Need: Many natural gas compressor stations which were previously located away from residential areas are now being encroached upon by surrounding building developments. Furthermore, an increased awareness of community noise issues has proved to be the impetus for investigating and developing more effective noise control methods and treatments for natural gas compressor facilities. This project investigates the feasibility of applying Active Noise Cancellation (ANC) to the exhaust of a large, internal-combustion reciprocating type engine. Large reciprocating internal combustion engines pose significant challenges for the noise control engineer. In the case of the engines employed at Tennessee Gas Pipeline Company Compressor Station 229, these engines radiate extremely low frequency exhaust noise into the surrounding environs. These engines produce discrete frequencies (i.e. “tonals”) in the exhaust spectra with a particularly strong component at 26.5 Hz, which corresponds to the fundamental firing frequency (the 5.0 rotational order) of the engine; significant attenuation of the raw exhaust noise can be particularly difficult due to the sound power and spectral content. Traditional methods would necessitate a very large silencer in order to realize improved attenuation of the exhaust noise, relative to the existing silencer.
Benefit: Measurements were conducted at the error microphone location, at 1.0 meter from the exhaust outlet and at the property line. At a distance of 1.0 meter the WNCT integrated active / passive silencer yielded 84.5 dBA (92.3 dBL) while the original equipment silencer yielded 92.7 dBA (98.8 dBL). Band-limited (DC - 200 Hz) measurements were taken at the error microphone location; control off (WNCT passive - only): 109.8 dBL overall, 107.7 dBL 26.5 Hz component. With control on (WNCT active + passive) at the same position overall noise was 99.7 dBL with the 26.5 Hz component reading 89.1 dBL. Far-field A-weighted reductions were inconclusive due to the presence of other contributing noise sources possessing similar noise characteristics. Flow resistance measurements indicated that back pressure had been reduced by 95% relative to the original equipment silencer through the use of the integrated WNCT active / passive silencer.
Result: While the application of the ANC silencer yielded substantial improvements in overall noise near the source, overall reductions in the far field will be realized only if all sources possessing similar sound power and content are likewise treated. A passive silencer yielding the equivalent insertion loss (as the active silencer) would have to be substantially larger than those in current use and may not be cost-effective, nor easily manufactured.
