Conventional Valve Stiction Analysis for Stable Control Loops
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional valves in process plants suffer from stiction issues, leading to oscillations, poor disturbance rejection, and poor set-point tracking, which can cause production shutdowns and quality issues, and existing methods for stiction analysis are not applicable to legacy valves.
Innovation Solution
A system and method for stiction analysis of conventional valves using inverse modeling to estimate stick and jump values, which includes a controller, a system processor, and a display to provide real-time feedback for valve operation optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If periodical valve maintenance is performed to resolve stiction, then valve reliability is improved, but production downtime and energy loss increase
Solution Approach 1:
The system continuously monitors valve operation parameters (process variable PV, controller output OP, manipulated variable MV) and provides real-time feedback on stiction levels. This enables proactive detection of stiction issues before they cause complete valve failure, allowing maintenance to be scheduled at optimal times rather than through forced periodic shutdowns.
Solution Approach 2:
The valve system performs self-diagnosis by automatically calculating stiction metrics from its own operational data without requiring external intervention or shutdown. The system computes stick values S and jump values J internally, enabling the valve to monitor its own health status and trigger maintenance alerts autonomously.
2Measurement precision
If smart valve technology is deployed to enable stiction monitoring, then stiction detection capability is improved, but device complexity and cost increase
Solution Approach 1:
The patent introduces a stiction processor as an intermediary component that sits between the existing conventional valve and the control system. This processor receives standard process variables (PV, OP, MV) from the existing control infrastructure and performs stiction analysis without requiring modification of the valve itself or the core control system, thus adding monitoring capability while minimizing disruption to existing simple architectures.
Solution Approach 2:
The stiction analysis system is designed to work with conventional control loops using standard process variables (PV, OP, MV) that already exist in virtually all process control systems. By utilizing existing universal control signals rather than requiring valve-specific sensors or communication protocols, the system achieves broad applicability across different valve types and control platforms without increasing complexity for each specific application.
3Device complexity
If conventional valves are operated without stiction compensation, then device simplicity is maintained, but control loop performance deteriorates
Solution Approach 1:
The system performs preliminary stiction analysis by continuously calculating stick values S and jump values J from operational data before stiction becomes severe enough to cause control failures. By detecting and quantifying stiction in advance, the system allows the controller to pre-adjust control strategies or schedule maintenance before performance degradation occurs, maintaining control loop reliability without complex real-time compensation mechanisms.
Data Source
AI summary
Stiction analysis of a conventional or legacy valve operation in a fluid process plant is described. The method includes receiving a series of setpoint signals, SP, generating, by the controller, a series of controller output signals, OP, driving the valve operation by applying the OP, to the valve, measuring a series of process variable signals, PV, downstream of the valve, receiving the series of process variable signals at the controller, performing inverse modelling on the PV, to estimate a manipulated variable, , receiving and analyzing by a system processor, the SP, the OP, and the , and outputting the estimated stick values, S, the estimated jump values, J, an average of the estimated stick values, an average of the estimated jump values, a confidence interval for the average of the stick values, and a confidence interval for the average of the jump values.