Valve Fault Detection

The valve is a piece of common equipment in industrial manufacturing, which is mainly used to control the flow and cut-off of the medium. Valves are subject to environmental, medium impact, corrosion, high-temperature creep and other factors, and are prone to failure. Taking nuclear power valves as an example, gate valves and electromagnetic valves have the highest failure probability, followed by check valves, regulating valves and pressure reducing valves. If the valve fails, it can't be found out in time and handled accurately, which will bring hidden dangers to industrial production and personnel safety. With the rapid development of information technology, valve fault detection technology can timely and accurately grasp the operating state of the valve, which provides an important way for valve maintenance and troubleshooting. However, at present, there are many methods for valve fault detection, with different applicable conditions and no uniform standard. Therefore, the summary and research progress of valve fault detection technology is of great practical significance to industrial stable production, cost reduction and resource utilization.
 
Failure Mode and Causes

The failure mode and cause analysis of the valve is the basis to ensure the reasonable design and normal operation of the valve. The main failure modes and causes are shown in Table 1.
 
Fig. 1 Principle of pressure drop method
 
Table 1 Main failure modes and reasons of valves
Leakage: ① Unreasonable design. There is no reliable design system combined with the operating environment, and there is no reasonable design for sealing surface, packing, sealing ring, material roughness and other factors.
② Unreasonable structure. The valve structure size, sealing interval, sealing specific pressure, etc. are not properly designed.
③ Unreasonable materials. The mechanical properties, corrosion resistance and composition of sealing materials are not reasonably selected.
Rupture of valve body: ① Improper selection of valve body materials or defects in materials, such as sand holes, air bubbles, shrinkage cavities, etc.
② External environmental factors, such as low temperature, high temperature, high pressure, etc., cause creep, fatigue, impact and other phenomena of the valve body, resulting in the valve body cracking.
③ Strong vibration, resonance or violent disassembly and assembly make the valve force uneven and crack. During the working process, due to the medium impact, the crack is prolonged and the valve body is damaged.
Stuck: ① The valve is stuck due to deformation caused by materials, stress superposition and other factors or physical reasons such as excessive and tight packing.
② The valve is stuck due to other reasons such as pollution caused by working environment or chemical corrosion.
Vibration and noise: ① Mechanical vibration. Unreasonable layout of the number and spacing of process pipes, elbows and valves leads to vibration or pressure pulsation of medium fluid in the pipes.
② The external environment resonates with the medium fluid, thus causing valve vibration.
③ cavitation vibration.
④ Hydrodynamic vibration.
 
Although many scholars have analyzed and improved the main failure forms of valves from the aspects of materials, design, structure and operation, there are still some problems.
 
(1) Most of the analysis improvements are aimed at specific situations. How to carry out quantitative analysis from qualitative analysis and obtain widely applicable methods needs further study.
(2) In the actual work of the valve, the flow of medium fluid is complex, and many factors such as fluid impact, corrosion and thermal stress may be superimposed together. However, in the simulation, some conditions will be ignored or idealized, which is unfavorable to the valve design and reliability analysis.
(3) After the valve failure occurs, it is of great significance to use the fault detection technology to improve the accurate identification of the occurrence and development of the valve failure to ensure the safe operation of the process system and the reliability research of the valve.
 
3 detection technology
The technology of valve fault detection is to use sensors to monitor the running state of the valve and obtain the working signal of the valve. By comparing the historical state information, the operating state of the valve is predicted and fed back. Extract and process the abnormal signal, analyze the cause and mechanism of valve failure, determine the type and level of valve failure, and decide whether to stop and maintain the valve. Timely and accurate acquisition of abnormal signals of valves is a prerequisite for valve fault detection. According to the failure forms and causes in Table 1, signal acquisition and sensing can be divided into two types: detection methods for valve leakage and detection methods for valve failure forms such as jamming and cracking.
 
3.1 Detection method of valve leakage
The earliest methods of valve leakage detection technology are pressure detection method, vacuum detection method and bubble method. This method requires the valve to stop working, and the valve should be removed from the pipe fittings for testing. This kind of method is easy to cause valve damage, which belongs to off-line detection and has low detection efficiency. In order to realize on-line nondestructive detection of valve leakage in industrial field, experts and scholars have conducted a lot of experimental research, including pressure drop method, vibration method, thermal infrared method, ultrasonic detection method, optical fiber leakage method and acoustic emission detection method.
 
(1) Pressure drop method
The pressure drop method is that clean and dry compressed air is led to the sealed detection valve, and pressure sensors are installed at the upstream and downstream, respectively. Through the processes of inflation, pressure keeping, detection and exhaust, the pressure change in the valve to be detected is monitored and recorded (Figure 1). If there is obvious change in the upstream and downstream pressure of the valve, it means that the valve has leakage. In order to reflect the leakage situation more intuitively, the pressure change can be converted into leakage (Figure 2).
 
Fig. 2 Conversion curve of pressure drop and leakage rate
 
 
The pressure drop method has the advantages of simple operation, low cost, easy to realize automatic detection, etc., but it is greatly influenced by temperature, and the detection effect of tiny leakage is not ideal. The pressure drop method is divided into direct pressure method and differential pressure method. The direct pressure method directly detects the change of the internal pressure of the valve under test, and the differential pressure method measures the pressure difference between the valve under test and the standard valve. Compared with direct pressure method, differential pressure method can reduce the influence of temperature change on pressure measurement. Xiao Xiangzheng studied the influence of temperature on the differential pressure method. Experiments show that when the temperature changes below 1℃, the influence on the pressure measurement can be ignored, otherwise, the pressure measurement should be corrected according to the temperature correction formula and effective volume formula obtained from experiments. Longjie used direct pressure method and differential pressure method to test the air tightness of PE pipe valve, established mathematical model and test device. Through simulation test, the influence of temperature and other factors on pressure measurement was investigated, and the leakage parameters were adjusted. The minimum detection leakage rate was 8.5× 10-4 Pa m3/s, which was in line with the experimental expectation.
 
(2) Thermal infrared method
Thermal infrared method refers to the use of detection instruments to capture the surface temperature distribution of objects that can't be directly seen by human eyes, and to capture the temperature difference between the upstream and downstream of the valve, so as to form an infrared image, so as to judge whether the valve leaks. The infrared method is mainly divided into active detection (Figure 3) and passive detection (Figure 4). It has the advantages of high detection sensitivity, convenience, rapidity and quick response, but it is only suitable for cases with obvious temperature difference. Guan Shitao and others [3] installed a temperature transmitter beside the valve to transmit the temperature signal to the gas valve leakage detection system based on ZigBee technology, thus realizing the detection of valve leakage.
 
Fig. 3 Active detection
 
 
Fig. 4 Passive detection
 
 
(3) Vibration method
The vibration method uses sensors to capture the vibration signals of valves and pipelines. By processing the vibration signals, the characteristics of time, frequency and acceleration are analyzed, and the modal parameters are obtained. Then, the valve leakage is judged by comparing with the normal modal values. Thompson and others have done a lot of experimental research on gas valves by vibration method, and the results show that this method is easy to operate and has high detection sensitivity. In laboratory environment, vibration method can detect valve internal leakage with low pressure difference and small leakage rate. The detection results are not affected by valve type, pipe material and thickness, but only related to pipe diameter. However, its disadvantage is that it is sensitive to industrial noise. Some scholars have studied it and found that the vibration method can effectively capture the noise between 0 and 0~20 kHz, which will bring errors and uncertainties to the valve leakage detection.
 
(4) Optical fiber detection method
Optical fiber detection method is mainly suitable for underground pipelines and valves. According to the principle of Joule-Thomson effect, when the pipeline leaks, the temperature near the leakage source will also decrease. By monitoring the local temperature change, the valve leakage can be monitored and located. The optical fiber method has high detection efficiency and positioning accuracy, but when the leakage is small, the temperature change near the leakage source is small, and the detection sensitivity of the optical fiber sensor is required to be high, so the cost is relatively high. It is close to the optical fiber pipeline valve, and the detection optical fiber and the leakage point cannot be located at both ends of the pipeline (Figure 5). Three optical fibers should be buried and evenly distributed around the pipeline (Figure 6).
 
Fig. 5 Detection optical fiber and leak point are located at both ends of the pipeline.
 
 
Fig. 6 Three optical fibers are evenly distributed around the pipeline.
 
(5) Acoustic emission detection method
Acoustic emission detection method is a widely used nondestructive testing technology at present. Its principle is to capture the continuous high-frequency fluctuation signal generated by the valve leakage medium scouring the valve body, and realize the valve leakage detection according to the law that the leakage acoustic emission characteristics are proportional to the leakage rate. In order to filter out environmental noise, the filtering band can be set at 100~400 kHz. Gao Qianxia [5] used acoustic emission technology to detect the leakage of valves under different conditions, and obtained three formulas for calculating the leakage rate based on acoustic emission characteristic values, thus realizing the quantitative diagnosis of valve leakage faults. The advantages of acoustic emission detection method are as follows.
① There is no need to disassemble the detected equipment, which will not damage the equipment, and the detection method is simple and convenient.
② There is no need to approach dangerous valves such as high temperature and high pressure.
③ The detection is fast, real, intuitive and quantitative. Acoustic emission detection method also has some limitations, because its central frequency is generally around 150 kHz, and its sampling frequency is as high as 1 MHz, which not only requires fast processing speed of hardware equipment and large memory, but also requires high requirements for signal analysis and processing, feature extraction and diagnostic modeling. The principle of acoustic emission detection method is shown in Figure 7.
 
Fig. 7 Principle of acoustic emission detection
 
 
(6) Ultrasonic testing method
When the valve leaks, the medium flows out from the leakage point. When the leakage hole is small and the pressure is high, the sound wave emitted is higher than 20 kHz, which can be captured by the ultrasonic sensor although it can't be heard by human ears. Using the characteristics of directional propagation of ultrasonic wave and inverse ratio to distance, the leak location can be quickly found. Considering the influencing factors of environmental noise, by comparing the spectrum distribution of leakage sound and noise, it can be found that if the central frequency of ultrasonic signal generated by valve leakage is 40 kHz, the accuracy of ultrasonic signal can be ensured and the surrounding environmental noise can be filtered out. Ultrasonic testing method also has its limitations. First, the pressure must be greater than 0.34 MPa or higher to ensure that the frequency of the sound emitted by the leak can reach the detectable range. Second, the leaking liquid can not be accurately identified, which is more suitable for gas or steam medium. Wang Chao [6] put forward that pressure and temperature can be used as auxiliary means to make up for the defects of ultrasonic detection and better detect the leakage of valves. The spectrum distribution of valve leakage sound and noise is shown in Figure 8.
 
Fig. 8 Spectrum distribution of leakage sound and noise
 
3.2 Detection method of valve sticking, rupture and other faults
(1) Install the positioner, which is mainly aimed at the regulating valve, and monitors the dynamic characteristics of the valve by detecting the valve stem displacement, input signal, unbalance force, friction and other parameters, and reducing the equivalent time of the secondary circuit. Judging the state of the valve according to the change of parameters.
(2) Using strain gauges or strain gauges to measure the stress, torque and other information of the measured part. The numerical simulation and strain test of the valve shell stress show that the shell surface is not unidirectional stress, so strain gauges should be used for testing, and the results are in good agreement. Clamp-type thrust sensor can be used to measure the thrust of valve stem, and acceleration sensor, dynamic strain gauge and pressure sensor installed at the end of valve can realize real-time measurement of valve data.
(3) Combining infrared technology with ultrasonic technology, cracks in important parts such as valve body can be detected. In addition to the common detection methods mentioned above, there are water detection method, halogen method, helium mass spectrometry, etc., but they are not suitable for large-scale industrial production due to errors, easy pollution and other reasons. The detection accuracy of different detection methods is shown in Table 2.
Table 2 Detection accuracy of different detection methods
Pressure drop method: detection accuracy/(pa m3/s): 10-5 ~ 10-3
Ultrasonic method: detection accuracy/(Pa m3/s): 10-4 ~ 10-3
Infrared method: detection accuracy/(Pa m3/s): 10-7
Ammonia method: detection accuracy/(Pa m3/s): 10-8
Water inspection method: detection accuracy/(Pa m3/s): 10-6 ~ 10-5
Helium mass spectrometry: detection accuracy/(Pa m3/s): 10-9
Halogen method: detection accuracy/(pa m3/s): 10-7 ~ 10-6
 
4 Conclusion
The research and development of valve fault detection technology is particularly important for industrial production. By discussing different detection methods and summarizing their advantages and disadvantages, it provides convenient conditions for experimental research and provides necessary technical support for the reliability of valves and the smooth progress of industrial production. Due to the influence of service conditions and detection environment, various methods should be considered to assist in order to improve the detection accuracy. For interference signals such as noise, corresponding treatment measures should be taken to prevent adverse effects on detection results. We should continue to study valve testing methods in depth, expand its application range, improve testing efficiency and reduce costs.