Neural Network Microseismic Data Classification for Integrity Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current microseismic monitoring systems generate vast amounts of data that require extensive manual review to identify significant incidents affecting oil and gas production operations, leading to inefficiencies in analyzing and classifying microseismic data for timely corrective actions.
Innovation Solution
A computer-implemented method using neural network analysis to detect and classify microseismic waves by processing data panels, determining noise events, calculating trigger values, and classifying events into categories based on event attributes such as magnitude, proximity, and spectral density, thereby automating the identification of potential casing failures and other integrity issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review methods are used to analyze microseismic data, then operators can identify significant incidents, but the analysis time and resource requirements increase significantly
Solution Approach 1:
The patent replaces manual mechanical review processes with automated neural network-based computer vision algorithms. The system processes microseismic data panels through trained neural networks that automatically detect and classify events, substituting human operators with computational models that achieve comparable or superior identification accuracy while dramatically reducing analysis time
Solution Approach 2:
The patent introduces an intermediary processing layer consisting of event detection algorithms and classification systems between the raw microseismic data and final incident identification. This intermediary layer pre-processes data panels, extracts relevant features, and filters significant events before presenting them for final analysis, reducing the overall processing burden and time
2Reliability
If extensive manual review is performed on all microseismic data, then comprehensive incident detection is achieved, but processing efficiency decreases
Solution Approach 1:
The patent segments the large volume of microseismic data into smaller, manageable data panels that can be processed independently and in parallel. Each data panel is analyzed by the neural network system, allowing for comprehensive coverage of all data while enabling efficient parallel processing that maintains both detection completeness and processing speed
Solution Approach 2:
The patent applies partial action by using automated neural network filtering to identify and prioritize only the most significant events for detailed review. Rather than manually reviewing all data equally, the system performs excessive automated screening to ensure no significant events are missed, then focuses human expertise only on the subset of events that require detailed analysis
3Productivity
If automated neural network analysis is implemented, then data processing speed increases, but system complexity increases
Solution Approach 1:
The patent implements a universal neural network platform that handles multiple functions including event detection, classification, and characterization within a single integrated system. This multi-functional approach increases processing speed by eliminating the need for separate manual analysis steps while managing complexity through consolidation rather than multiplication of components
Data Source
AI summary
Methods are disclosed for monitoring operation integrity during hydrocarbon production or fluid injection operations. According to the methods, received microseismic data is processed to obtain a plurality of data panels corresponding to microseismic data measured over a predetermined time interval. For each data panel, trigger values are calculated for data traces corresponding to sensor receivers of the microseismic monitoring system. At least one data panel is selected as a triggered data panel that satisfies predetermined triggering criteria. A value is calculated for each of at least two event attributes of a plurality of event attributes of the event. An event is classified into at least one event category of a plurality of event categories based on the event score. Related non-transitory computer usable mediums are also disclosed.