Pressure Transmitters

AMS provides equipment, training, and services for response time testing of pressure transmitters in nuclear power plants and for detection of blockages, voids, and leaks in pressure sensing lines. There are both conventional testing methods and modern testing methods. The latter provides the means to perform response time testing remotely from the control room area while the plant is on-line. As such, the modern methods are also referred to as on-line methods.

 

1. Conventional Methods

The conventional methods for response time testing of pressure transmitters usually involve a hydraulic pressure generator which produces a test signal in the form of a step or ramp. The ramp test is more commonly used than the step test because design basis accidents in nuclear power plants usually assume pressure transients which approximate a ramp.

The pressure test signal, as generated by the ramp test equipment, is fed simultaneously to the subject transmitter and to a high-speed reference transmitter. The outputs of the two transmitters are recorded on a dual channel recorder and used to identify the response time of the transmitter. (See Figure) This response time value is sometimes referred to as the asymptotic ramp time delay. In practice, however, the response time of a pressure transmitter is usually defined as the time delay between the response of the reference transmitter and the test transmitter as they pass through a setpoint.

For step tests, the response time is either determined as the time that it takes for the transmitter to pass through a setpoint, in which case the response time is sometimes referred to as "time to trip", or is defined as the time that is required for the sensor output to reach 63.2 percent of its final steady state value. The latter definition corresponds to the response time of a first order system, and is used in defining the step response of pressure transmitters, even though pressure transmitters are not necessarily first order.

2. On-Line Methods

Two methods have been developed and validated for in-situ response time testing of pressure transmitters as installed in operating processes. These methods are referred to as noise analysis technique and Power Interrupt (PI) test. The noise analysis technique can be used for response time testing of most pressure transmitters, while the PI test is applicable only to force-balance pressure transmitters. Force-balance pressure transmitters are also testable by the noise analysis technique, but the PI test is used more often than noise analysis because the PI test involves a simple analysis.

 

Noise Analysis Technique

The noise analysis technique is based on analyzing the natural fluctuations that exist at the output of pressure transmitters while the plant is operating. These fluctuations (noise) may be due to turbulence induced by the water flow, core random heat transfer, or other naturally occurring phenomena. The noise is extracted from the transmitter output by removing the DC component of the signal and amplifying the AC component. The AC component is then passed through a low-pass filter for anti-aliasing and removal of high frequency electrical noise and interferences. (See Figure) After this, the signal is digitized and stored for subsequent analysis.

The analysis of noise data is performed in frequency domain and/or time domain, and is based on the assumption that the dynamic characteristics of the transmitter are linear. The frequency domain and time domain analyses are two methods for response time determination of pressure transmitters. It is the AMS practice to analyze the data with both methods and average the results.

For frequency domain analysis, the Power Spectral Density (PSD) of the noise signal is obtained through a Fast Fourier Transformation (FFT) algorithm. An appropriate mathematical function is then fit to the PSD from which the response time of the transmitter is calculated.

In the time domain analysis, the raw noise data is used in a Univariate Autoregressive (UAR) model to obtain the impulse response and then the step response from which the transmitter response time is calculated.

The validity of the noise analysis technique has been established by AMS through laboratory testing of representative transmitters from such manufacturers as Barton, Foxboro, Rosemount, Veritrak, Tobar, and Fischer & Porter. The details of the validation effort are covered in two reports which AMS wrote for the U.S. Nuclear Regulatory Commission (NRC). These reports are referred to as NUREG/CR-5383 and NUREG/CR-5851 and are available from the U.S. Government Printing Office.

Prior to any time domain or frequency domain analysis, the suitability of the noise data for a reliable analysis must be examined by a computer scanning and screening of the raw data. This is accomplished using data qualification algorithms that check for the stationarity and linearity of the data. This includes plotting the amplitude probability density (APD) of the data for visual inspection of skewness and nonlinearity. Calculating the skewness, flatness, or other descriptors of noise data is performed to ensure that the data has a normal distribution and does not contain saturated blocks and other undesirable characteristics.

Power Interrupt Test

 

The Power Interrupt (PI) test is a method for in-situ response time testing of force-balance pressure transmitters. The test is based on a momentary interruption of the electrical supply power normally used to activate the transmitter. The test is performed by turning the power to the transmitter OFF for a few seconds, and then ON. When the power is turned ON, the transmitter provides a transient output that is digitized and then analyzed to give the response time of the transmitter (see Figure). This response time corresponds to the response time that would be obtained for the transmitter using the conventional step or ramp test.

The analysis of PI data involves a proprietary AMS software which employs a data stripping algorithm followed by a data fitting routine to yield the response time of the transmitter. The stripping algorithm peels off the extraneous data leaving a response transient that is then fit to a mathematical model to give the response time of the transmitter.

The PI test accounts for the response time of both the mechanical and the electronic components in the transmitter and thereby gives the response time of the complete electromechanical system of the transmitter; (does not account for any sensing line blockage). The validity of the PI test has been established in laboratory tests by comparing the results of the PI test with those of the conventional ramp or step tests performed on a number of force-balance pressure transmitters. The details are covered in two reports which AMS wrote for the U.S. Nuclear Regulatory Commission (NRC). These reports are referred to as NUREG/CR-5383 and NUREG/CR-5851.