Subsurface Geological Modeling Across Faults With Dynamic Kernel Masks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing geological modeling techniques are labor-intensive, computationally expensive, and inefficient in handling subsurface discontinuities like faults, leading to inaccurate and resource-intensive model generation.
Innovation Solution
A subsurface modeling system using a horizon-fault machine learning model with a dynamic kernel mask and U-Net architecture processes sparse geological data to isolate continuous zones, employing interpolation and targeted convolution to generate accurate geological models with reduced computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional geological modeling techniques are used to handle subsurface discontinuities like faults, then model accuracy is maintained, but computational cost and time consumption increase significantly
Solution Approach 1:
The patent segments the geological modeling process by introducing a dynamic kernel mask that divides the computational domain into continuous zones separated by faults. This segmentation allows the model to process each zone independently, reducing the overall computational complexity while maintaining accuracy at fault boundaries. The U-Net architecture further segments the processing into encoding and decoding paths with skip connections, enabling efficient handling of discontinuities.
Solution Approach 2:
The patent applies local quality by using a dynamic kernel mask that adapts its properties based on the local geological structure. The kernel mask modifies the convolution operations to have different effective receptive fields in continuous zones versus fault zones, allowing the model to focus computational resources where they are most needed while maintaining high accuracy at discontinuities.
2Reliability
If traditional geological modeling techniques are used, then comprehensive geological features are captured, but system complexity and resource requirements increase
Solution Approach 1:
The patent introduces a dynamic kernel mask as an intermediary element that mediates between the input seismic data and the geological model output. This mask acts as a adaptive filter that selectively enhances or suppresses features based on the local geological context, particularly at fault boundaries, thereby improving reliability without requiring a more complex overall system architecture.
Solution Approach 2:
The patent employs dynamics by making the kernel mask adaptive rather than static. The mask dynamically adjusts its properties during the forward propagation process based on the input data characteristics, allowing the system to automatically adapt to different geological scenarios without increasing structural complexity.
3Quantity of substance
If sparse geological data is processed using conventional methods, then complete geological models are generated, but computational resources are excessively consumed
Solution Approach 1:
The patent applies preliminary action by pre-processing the sparse geological data through a U-Net encoder that extracts key features and generates a compressed representation before the main decoding process. The dynamic kernel mask is also prepared in advance to guide the subsequent reconstruction process, enabling efficient generation of complete geological models from sparse input data with reduced computational resource consumption.
Data Source
AI summary
A method of modeling subsurface geology includes receiving interpreted data including fault data indicating at least one fault within the interpreted data, and geological feature data including a plurality of horizon picks indicating a subsurface geological feature. The method includes, based on the fault data, creating a dynamic kernel mask. The method includes generating a geological model using a horizon-fault machine learning (ML) model that is generated to process input geological feature data to predict continuous geological features across subsurface discontinuities based on applying the dynamic kernel mask to isolate continuous zones in the input geological feature data. The method also includes providing the geological model for simulating one or more properties of a geological feature indicated in the geological feature data.