Ultrasonic cross-correlation flow measurement: Theory, noise contamination mechanisms, and a noise mitigation technique

Based on past experience with ultrasonic cross-correlation flow meters in power plant environments, the presence of spatially correlated noise due to pressure waves, vibration, or sources other than transport of turbulent eddies will cause a bias in the time delays measured by the meter. Several techniques were developed to detect the existence of such correlated noise and correct for its effect at plant conditions. An analytical and experimental investigation was performed to further understand the basic physics of the noise mechanisms. The dominant error mechanisms investigated in this work were speed of sound perturbations due to pressure fluctuations and beam path length changes due to wall vibration. An analytical model was formulated which estimates the signal level of the flow meter based on the turbulent velocity field. From this model, an estimate of the system noise which would cause contamination could be determined. A test at a water tunnel facility was performed in order to evaluate the noise mechanisms. During this test, measurements were taken with and without controlled noise sources. Pressure and acceleration measurements were used to evaluate a coherent noise removal technique developed to mitigate the impact of noise in the ultrasonic cross-correlation flow measurement. The coherent noise removal technique was shown to be effective in removing noise during the water tunnel test.