Electronic Thesis and Dissertation Repository

Degree

Doctor of Philosophy

Program

Electrical and Computer Engineering

Supervisor(s)

Dr. Jin Jiang

Abstract

The dynamic response of the process sensors that supply real-time data to the safety systems in nuclear power plants (NPP) play a vital role in preventing plant accidents. If a critical process temperature, pressure, level, or flow experiences a step change, for example, the sensors that measure the process variable must act quickly to actuate the safety systems that will mitigate the consequence of an undesirable process excursion. The research conducted for this dissertation has been performed to ensure the prompt response of critical sensors by advancing, refining, validating, and implementing new methods for measuring the response time of temperature, pressure, level, and flow sensors in NPP safety systems. The essential significance of the new methods is that they can be performed remotely on installed sensors at operating conditions, thereby providing the actual in-service response time as opposed to the unrealistic response time provided by the manufacturer or by offline testing.

The in-situ response time testing technique for temperature sensors is referred to as the Loop Current Step Response (LCSR) test. This technique is based on heating the sensor internally by applying a step change in the DC current to the sensor extension leads in the plant control room. The DC current heats the sensing element of the sensor, resulting in a temperature transient that is then analyzed to provide a true sensor response time, which accounts for all process conditions as well as for the effects of installation and aging. This dissertation presents the theoretical foundation of the LCSR, the details of the author’s extensive experimental research to validate and refine its use in multiple nuclear plant safety applications, and the assumptions that support the validity of the author’s research and experimental results.

The in-situ response time testing technique for pressure, level, and flow transmitters is the so-called noise analysis method. This method is based on recording and analyzing the inherent process fluctuations present at the output of transmitters while the plant is operating. These fluctuations (noise) arise from random flux, turbulent flow, random heat transfer, process control action, and vibration. They are separated from the output of the transmitter by signal conditioning, recorded for about an hour, and analyzed in frequency and/or time domain to yield the response time of the pressure sensing system. This dissertation describes the theoretical foundation of the noise analysis technique, the details of the experimental research that the author has conducted for this dissertation to validate and expand the scope of this technique in actual plant applications, and the assumptions informing the author’s confidence that the research in this dissertation validates the noise analysis technique. The significance of the noise analysis technique is that it not only measures the in-service response time of the transmitter but also of its sensing lines. In contrast to other methods, it can thereby account for the effect of sensing-line length, blockages, and voids on sensor response time.

As part of this research, both the LCSR and noise analysis techniques were validated through extensive laboratory measurements performed on temperature and pressure sensors of the types used in nuclear power plants. The author has used these results to indicate where these methods are most effective but also where they may pose significant uncertainties or may fail.


Share

COinS