Reciprocating Compressor Failure Prediction Using Slope Signatures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems for reciprocating compressors are inadequate for providing predictive indications of pending failures, relying on manual inspections and basic sensor data interpretation, which are prone to human error and limited in identifying impending component failures.
Innovation Solution
A control system with sensors that measure operational parameters, calculate the rate of change of these measurements over time, correlate them with a slope signature library to predict time-to-failure, and automatically adjust operations to prevent failures, using machine learning to refine predictions and reduce false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual inspections and basic sensor data interpretation are used, then the system is simple and easy to operate, but the ability to predict component failures is insufficient and prone to human error
Solution Approach 1:
The system performs preliminary analysis by calculating rate of change (slope) of operational parameters and comparing them against pre-stored slope signature patterns that represent known failure modes. This preliminary action enables early detection of degradation trends before actual failure occurs, improving prediction accuracy without requiring complex real-time analysis of every parameter fluctuation.
Solution Approach 2:
The patent introduces slope signatures as an intermediary layer between raw sensor data and failure prediction. Instead of directly interpreting complex sensor data, the system computes slopes (rates of change) and matches them against reference slope signatures. This intermediary transformation simplifies the prediction process while maintaining high reliability, as the slope comparison approach is computationally efficient yet effective at identifying degradation patterns.
2Reliability
If continuous monitoring of operational parameters is implemented, then the ability to identify impending failures is improved, but the loss of time for data processing and analysis increases
Solution Approach 1:
The system extracts only the most critical feature from continuous sensor data - the rate of change (slope) of operational parameters. By focusing on this single extracted feature rather than analyzing all raw data continuously, the system maintains high reliability in failure detection while minimizing data processing time. The slope calculation provides a condensed representation of degradation trends that can be quickly compared against reference patterns.
3Measurement precision
If basic sensor monitoring is used, then the device complexity is low, but the measurement precision for detecting early signs of failure is insufficient
Solution Approach 1:
The system transforms operational parameters by calculating their rates of change (slopes) over time. This parameter transformation enhances measurement precision for detecting early degradation, as the slope of parameter change often reveals degradation trends before the parameters themselves deviate significantly from normal ranges. The transformation from static parameter values to dynamic rate-of-change measurements improves sensitivity to early failure signs without requiring additional sensors or complex hardware.
Data Source
AI summary
A system and process for predicting the failure of a machine begins with the step of loading a slope signature library into the control system, in which the slope signature library correlates time-to-failure based on rates of change of one or more measured conditions. The process includes the steps of activating the machine, determining baseline measurements, and detecting an out-of-spec measurement. Once an out-of-spec measurement is made, the process includes the determination of the rate of change for the out-of-spec measurement. A slope signature is calculated based on the rate of change for the measured condition, which is compared against the slope signature library to determine a predicted time-to-failure based on the calculated slope signature, and outputting the predicted time-to-failure. The process can be used to modify the operation of the machine to extend the predicted time-to-failure.