Well Site Edge Analytics for Real-Time Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Oil and gas wells often operate unattended in remote areas, leading to costly and time-consuming maintenance when issues arise, resulting in productivity and profitability losses, as well as safety risks for field personnel due to the need for physical inspections and repairs.
Innovation Solution
Implementing machine learning (ML) based edge analytics at well sites to detect unusual events and operating conditions in real-time, enabling automatic alerts and responses to minimize downtime and safety risks, using edge devices equipped with processors and storage for monitoring and control applications that analyze operational parameter data from remote programmable automation controllers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If field personnel physically inspect and repair equipment at remote well sites, then equipment reliability is maintained, but productivity is lost and safety risks increase due to travel time and remote location
Solution Approach 1:
The system enables self-service through automated monitoring and alerting. Sensors continuously monitor equipment parameters, and the system automatically detects anomalies and alerts operators, eliminating the need for frequent manual inspections and allowing equipment to operate autonomously until actual failure occurs.
Solution Approach 2:
Physical inspection by field personnel is replaced with electronic sensing and automated analysis. The system uses sensors, data transmission, and automated diagnostic algorithms to substitute human mechanical inspection with electronic monitoring, reducing travel time and increasing productivity.
2Ease of repair
If field personnel travel to remote well sites for inspections and repairs, then equipment issues are resolved, but safety risks increase and costs increase due to travel requirements
Solution Approach 1:
The system performs preliminary detection and diagnosis before field personnel need to travel. By continuously monitoring equipment parameters and detecting anomalies early, the system identifies issues before they become critical, allowing operators to plan repairs in advance and reduce the urgency and risk of remote field interventions.
Solution Approach 2:
Dangerous physical inspections in remote locations are replaced with remote electronic monitoring. The system uses sensors and automated analysis to detect equipment issues without requiring personnel to physically visit hazardous remote well sites, thereby eliminating safety risks associated with field travel.
3Ease of operation
If manual monitoring and inspection methods are used, then equipment operation is maintained, but downtime increases and cost losses increase when issues arise
Solution Approach 1:
The system implements continuous feedback through real-time monitoring of equipment parameters. Sensors continuously collect data, the system analyzes trends and detects deviations from normal operation, and automatically alerts operators immediately when anomalies are detected, enabling rapid response and minimizing downtime compared to periodic manual inspections.
Solution Approach 2:
Manual monitoring and inspection processes are replaced with automated electronic systems. The system uses sensors, automated data collection, and computerized analysis to continuously monitor equipment without human intervention, enabling immediate detection of issues and reducing the time lag inherent in manual inspection methods.
Data Source
AI summary
Systems and methods for real-time monitoring and control of well site operations employ well site edge analytics to detect abnormal operations. The systems and methods receive well site data from a remote programmable automation (PAC) controller at the well site, the well site data representing one or more operational parameters related to the well site operations. A probability is derived for a given slope for each one of the one or more operational parameters as correlated to a different one of the one or more operational parameters to produce correlated probabilities for the one or more operational parameters. A resultant probability is derived from the correlated probabilities for the one or more operational parameters and it is determined whether the resultant probability meets a preselected threshold probability value. A responsive action is initiated if the resultant probability fails to meet the preselected threshold probability value.