Quiet that compressor

The general intent of noise mitigation is to neutralize its objectionable characteristics by modifying tonal, temporal and loud noise into a continuous broadband noise at a lower sound level. Identifying the types of noise generation and paths of travel leads to uncomplicated, feasible, economical ways to attenuate and isolate both airborne- and structure-borne sound.

Excessively loud air compressor systems in the plant or operations area are associated with hearing protection, difficult workplace communications, fatigue and annoyance. Central plant or building services noise can intrude into conference rooms, offices and laboratory spaces, resulting in speech interference or annoyance.

The complaints and their sources make it apparent that noise is more complicated than mere loudness. Sound quality, or tonal and temporal noise characteristics, affects perception and annoyance. But you can solve facility noise issues by considering the source, the path and the receiver.

Determine where the noise originates and how it radiates. The source may be machine rotation, reciprocating vibration or impact contact. Turbulence and gas escaping a nozzle or other opening also produces noise. If it’s unclear which of several potential noise sources is at fault (Figure 1), turn off each one sequentially until the true source is obvious.

Then characterize the noise as tonal (low-, mid- or high-frequency) or as broadband (sound distributed evenly over the audible spectrum). Figure 2 illustrates different noise spectra. The blue spectrum is a whine, peaking at 1,000 Hz. The red spectrum, heavily unbalanced on the low frequency side, would be perceived as rumbly. The purple spectrum has a tonal characteristic with sound level in the 500 Hz band that is much louder than the side bands. The green spectrum is a relatively balanced or flat spectrum, relative to the noise criteria lines. Of the four shown, this would be least objectionable.

Next, determine how the noise differs from the continuous background sound in the area. Identify tonal components, such as buzz, hum, hiss or rumble by subjective listening or by measurement. Also, identify temporal or time-varying nature, such as on and off.

Determine whether the sound travels through the air, the floor and walls, or both. Determine whether the airborne- or structure-borne path is stronger, or whether noise travels over both paths. This requires either careful listening or sound and vibration measurements along the suspected route. Sometimes structure-borne sound can be detected by putting an ear against wall surfaces and comparing that noise with the general room noise. As noted above, turning suspect sources on and off may confirm the offender. Along the path, consider conditions that affect airborne sound:

• Amplification of airborne sound by reflection.
• Attenuation of airborne noise by absorption.
• Reduction by transmission through solid partitions and other barriers.

Look for holes in what may appear to be solid walls. Look for gaps in mortar between bricks or masonry units. Drywall partitions might not be caulked at floor and penetrations. Find holes and penetrations for ducts, piping and conduit. Look for partitions that terminate at a suspended ceiling. Locate and evaluate vertical shafts and pipe chases. Determine whether a discontinuity could be introduced into a structure-borne path to either isolate or break the vibration path or damp vibrating surfaces (ringing or free vibration).

Potential solutions might include adding absorptive surfaces to reduce reflective or reverberant noise buildup along the path. Improve partition barriers by increasing their mass and stiffness while closing gaps, penetrations and other holes. Often a combination of measures at the source and along the path mitigate noise disturbance most effectively.

Absorption, though, is often misapplied. It helps to reduce reverberant and reflective sound, particularly near the source and along the path, but it may be of little benefit in the room where the noise is objectionable. The reason is that absorptive materials reduce the sound level of everything in the room. If the intrusive noise is 3 dB louder than the background before absorption is added, it will probably remain about 3 dB above the background afterward.

Noise is perceived relative to other environmental sounds. If the ambient level is quiet, relatively modest noise can be annoying. If the background is loud but broadband, tonal or temporal components of noise can still be perceptible and annoying. Evaluate the severity, perceptibility and annoyance of the noise by comparing the noise (with equipment operating) to the background sound at the receiver location (with equipment off).

First, evaluate or categorize the continuous background or ambient conditions at the receiver in terms of loudness and spectrum (preferably without the offending or disturbance noise). Listen to the background noise and describe it subjectively as loud or quiet, and with descriptive adjectives, such as roar, whine, buzz, hum, rumble or whistle.

Take note of how far you can be from a person speaking in a normal conversational voice and still understand most of what is said. If you can conduct measurements, the A-weighted levels from a simple hand-held meter provide a start. With a two-band meter, compare A- and C-weighted measurements to evaluate the noise spectrum. If the difference between A and C levels is 10 dB or more, you probably have an unbalanced spectrum with a lot of low-frequency rumble. Using a spectrum analyzer to determine octave band levels makes tonality much easier to identify.