Machine Learning Wellsite Control for Alarm Validation and Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wellsite control systems often raise incorrect alarms, leading to operator distrust and response delays due to improper notification routing and validation, which can result in ignored alerts and inefficient wellsite management.
Innovation Solution
A system utilizing machine learning models to process sensor information from various wellsite locations, enabling accurate notification validation, routing, and automated responses, leveraging remote computational resources for improved decision-making and reducing human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional alarm systems are used at wellsites, then operators receive notifications, but incorrect alarms are raised leading to operator distrust and response delays
Solution Approach 1:
The system implements feedback loops where alarm notifications are sent to remote users who can validate or dismiss alarms. This feedback mechanism filters out incorrect alarms before they reach operators, improving alarm accuracy while maintaining rapid response times through automated validation processes
Solution Approach 2:
Remote users act as intermediaries between the alarm system and operators. They receive and validate alarm notifications, serving as a filtering layer that prevents incorrect alarms from triggering operator responses, thereby improving reliability without adding significant time delays
2Reliability
If alarms are routed to multiple operators, then coverage is improved, but response delays occur due to notification routing issues
Solution Approach 1:
The notification system is segmented into multiple stages: initial alarm generation, remote user validation, and operator notification. This segmentation allows the system to route to multiple potential recipients simultaneously while maintaining clear progression, ensuring reliable delivery without unnecessary delays
Solution Approach 2:
Remote users perform preliminary validation of alarms before they are routed to operators. This preliminary action ensures that only valid alarms proceed to the operator notification stage, improving delivery reliability while reducing the total number of operators who need to be notified
3Reliability
If manual validation of alarms is required, then alarm accuracy improves, but response time decreases due to human intervention
Solution Approach 1:
Remote users serve as intermediaries who perform validation with minimal time investment. The system design allows these intermediaries to quickly assess and validate or dismiss alarms, maintaining high validation accuracy while minimizing the time added to the response process
Solution Approach 2:
The system replaces traditional manual operator validation with a distributed network of remote users who validate alarms remotely. This substitution maintains validation accuracy while enabling parallel processing of multiple alarms, thereby improving overall response speed
Data Source
AI summary
A system and method include receiving a drill plan and wellsite information associated with a well site that includes at least one of: surface information and downhole information. The system and method also include inputting the drill plan and the wellsite information into a machine learning model and determining at least one at least control command based on at least one of: the wellsite information and the drill plan. The system and method additionally include transmitting the at least one control command to a notification destination to change at least one operating parameter of at least one equipment at the wellsite.