Compressor Surge Cycle Detection via Phase Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Compressor surge events, characterized by flow reversal and pressure instability, can lead to vibration, wear, and potential damage to compressor components, and existing detection systems may not accurately identify and count consecutive surge cycles without false positives or negatives.
Innovation Solution
A method and system for detecting compressor surge cycles by monitoring and filtering surge parameters such as flow differential pressure, suction pressure, and rotational speed, identifying three distinct phases of a surge cycle, and validating the occurrence of a surge event based on threshold differences and time indices, allowing for accurate counting of consecutive cycles without phase lag or double-counting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing detection systems monitor surge parameters to identify compressor surge events, then surge detection capability is provided, but accurate identification and counting of consecutive surge cycles is compromised due to false positives or negatives
Solution Approach 1:
The surge cycle detection is divided into three distinct phases (surge onset, surge peak, surge termination) with specific identification criteria for each phase. This segmentation allows the system to accurately track and count complete surge cycles by recognizing the characteristic pattern of all three phases, thereby reducing false detections while maintaining high detection accuracy
Solution Approach 2:
The system dynamically adjusts surge cycle counting by validating that detected phases form complete cycles rather than isolated events. By requiring the sequence of all three phases to constitute a valid surge cycle, the system adapts to varying surge conditions and accurately counts only genuine consecutive surge cycles, eliminating false positives and negatives
2Speed
If surge parameters are monitored continuously to detect surge events, then real-time detection is achieved, but phase lag and double-counting errors occur in cycle validation
Solution Approach 1:
The system stores the identified phases and their time indices during the surge event, creating a complete record of the surge cycle pattern before final validation. This preliminary action allows the system to validate complete cycles accurately and count them correctly, preventing both phase lag errors and double-counting while maintaining real-time detection capability
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
The subject matter of this specification can be embodied in, among other things, a method that includes receiving one or more surge value parameters, selecting a first value from a selected parameter of the one of the parameters, determining a moving average, determining a rate of change, identifying a first phase of a surge cycle based on determining that the rate of change exceeds a threshold rate change, memorializing the first value, the moving average value, and a first time index, receiving a second value of the selected parameter, identifying a second phase of the surge cycle, receiving a third value of the selected parameter, identifying a third phase of the surge cycle, and providing an indication that one or more compressor surge cycles have occurred.