Airede, Yusuf Okahamame (2019) Intrusive and Non-intrusive Methods for Fluid Diagnostics and Flow Condition Monitoring in a Control valve. Doctoral thesis, University of Huddersfield.

Process control systems consist of numerous control loops, linked together in producing a particular product to be used for carrying out experiments or to be sold. For each of these loops, they are designed to ensure that significant flow parameters such as temperature, velocity, pressure, level, flow, etc. operate within a set range to ensure end product quality. Signal and disturbances for external control loops are sent to these control loops, or sometimes, generate their own internal disturbances, adversely affecting these variables. After measurement, comparison and calculation, the strategy that has been selected by the specific controller such as (Proportional, Proportional Integral, Proportional Integral Derivative, etc.) must be implemented by a final control element.

One common FCE (final control element) largely used in the industry for process control is the control valve. The manipulation of the flowing fluid, such as water, chemical compounds, steam, gas, etc., are done by the control valve for load disturbance to be compensated, and ensure that the process variable and the desired set point are as close as possible. Due to its importance in a flow loop, it is very important to know the control valve flow characteristics. Because of the fast development in simulating flow and using numerical technique for flow diagnostics, understanding flow inside the valve has become possible for the valve performance to be estimated.

This thesis investigates methods for diagnosing flow for both single phase and multiphase flow under various flow conditions and probe designs.

This thesis reports the results of experiments on the control valve flow characteristics using a range of statistical parameters obtained from: airborne acoustic signal microphone, vibration signal from an accelerometer on the valve body and fivehole multihole probes (MHP) placed ingeniously inside the valve at different sections of the valve inlet and outlet. An assessment on the behaviour of the fluid using these three methods has been presented.

For the multihole probe (MHP), different probe lengths and heights have been used to obtain local flow parameters and analyse how these parameters vary at different sections from inlet to outlet.

Detailed experimental investigations were conducted for diagnosing the valve flow behaviour. The signatures from these sensors (vibration and acoustic) have been analysed by making use of different statistical parameters including peak amplitude, kurtosis, RMS (root mean square), variance and Peak-to-Peak. In addition, both signals from the vibration and acoustic sensors were then transformed to frequency domain and analysed using Fast Fourier Transform (FFT) and mean frequency. Analyses for both signals in frequency domain were also carried out under various ranges of frequency. The results revealed that vibration technique gives a clearer picture into how the signal changes with different flow and valve behaviour. This research also found that frequency range between 0 to 1 KHz was more sensitive in determining the valve behaviour at various conditions. Concerning this research, it can be suggested that lower frequency range sensors (vibration and accelerometer) can be used to capture the behaviour of the control valve and detect any form of fault, which are cheaper than higher frequency range sensors.

The results also reveal that the combination of all of these techniques could be applied to increase the reliability of the valve and fault detection. The combination of all experimental methods can be considered robust, providing detailed information about the valve performance and flow diagnostics, hence, assisting in prolonging the valve life and protect the control system from any emergency shutdown.

FINAL THESIS - Airede.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (18MB) | Preview


Downloads per month over past year

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email