Sensor-based performance monitoring of a control valve unit

Abstract This is the second of a series of papers dedicated to the research and development work in the field of industrial control valve condition monitoring and fault diagnosis. This paper reports new results that were produced after conducting a series of tests to monitor the dynamic performance of a typical industrial control valve unit (which consisted of a control valve, an actuator and a digital valve controller) under various operating conditions. These included the gradual blockage of the actuator's vent hole, the build-up of deposits on the valve stem and the incorrectly replaced valve stem seals. These conditions were created by using a precision in-line flow regulator for the actuator's vent hole and a thin layer of a special type of adhesive film that crystallized and bonded to the valve stem and by replacing the polytetrafluorethylene valve stem seals with graphite-type seals under controlled conditions. The effects of these abnormalities on the pressure that was developed in the upper diaphragm casing of the actuator as well as the flowrate were carefully monitored using a range of sensors that were integrated into an intelligent data acquisition and processing system. It was evident that certain types of fault, including the fully blocked actuator's vent hole, could be identified with great accuracy using the responses of the flow transmitter and the pressure transducer. On the other hand, other types of fault, such as the simulated build-up of deposits on the valve stem, required a comparison of their statistical properties in addition to the transducers' responses to determine the condition reliably. The current-to-pressure converter is a vital element of any digital valve controller and its performance was, therefore, monitored under various conditions to establish the effects of the faults on its operation. The paper concludes by outlining the main results and the limitations of the diagnostic techniques utilized as well as the future work, which is to integrate the results into a user-friendly diagnostic system based on the Microsoft visual basic programming language.