Objective With the development of compressed air energy storage (CAES) technology, compressed air transmission pipelines are expected to operate under higher pressures and larger flow rates, making the issue of aerodynamic noise increasingly prominent. During the design phase of a 300 MW CAES power station demonstration project, the first-stage compressor exhaust pipeline was identified as having a risk of high flow-induced noise. Therefore, it is necessary to conduct a predictive study to quantify the noise level and explore potential noise reduction measures.

Method The computational fluid dynamics (CFD) method was employed, using large eddy simulation (LES) to simulate the broadband pressure fluctuations induced by turbulence. A mathematical model for analyzing flow-induced noise in pipelines was established to investigate the internal flow characteristics and noise generation mechanisms. The noise spectrum distribution during pipeline operation was determined, and noise mitigation strategies were proposed based on local structural modifications.

Result The flow field distribution of the exhaust transmission pipeline of the first-stage compressor was obtained, and it was found that there was a large flow field disturbance in the section where the flow channel changed drastically, such as the elbow, reducer, and tee, which became a potential noise source. The spectral density of noise and pipe wall pressure based on frequency domain was obtained, and it was determined that the fluid pressure ripple caused by the change of pipe structure mainly existed in the low-frequency region below 10 Hz, and the pressure pulsation energy decreased exponentially with the increase of frequency.

Conclusion Optimization analysis showed that adding beveled chamfers to the tee connections of the pipeline can effectively reduce noise. The noise reduction effect improves with increasing chamfer size, with the maximum sound pressure level reduced by approximately 4.99%.