Conventional Control Valve Diagnostics via Non-Linearity Index
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional control valves in process plants lack real-time performance monitoring diagnostics, leading to potential unscheduled shutdowns due to unchecked deterioration, as they do not have smart positioners or feedback mechanisms, and their low cost contributes to high usage despite the risk of undetected issues.
Innovation Solution
A method and apparatus that obtain data samples from a database to compute a non-linearity index, chart plots, and apply heuristic rules and fuzzy logics to detect valve problems, including stiction and oscillation analysis, enabling real-time performance monitoring and diagnostics without requiring smart positioners.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional control valves are used without smart positioners, then cost is reduced and usage is increased, but real-time performance monitoring capability is lost leading to undetected deterioration
Solution Approach 1:
The patent introduces an intermediary diagnostic system consisting of a data acquisition module, data preparation module, and analysis module that mediates between the conventional valve and the user. This intermediary captures process variable, set-point, and manipulated variable data from the control loop and performs offline analysis to detect valve deterioration, stiction, and oscillation without requiring modifications to the valve itself.
Solution Approach 2:
The patent replaces the mechanical/electrical feedback mechanisms of smart positioners with a computational diagnostic system. Instead of using physical sensors and feedback loops integrated into the valve, the system uses data acquisition from existing control loop signals and applies computational algorithms (non-linearity index, elliptical fit analysis, stiction detection) to monitor valve health.
2Device complexity
If conventional control valves without feedback mechanisms are used, then device complexity is reduced, but detection of valve deterioration becomes difficult
Solution Approach 1:
The patent creates a feedback loop through data acquisition and analysis. The system continuously collects data from the control loop (process variable, set-point, manipulated variable), analyzes the data using multiple diagnostic techniques, and provides feedback about valve condition including stiction detection, non-linearity assessment, and oscillation identification, enabling users to detect deterioration without complex valve-mounted sensors.
Solution Approach 2:
The patent creates a virtual model or copy of the valve's performance characteristics by analyzing the relationship between manipulated variable and process variable data. Through techniques like elliptical fit analysis and non-linearity index calculation, the system generates a computational representation of valve behavior that reveals deterioration patterns without physically instrumenting the valve.
3Reliability
If real-time performance monitoring is implemented for conventional valves, then reliability is improved, but device complexity increases due to additional sensors and feedback mechanisms
Solution Approach 1:
The patent uses an intermediary diagnostic system that sits between the existing control loop and the user, acquiring data from standard control signals without adding sensors to the valve. The data preparation module filters and validates data, while the analysis module applies multiple diagnostic algorithms to detect valve conditions, providing comprehensive monitoring through software rather than hardware additions.
Solution Approach 2:
The diagnostic system performs multiple functions using a single integrated platform: data acquisition from control loops, data preparation and filtering, non-linearity analysis, stiction detection, oscillation identification, and performance reporting. This multi-functional approach provides comprehensive valve monitoring without requiring separate specialized devices for each diagnostic function.
Data Source
AI summary
A method of qualifying performance of a conventional control valve in a process plant, the valve being controlled by a controller, the method comprising a processor obtaining data samples from a database stored on a server of the process plant, each data sample comprising a process variable, a set-point, and a manipulated variable; the processor computing a non-linearity index from the data samples and determining if the non-linearity index is greater than a threshold value; if the non-linearity index is greater than the threshold value, the processor charting a plot of the process variable against the manipulated variable and determining if the plot has an elliptical or rectangular fit; and if the plot has an elliptical fit, the processor determining if a percentage of the total number of data samples lying within a theoretical ellipse encompassed within the elliptical fit is less than or equal to a preset percentage.