Anomaly Detection Using Combined Surface and Downhole Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mathematical models for detecting operational anomalies in wellbore operations rely solely on surface data, which may not accurately identify issues due to differences in operational parameters between surface and downhole equipment, leading to potential equipment failures and production losses.
Innovation Solution
The use of machine-learning models that combine surface and downhole measurement data to predict equipment malfunctions, trained on historical data and updated in real-time, to identify anomalies in operational parameters such as pressure, temperature, force, torque, and vibrations, enabling early detection of equipment irregularities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional mathematical models rely solely on surface data for anomaly detection, then the system complexity is reduced and ease of operation is improved, but the measurement precision and reliability of anomaly detection deteriorate due to differences in operational parameters between surface and downhole equipment
Solution Approach 1:
The patent combines surface data and downhole data into a unified machine learning model for anomaly detection. The system integrates multiple data sources (surface operational parameters and downhole operational parameters) to improve detection accuracy while managing complexity through automated processing pipelines.
Solution Approach 2:
The patent replaces conventional mathematical models with machine learning models that can automatically learn and adapt to the complex relationships between surface and downhole operational parameters, substituting traditional physics-based approaches with data-driven algorithms that handle multi-source data more effectively.
2Reliability
If machine-learning models combine surface and downhole measurement data, then the reliability of equipment anomaly detection is improved, but the device complexity and difficulty of detecting and measuring increases
Solution Approach 1:
The machine learning model performs self-training and automatic adaptation to the specific operational conditions. The system learns from historical data and continuously improves its detection capabilities without requiring manual recalibration or complex configuration, reducing the operational burden despite the sophisticated underlying technology.
Solution Approach 2:
The system implements feedback loops where detection results and operational outcomes are fed back into the machine learning model to continuously improve its accuracy. This automated feedback mechanism enhances reliability while reducing the need for manual intervention in the complex detection process.
3Productivity
If real-time machine-learning models are used for anomaly detection, then the productivity and response time are improved, but the use of energy and computational resources increases
Solution Approach 1:
The system applies machine learning models selectively to detect anomalies rather than continuously processing all data at maximum capacity. The model processes data in real-time only when needed for anomaly detection, reducing overall computational energy consumption while maintaining high productivity during critical detection events.
Data Source
AI summary
The disclosed technology provides solutions for performing equipment anomaly detection. In particular, a process of the disclosed technology includes steps for receiving surface data from one or more surface sensors, receiving downhole data from one or more downhole sensors, and analyzing a combination of the surface data and the downhole data to determine if an operational anomaly is detected with respect to the surface equipment devices or the downhole equipment devices. Systems and computer-readable media are also provided.