Stratigraphic Model Update via Facies Classification and Dissimilarity Minimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing stratigraphic simulation methods struggle to accurately update models due to uncertainties in input parameters and the challenge of comparing discrete data like facies, leading to models that may not accurately represent the real stratigraphy of a sedimentary basin.
Innovation Solution
A computer-implemented method that reorganizes data into classes based on value ranges, allowing for the extraction of discriminant features and treating facies as categorical data. This method updates stratigraphic models by minimizing an objective function formed from local dissimilarities between measured and simulated data, while accounting for uncertainties and focusing on the geometry of basin layers and sediment properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If stratigraphic simulation uses interpreted input parameters from measurements, then the model can be constructed, but the accuracy of the model decreases due to parameter uncertainties
Solution Approach 1:
The patent implements an iterative feedback mechanism where the stratigraphic model is repeatedly simulated, compared with measured data, and updated by adjusting input parameters. The objective function quantifies the mismatch between simulated and measured data, guiding parameter optimization. This closed-loop feedback process progressively improves model accuracy despite initial parameter uncertainties.
Solution Approach 2:
The patent systematically varies stratigraphic simulation input parameters (sediment supply rates, transport coefficients, basin deformation rates) to find optimal values that minimize the objective function. By changing parameters iteratively based on model-data comparison, the method transforms uncertain input parameters into optimized values that better represent reality.
2Ease of operation
If traditional cell-to-cell comparison is used to update the model, then the process is simple, but it fails to capture spatial variations and produces inaccurate models
Solution Approach 1:
The patent segments the basin into discrete columns and further divides each column into stratigraphic layers. This segmentation allows independent comparison of simulated and measured data at each column-layer intersection, capturing spatial variations in sediment thickness and properties across the basin while maintaining a manageable computational structure.
Solution Approach 2:
The patent extends the comparison from simple cell-to-cell (0D) to incorporate spatial dimensions by comparing data across multiple columns and depth layers. The objective function integrates mismatches across the 3D stratigraphic space, transforming the update process from a single-point comparison to a multi-dimensional spatial optimization that captures geological variability.
3Ease of operation
If discrete data like facies are treated continuously, then the model can be processed easily, but the categorical nature of facies is lost leading to poor model representation
Solution Approach 1:
The patent applies different treatment strategies to different data types based on their local characteristics. Continuous properties (sediment thickness, grain size) are processed using numerical optimization, while discrete facies data are handled through categorical comparison methods. This localized quality approach ensures each data type is processed according to its inherent nature, improving overall model reliability.
Solution Approach 2:
The patent introduces an intermediary classification system that maps facies to discrete categories before comparison. By using facies classification schemes as intermediaries, the method preserves the categorical nature of geological units while enabling systematic comparison between simulated and measured data through standardized facies codes or labels.
Data Source
AI summary
The present invention is a method for exploiting a sedimentary basin using an updated stratigraphic model based on updating a stratigraphic model according to measurements performed in a basin. The method comprises steps of i) determining at least a first spatial distribution of the values of an attribute representative of the basin stratigraphy, and applying a classification method for converting this first spatial distribution to a first classified image; ii) performing a stratigraphic simulation with first parameter values of the simulation, deducing therefrom a second spatial distribution of the values of the attribute and applying the same classification method identically to determine a second classified image; iii) determining a distribution of local dissimilarities between the first and second classified images, and modifying the values of at least one of the stratigraphic simulation parameters to minimize an objective function formed from at least the local dissimilarity distribution; and iv) updating the stratigraphic model by performing at least one stratigraphic simulation with the modified simulation parameter values.