Geological Model Formation Using Seismic Uncertainty Propagation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for integrating seismic and well marker data in petroleum reservoirs fail to effectively capture three-dimensional uncertainty in structural models, leading to biased decisions in drilling and drainage strategies due to the inability to unify diverse data at a common scale and manage positional uncertainties accurately.
Innovation Solution
A method involving seismic data processing using image ray tracing to determine three-dimensional positions and uncertainties, combined with well marker data integration to form a unified geological model, which includes seismic traveltime and velocity uncertainties, and adjusts the model using non-seismically obtained data to preserve interface and fault intersections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional vertical stretch method is used to convert seismic traveltimes to depth, then the processing is simple and fast, but the three-dimensional structural uncertainty cannot be captured and bias is introduced for dipping layers
Solution Approach 1:
The patent introduces an intermediary statistical framework that acts as a mediator between the simple vertical stretch method and the more accurate ray-tracing method. This framework uses covariance matrices to represent and propagate uncertainties through the depth conversion process, allowing the simple method to be used while quantifying and managing the introduced biases and uncertainties in a rigorous statistical manner.
Solution Approach 2:
The patent changes the parameter representation from deterministic single values to probabilistic distributions characterized by mean values and covariance matrices. This allows the model to capture three-dimensional uncertainties in east, north, and depth coordinates simultaneously, transforming the precision issue from a qualitative problem to a quantifiable statistical parameter.
2Device complexity
If well marker data are treated as fixed known measurements, then the integration process is simple, but the positional uncertainties of well markers are not accounted for leading to biased structural models
Solution Approach 1:
The patent merges the uncertainty representations of multiple data sources (seismic traveltimes, velocity models, and well marker positions) into a unified statistical framework. All inputs are treated as probabilistic measurements with associated uncertainties, and their combined effect is propagated through the depth conversion to produce a final model with quantified total uncertainty, thereby improving reliability while maintaining manageable complexity.
3Measurement precision
If seismic data are processed with improved signal to noise ratio, then the data quality is improved, but the lateral continuity information is insufficient when wells are sparse
Solution Approach 1:
The patent addresses the loss of lateral continuity information by moving from traditional two-dimensional seismic sections to a full three-dimensional statistical model. The covariance matrices capture correlations in all three spatial dimensions (east, north, depth), allowing lateral continuity to be represented and propagated statistically even when well spacing is large, thereby recovering information that would be lost in conventional processing.
Data Source
AI summary
A method of forming a geological model of a region of the earth includes obtaining seismic data relating to the region, the seismic data including seismic traveltime uncertainty. A seismic velocity model of the region may also be provided and includes velocity uncertainty. Three dimensional positions of a plurality of points of the region can then be determined. The three dimensional positional uncertainties of at least some of the points can be calculated from the traveltime uncertainty and the velocity uncertainty. This can be combined with the positions to form a geological model.