Flow noise identification using acoustic emission (AE) energy decomposition for sand monitoring in flow pipeline

Abstract In pipelines used for petroleum production and transportation, sand particles may be present in the multi-phase flow of oil and gas and water. The Acoustic Emission (AE) measurement technique is used in the field of sand monitoring and detection in the oil and gas industry. However, as the AE signals recorded are strongly influenced by flow conditions in the pipe, identification of sand particle related signals or events remain a significant challenge in interpretation of AE signals. Therefore, a systematic investigation of sand particle impact AE energy measurements, using a sensor mounted on the outer surface of a sharp bend in a carbon steel pipe, was carried out in the laboratory to characterise flow signals using a slurry impingement flow loop test rig. A range of silica sand particles fractions of mean particle size (212–710 μm) were used in the flow with particle nominal concentration between (1 and 5 wt.%) while the free stream velocity was changed between (4.2 and 14 ms−1). A signal processing technique was developed in which the total AE energy associated with particle-free water impingement was divided into static and oscillated parts and a demodulated frequency analysis was carried out on the oscillated part to identify major spectral components and hence the sources of AE signals. A simple theoretical model for water impingement AE signals was then developed to show the dependence of AE energy components on different flow speeds. A similar decomposition of AE energy into static and oscillatory components was used to analyse AE signals for particle-laden flows. The effect of flow speed on the spectral AE energy for different sand concentrations and particle size fractions was investigated and the results show that the 100 Hz band is attributed to mechanical noise, the 42 Hz band is due to fluid turbulence and the dominant band is broad oscillated component. The AE energy decomposition method together with the water impingement model and coupled with spectral peaks filtering enable isolation of AE energy associated with particle impact from other AE sources and noise and, hence, the proposed decomposition approach can enhance the interpretation of AE data in pipeline flows.