Design of ultrasonic flowmeter for wet gas measurement

Abstract When an ultrasonic flowmeter (UFM) is applied in the measurement of wet gas, the ultrasonic signal attenuates and fluctuates strongly because of the liquid film and droplets in gas. It is hard to obtain the accurate time of flight (TOF) of ultrasound. Besides, the gas velocity distribution on a pipe cross section changes with different flow patterns, making it difficult to determine the measurement model. In order to solve these problems, a dual-channel UFM is specially designed and a measurement system is built. Then, a dynamic threshold method based on statistical average signal amplitude is proposed to obtain the TOF of ultrasound in wet gas. Combined with the classical void fraction model, the wet gas velocity measurement model for the dual-channel UFM is established based on a neural network. Finally, experiments are carried out on a DN50 wet gas device, where pressure is 0.1 MPa, gas superficial velocity is 5–20 m s −1 and liquid volume fraction is 0.2%–5%. The results show that the average relative error of gas velocity is 0.52% and the maximum relative uncertainty is 0.491%.