Sign In

UTP Ignite



Capture.PNG

Project Leader: AP. Ir. Dr Haslinda Zabiri
Department: Chemical Engineering
Expertise: Control Valve Stiction Detection, Quantification and Compensation methods; Modeling, Simulation and Control of CO2 capture systems; Multivariate statistical methods for monitoring and diagnostics of industrial processes; Hybrid Model Predictive Control

Automated Control Valve Health Monitoring System (ValveStic)




TECHNOLOGY READINESS LEVEL: 5​

BRIEF TECHNOLOGY
ValveStic is an automated control valve health monitoring system that uses Artificial Intelligence (AI) for both the detection of stiction, as well as identification of the severity of stiction.

Valvestic 3.PNG

Valvestic 2.PNG 


PROBLEM STATEMENT & CURRENT ISSUES
​Valves are essential assets to processing plants. Through a control valve, control is brought to bear on a process. A 250,000-bpd refinery might have as many as 2,500 of control valves installed for regulating flows.

Underperforming control valves cost the typical process plant thousands of dollars annually. Almost 70% of the total life cycle cost goes for maintenance, which represents a huge opportunity for cost reduction.

Valvestic 1.PNG

Due to its continuous movement, control valves’ wear and aging are inevitable. Based on the faults that occur inside control loops, stiction is a major contributor. Valves that have stiction will cause a sustained oscillation in process variables which subsequently lead to decreased control valves lifespan, and inferior quality end products thus reducing profitability. 

Our research observed that the current OEM-based control valves health monitoring methods suffer from few drawbacks such as usually highly invasive, brand-specific and require external assistance to analyse the results from diagnosis.

USEFULNESS & APPLICATION
ValveStic is a non-invasive valve condition checking software which enable users to check the valves’ condition by just providing the readily available OP (valve opening percentage) and PV (corresponding process variable) data. In return, ValveStic will provide verdict whether the valves are healthy, have stiction or other potential problems. The performance of ValveStic has been validated against commercial OEM-based methods and has shown high accuracy while having the advantages of being:
  • Completely non-invasive (require only routine operating data to diagnose the valve’s health),
  • Widely applicable to various industrial processes regardless of valve brands, and
  • Automated easy to understand diagnostic analysis without the need for external expert inputs.

IMPACT OF THE PRODUCT

Valvestic 4.PNG

MARKET POTENTIAL
  • Opportunity to apply ValveStic at candidate critical processes in the plant. The outcome is to create value from maintenance optimization and increase productivity against conventional baseline practices.
  • Opportunity to recognize ValveStic as a validated technology inventory available at the disposal for existing plants.
  • Opportunity for the long-term involvement of local universities for knowledge transfer and technology validation as well as local vendor companies for sustainability and development of local Thailand technology ecosystem.

INDUSTRY COLLABORATION
  • ​-
INTELECTUAL PROPERTY (IP)
Patent         : PI2019004306
Copyright   : LY2021P06583​