UUKI Yoshitaka1 MIYAJI Nobuo1
The aim of VigilantPlant AE (Asset Excellence) is to achieve high performance in plant operation. Improvement of the plant operation rate is a key to reliable and safe operation, as well as to reducing operation costs and improving efficiency. In addition to conventional periodic inspection and maintenance, advanced diagnostics contributes to the realization of reliable and safe operation by using intelligent sensors even during plant operation. Field instruments, such as differential transmitters and flow meters, are able to measure process values, calculate process variables, and send diagnostics information to the host system. This paper introduces diagnostics technology that provides great improvements in plant performance and system availability for less cost.
Yokogawa Electric Corporation promotes the standardization of FOUNDATION™ fieldbus as the communication foundation supporting Asset Excellence for VigilantPlant, which is an approach aimed at the advancement of plant instruments and equipment. With the recent dramatic advancement of fieldbus technologies, information processing technologies, and network speed, it has become possible to process various kinds of information inside the field sensors and transmit the results to the distributed control system (DCS) and host computer. Plant Resource Manager (PRM) that is designed to manage online diagnostic information in an integrated manner has already been presented in other documents. It is, however, a group of individual diagnostic technologies that supports the substantial foundation of diagnosis. This paper presents the diagnostic technologies to be used for prognostic diagnosis, which employs multiple-sensor information that is available thanks to the recent advancement of field instruments (multiple functions, higher speed, and transition to digital).
|Figure 1 Configuration Example of Diagnostic Functions for a Field Instrument|
FOUNDATION fieldbuses are applied to diagnostic technologies at plants. Figure 1 shows the configuration of diagnostic functions inside a field instrument. The diagnostic functions are implemented in the Sensor Transducer Block (STB) and the diagnostic results are passed as the status signal to the Analog Input Function Block (AIFB). In STB, diagnostic variables are obtained by a calculation based on the measured process values for flow rate, pressure, temperature, and the like, and logical operations are performed for diagnosis. The diagnostic results are transmitted as alarm signals to a host computer and used to support the overall maintenance plan.
Processing the diagnostic variables inside the field instrument allows for significantly reducing the volume and time of communication for transmitting the raw data to the host computer, thus shortening the diagnostic interval, and ultimately, increasing the speed and accuracy of diagnosis.
Since diagnostic programs are updated continuously, downloading the software using digital communication allows for obtaining the latest diagnostic algorithms without replacing the printed circuit board.
Figure 2 Diagnostic Analytical Technologies and Applications
A field instrument is responsible for measuring the physical value and chemical composition of elements at a plant, such as temperature, pressure, flow rate, pH value, and concentration, and for providing the measurement information for the control system at the plant. A field instrument equipped with digital communication functions, such as a fieldbus, is required to provide many of the functions shown in Figure 2, in addition to measurement accuracy, measurement range, environmental resistance, or cost efficiency. Specifically, a field instrument is required by customers to provide information that supports maintenance and security for plant operations based on the results of the instrument self-diagnosis, the diagnosis of the field around the instrument, the control loop and equipment diagnosis, the instrument performance diagnosis, and the like. It is necessary to obtain diagnostic information based on multifunctional information of the existing fieldbus transmitters, instead of installing new sensors.
|Figure 3 Configuration of Diagnostic Algorithms Using Diagnostic and Process Variables|
As shown in Figure 3, in order to obtain diagnostic information, it is necessary to combine the process variables for the physical value and chemical composition with the diagnostic variables related to signal fluctuations, drifts and other variables, which were conventionally viewed as disturbances. In order to evaluate the diagnostic result, a proper diagnostic evaluation function needs to be defined by combining multiple diagnostic variables. A diagnostic variable includes an irregular disturbance than a process variable is, so the diagnostic variable is handled as a random variable. The diagnostic evaluation functions used combine the statistical processing operations such as the mean value and standard deviation for random variables, the differential and ratio operation for eliminating the disturbance, the compensation operation based on other measurement information, and the operation that performs comparison with the standard state, and the like. In addition, a Fast Fourier Transform (FFT) for transforming a time range variable to the frequency range variable, a Wavelet Transform for accumulating the result of multiplication of a nonstationary process variable by a properly assumed basis function, and the like can be used as a diagnostic evaluation function.
|Figure 4 Example of Steam Trace Instrumentation|
|Figure 5 AFX Resistance Value Trend of Electrode Fouling Material|
|Figure 6 Impulse Line Pressure Instrumentation of a Differential Pressure Transmitter|
The following shows specific diagnostic applications.
The impulse line blocking diagnosis is performed based on changes caused by the fluctuations of a pressure signal that can be measured in the flow of the pipe line during normal operation. A diagnostic function F called "Blockage Factor," which is obtained by combining the fluctuation correlation coefficients of differential pressure, high-pressure-side pressure, and low-pressure-side pressure, is employed to identify which impulse line is blocked.
Figure 7 shows the difference in the fluctuation correlation coefficient CorL for the differential pressure and the low- pressure-side pressure between the normal state and the high- pressure-side impulse line blockage state. When the low- pressure-side pressure fluctuation and the differential pressure fluctuation are plotted as 2 random variables in the 2-dimensional diagram, the CorL value for the high-pressure side blockage state is close to –1. On the other hand, although not shown in the figure, the correlation coefficient CorH value for the random variables for the high-pressure-side pressure fluctuation and the differential pressure fluctuation on the low pressure side is close to +1. The diagnostic function F, which is obtained by combining the CorL and CorH values based on the principle equation in Figure 8, and which is normalized to ± 1, is used for impulse line blocking diagnosis. When this function is close to +1, the high- pressure-side impulse line is diagnosed as being blocked, and when the value is close to –1, the low-pressure-side impulse line is diagnosed as being blocked.
|Figure 7 Correlation between Low-pressure-side Static Pressure
Fluctuation and Differential Pressure Fluctuation in the case of
A field instrument equipped with digital communication functions such as a fieldbus is able to perform diagnosis on its own. However, when performing advanced diagnosis, such as trend diagnosis based on a trend graph, or multivariable range diagnosis, it is necessary to understand the results visually by combining a Plant Resource Manager (PRM) system and fieldbus instruments. Furthermore, in many cases of actual diagnostic practices, the standard characteristics of diagnostic variables are established as reference data in a normal state and then compared with the variables of diagnosis that is performed each time. In some cases, a lot of reference data is prepared in advance for future changes in the plant operation conditions. When using diagnostic results, for example, setting a proper threshold value based on the trend graph, or revising a maintenance plan before a problem occurs, dedicated diagnostic packaged software installed in the PRM system is required to perform advanced diagnosis.
Figure 9 shows the screens of the Impulse Line Blocking Diagnosis (ILBD) software, which enable the registration of instruments for impulse line blocking diagnosis (B), the observation of the diagnostic result trend and the setting of thresholds for the blocking level (C), or the registration of reference data for diagnosis (D).
|Figure 8 Relation between One-side Blockage
Factor F and Blockage
The diagnostic technologies presented in this paper allow for always monitoring the process state onsite using intelligent sensors, for analyzing the trend in changes, and for providing information for preventive maintenance at earlier stages. It is necessary to further improve the practical issues; for example, the application range of diagnostic technologies and the method to evaluate the diagnostic accuracy.
It is also expected that the variation of input signals such as ultrasonic and optical signals will be increased, and faster diagnosis based on the measured values in the high-frequency range will be developed. The existing diagnosis will ultimately progress from the localized diagnosis to the more advanced diagnosis that is based on the 2-dimensional and 3-dimensional information obtained by sensors at multiple points.
Figure 9 ILBD Diagnostic Screen for PRM
The AXF magnetic flow meter is a sophisticated product with outstanding reliability and ease of operation, developed on the basis of decades of field experience. Based on FOUNDATION™ fieldbus specifications, AXF Fieldbus Magnetic Flow meter models offer more flexible instrumentation.
The AXFA11 magnetic converter has been developed based on Yokogawa's decades long experience in magnetic flowmeters. The AXFA11 continues the tradition of high quality and reliability that has become synonymous with the Yokogawa name and in addition features an even higher level of performance and increased functionality.
Traditional-mount Differential Pressure Transmitter based on the EJX-A Series.
Differential Pressure Transmitter attached to an IFO assembly based on the EJX-A Series.
Differential Pressure Transmitter with Remote Diaphragm Seals based on the EJX-A Series.
Traditional-mount Differential Pressure Transmitter designed for Draft Range applications based on the EJX-A Series.
Traditional-mount High Static Differential Pressure Transmitter based on the EJX-A Series.
Flanged-mounted Differential Pressure Transmitter designed for Liquid-level applications based on the EJX-A Series.
Traditional-mount Absolute Pressure Transmitter based on the EJX-A Series.
Traditional-mount Gauge Pressure Transmitter based on the EJX-A Series.
Gauge Pressure Transmitter with a Remote Diaphragm Seal based on the EJX-A Series.
Traditional-mount High Gauge Pressure Transmitter based on the EJX-A Series.
In-Line Mount Absolute Pressure Transmitter based on the EJX-A Series.
In-Line Gauge Pressure Transmitter based on the EJX-A Series.
High Performance In-Line Mount Absolute Pressure Transmitter based on the EJX-A Series.
In-Line Mount High Performance Gauge Pressure Transmitter based on the EJX-A Series.
This transmitter precisely measures differential pressure, static pressure, and process temperature; then uses these values in a high-perfomance on-board flow computer to deliver fully compensated Mass Flow.
Designed specifically for high static pressure applications, this transmitter precisely measures differential pressure, static pressure, and process temperature; then uses these values in a high-perfomance on-board flow computer to deliver fully compensated Mass Flow.
Plant Resource Manager (PRM) is a key platform for the Yokogawa VigilantPlant Asset Excellence initiative which aims to improve operations and maintenance and maximize the reliability and availability of plant assets by achieving greater predictability.