6.5 Noise Analysis Technique

The output of all process sensors in nuclear power plants normally contains fluctuations due to random flux, random heat transfer, turbulence, vibration, and other mechanical and thermal hydraulic phenomenon. These fluctuations (noise) can be extracted from the sensor output and analyzed to yield the sensor's response time. The noise analysis technique is normally used for in-situ response-time testing of pressure, level, and flow transmitters. Furthermore, the technique is used to detect blockages, voids, and leaks in pressure sensing lines (see Chap. 9), to measure vibration of reactor internals using neutron sensors, and to detect core flow anomalies by cross-correlating noise signals from neutron detectors and core exit thermocouples.[17]

To test the response time of temperature sensors, the LCSR method is the most commonly used technique because it can yield a sensor's response time so accurately. If accuracy is not critical (e.g., if a large margin exists between the sensor's nominal response time and the plant's required response time), then noise analysis may be used. Alternatively, noise analysis can be used to identify if a sensor's response time has changed. If the sensor's response time has changed significantly, then the exact response time can be measured using the LCSR method. The advantage of the noise analysis technique is that it does not require that the sensor be removed from service, and many sensors can be tested simultaneously using a single multichannel noise data acquisition system.