The development of a simple, reliable and low-cost on-line particle detection method for liquid flows has been recognized as a global challenge in the chemical engineering industry. To overcome the existing limitations in the detection of sand particles in water-gas pipe flows, a non-invasive acoustic method with two special acoustic devices (installed separately on a straight pipe and a bend pipe) for the detection of miniscule sand particles was evaluated. A series of detection methods, namely, time-frequency analysis, mutual-coherence analysis, and multi-band elimination filter methods, were developed to extract weak sand features submerged in the background noise of strong flows. Verification experiments were conducted to record acoustic signals from sand-carrying flows, thereby revealing five sand characteristic frequency bands in a water-gas bubbly flow: 20.59-21.31 kHz, 24.10-24.82 kHz, 27.64-28.36 kHz, 31.10-31.84 kHz, and 34.61-35.33 kHz. Additionally, good agreement was found between the sand concentration (0.06-0.24 wt.%) with a particle size of 150 mu m and the acoustic sand root mean square (RMS) level within the discovered sand frequency multi bands, verifying the use of the proposed acoustic detection devices and the corresponding weak sand identification method. This study provides the foundation for the future detection of solids in more complex multiphase flows. (C) 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.