Saltwater Disposal Pump Alarm Validation Using Regression Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current monitoring systems for saltwater disposal pump systems often fail to accurately detect abnormal operating conditions, leading to false alarms and costly downtime due to unreliable data, and there is a need for a system that can reliably predict future machine performance.
Innovation Solution
A saltwater disposal pump system equipped with sensors to gather data on suction strainer differential pressure, suction flow, discharge pressure, suction pressure, vibration, and temperature, using statistical regression analysis to validate alarms and automatically switch to functioning equipment when abnormalities are detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional sensor-based monitoring is used to detect abnormal operating conditions, then the system can identify potential issues, but the data reliability is poor leading to false alarms and unnecessary downtime
Solution Approach 1:
The system uses feedback by continuously comparing predicted sensor values (from the digital twin) with actual sensor readings. When deviations occur, the system adjusts and refines the digital twin model, creating a closed-loop validation mechanism that improves alarm reliability over time by filtering out false alarms while maintaining sensitivity to real issues
Solution Approach 2:
The patent creates a digital copy (digital twin) of the saltwater disposal pump system that mirrors the physical system's behavior. This digital replica is used to predict sensor readings and validate alarm conditions, allowing the system to distinguish between actual equipment issues and sensor false readings without requiring physical intervention or downtime
2Measurement precision
If multiple sensors are deployed to monitor pump operations, then more data is available for detection, but the complexity of validating and correlating multiple data streams increases
Solution Approach 1:
The digital twin acts as an intermediary between multiple sensors and the alarm validation system. Instead of directly correlating and validating data from multiple sensors, the system compares actual sensor readings against predictions from the digital twin model, simplifying the validation process while maintaining high measurement precision through the intermediary's predictive capabilities
Solution Approach 2:
The system changes parameters by using the digital twin to predict what sensor values should be under current operating conditions. This transforms the validation approach from checking absolute threshold violations to evaluating deviations between predicted and actual parameter values, improving detection accuracy while managing complexity through parameter transformation
3Reliability
If alarm thresholds are set to be sensitive to detect all abnormalities, then more issues are detected, but false alarms increase leading to alarm fatigue and reduced operator response
Solution Approach 1:
The system applies partial action by using the digital twin to predict only the critical portion of sensor behavior that indicates real problems. Rather than setting thresholds for all possible sensor variations, the system focuses validation on deviations from predicted values, effectively filtering out normal operational variations (noise) while maintaining sensitivity to actual abnormalities
Data Source
AI summary
Embodiments of the present disclosure are directed to a system for monitoring operating conditions of a saltwater disposal pump system, which includes a processing device that causes a controller to receive an alarm from the at least one machine that exceeds a predetermined alarm level indicative of an abnormal machine operation, receive a plurality of data from a plurality of sensors that are indicative of a current operating conditions of a saltwater disposal pump assembly, derive an expected value utilizing a statistical regression analysis, determine whether the expected value and a history of the plurality of data correlate at a point where the plurality of data exceeds the predetermined alarm level, determine whether a deviation from a predefined operating profile exceeds a predetermined machine data validation threshold value, validate the alarm, output an alert of the deviation and inhibit further operation of the saltwater disposal pump assembly.