simplified velocity and anisotropy assumptions to using probabilistic parameter ranges. By defining uncertainty bounds for these parameters and generating multiple realizations within those bounds, the method maintains ease of operation through systematic sampling while significantly improving measurement precision by capturing the full range of possible subsurface properties.
Data Source
AI summary
A method is described for stochastic modeling of seismic velocity and anisotropic parameters, including receiving 3D bounds of normal moveout velocity (Vnmo) and anisotropic parameter η; modeling 3D bounds for vertical velocity V and anisotropic parameter δ based on the 3D bounds of Vnmo and η; generating 3D model realizations of V, η, and δ within the 3D bounds; and testing detectability of each of the 3D model realizations to create a detectable subset of model realizations wherein the detectability identifies which 3D model realizations will produce images with flat migrated gathers. The method may be executed by a computer system.