Controlling mechanical system noise

by Katie Daniel | December 10, 2015 11:03 am

[1]

by Aaron Bétit
A typical urban and suburban environment has numerous sources contributing to the exterior ambient noise. Among these are the environmental sounds from a building’s heating, ventilating, and air-conditioning equipment. How can design professionals help mitigate the distraction caused by HVAC?

In addition to specific air exchanges and heating/cooling requirements, mechanical systems must comply with the local noise ordinances and operate without disturbing adjacent properties. The proliferation of mixed-use developments can increase the concerns of mechanical system noise due to the blend of commercial, cultural, residential, and industrial uses in directly adjacent properties. For development to succeed, it is important to design buildings that control noise appropriately.

Environmental noise
Noise is unwanted or objectionable sound as translated from minute air pressure fluctuations by our ears. The fundamental measure of sound amplitude is the decibel, which is abbreviated to dB; as people can hear a large change in sound pressure, the decibel was created as a logarithmic metric.

Human reaction to a sound varies depending on the frequency content, the duration, the amplitude of other noise sources, and the time the noise level
is generated. Some examples:

• 
frequency content—the noise generated by a window fan can be significantly less distracting compared to the sound of a garbage truck backing up, even when compared at the same amplitude;
• 
duration—a short report of a warning siren once
a day at noon is considerably less disturbing than the sound of a neighbor practicing on a drum kit in an adjacent garage for hours on end; and
• 
time of day—if an air-cooled chiller is operating at a consistent amplitude directly outside an apartment façade, it will typically be less distracting in daytime hours when the receptor is active than at night, when a receptor is attempting to sleep. The noise may seem louder at night than during the day.

The ear does not respond equally to high- and low-pitch noises. Starting in the 1930s, scientists developed response characteristics to represent the sensitivity of a typical ear. Adjustments to amplitude based on frequency called ‘A-weighting’ were developed to allow for a metric that more closely follows the human reaction to noise. A-weighted decibels—dBA—are modeled after the sensitivity of the ear at sound levels commonly found in the environment. In most cases, the A-weighting decibel is the standard metric used to evaluate exterior noise.

The ambient noise level’s duration is another important aspect of human response to noise. Numerous metrics have been developed to help determine the impact of ambient noise levels over a period. In the most basic form of evaluation, sound pressure is averaged over a specific length of time. This is referred to as the Equivalent Sound Level (Leq).

While the Leq metric is appropriate for evaluating continuous noise sources, there are times when the noise of concern is not continuous—some mechanical equipment has a cyclical noise, or only occurs in short durations. To evaluate these conditions, the Maximum Sound Level (Lmax) can be used to document the highest sound pressure level from an activity. However, it is important to note Lmax can be compromised by anomalies in the data or by extraneous sound sources. In part for that reason, percentile noise metrics (Ln, where n is the percentile of interest) are often used to characterize environmental noise.

The 90^th percentile noise level (L90) is defined as the noise level exceeded 90 percent of the time during the measurement period; it is a reasonable proxy for ‘ambient’ levels used in some local and state regulations. The 10^th percentile noise level (L10)
is sometimes used to characterize transient noise. Similarly, the median noise level (L50) is used in lieu of the mean or Equivalent Noise Level (Leq)
by some local ordinances.

Twenty-four-hour metrics document an ambient noise environment over a typical day and are often used when evaluating community noise. The Day-Night Sound Level (Ldn) is measured in dBA and describes the receptor’s cumulative noise exposure from all noise events during a 24-hour period. The events between 10 p.m. and 7 a.m. are increased by 10 dB to account for greater nighttime sensitivity to noise.

The State of California developed a 24-hour metric, the Community Noise Equivalent Level (CNEL), to document the ambient noise levels during a typical day. Rather than dividing the day into two periods like the Ldn metric, the CNEL metric includes a third period to address the time when people are likely to be engaged in outdoor activities around the home. Between the period of 7 p.m. and 10 p.m., the measured noise levels are increased by +5 dB to reflect additional annoyance noise causes during this time. In most cases, the difference between Ldn and CNEL is slight, and the two measures are fairly interchangeable.

State and local noise limits for intrusion across a property line have been established for most developed areas in the United States. Guidelines have also been published to help evaluate the appropriate use of specific land areas based on the ambient noise environment. In the absence of local noise limits, the land use table in Figure 1 can be a good resource for evaluating impact to adjacent properties.

[2]

[3]

Published by the State of California Department of Health Services’ Environmental Health Division, the table provides guidelines for land use compatibility, dividing into four categories: normally acceptable, conditionally acceptable, normally unacceptable, and clearly unacceptable.

Normally acceptable
Specified land use is satisfactory, based on the assumption any buildings involved are of normal conventional construction without any special noise insulation requirements.

Conditionally acceptable
New construction or development should be undertaken only after an extremely detailed analysis of the noise reduction requirements is made and needed noise insulation features have been included in the design. Conventional construction with closed windows and fresh-air-supply systems or air-conditioning normally suffices.

Normally unacceptable
New construction or development should be discouraged. If new construction or development does proceed, a detailed analysis of the noise reduction requirements must then be made and any needed noise insulation features included in the design.

Clearly unacceptable
New construction or development should generally not be undertaken.

Acoustical mitigation for specific mechanical equipment
Just as in real estate, the most important factor to consider in mitigating HVAC noise is location. Prior to construction, noisier building systems components should be located away from more noise-sensitive properties. Ideally, noisy components like chillers and pumps are located inside mechanical rooms or in a central plant. At a minimum, portions of the structure would be used to block the line of sight to noise-sensitive receptors. Using a central plant, rather than packaged units, allows the noisier components to be inside, helping contain the sound and reduce the impact to 
the community.

With a central plant system, cooling towers are often the major offender 
to the community noise level. Sound generated by a cooling tower comes from the fan, the motor, cascading water, and the gear boxes.

[4]

Two common cooling tower types are forced draft and induced draft cooling towers. Forced draft towers usually have centrifugal blower fans on the side of the air stream entering the tower. These towers are usually a little louder compared to induced draft cooling towers. However, they are more contained with fewer components directly exposed to the exterior, and some can be operated indoors. Since their fans are capable of additional static pressure, forced draft towers are able to include ductwork, and sound attenuators can be applied on the inlet and outlet if necessary for noise control.

Induced draft towers have the fan on the exit of the air stream, pulling the air through the tower. Their fans generally cannot handle the additional static pressure that sound attenuators cause. Consequently, beyond selecting cooling towers with efficient and quiet fans, the typical mitigation options for induced draft cooling towers are location, oversizing the tower so the fans run at a slow speed, and sound barrier walls. To limit the noise generated by cooling towers, the project specification should include sound limits for the specific tower based on measurements conducted using Cooling Technology Institute (CTI) 128, Acceptance Test Code for Measurement of Sound from Water-cooling Towers.

Mechanical chillers—which can
be another important source of environmental noise—are classified by their compressor type. These include reciprocating, rotary screw, frictionless centrifugal, and absorption.

A reciprocating chiller is similar to an automotive engine. An electric motor turns a crankshaft that causes pistons to compress a gas. As the piston compresses the gas, it heats up and is exhausted into the condenser. Capacity in a reciprocating chiller is typically controlled by keeping specific exhaust and inlet pistons open; consequently, the electrical motor turns at the same rate regardless of the required capacity.

Rotary screw chillers use two mating helically-grooved rotors. Gas is compressed by rotating the screws. Capacity of a rotary screw chiller is controlled by variable-speed drives. While all chillers have a tonal aspect to the noise, the tonal characteristic of screw chillers is significantly more pronounced.

[5]

Less common in the United States, the frictionless centrifugal compressor is similar to a centrifugal water pump or fan—the impellor spins to compress the refrigerant. They use magnetic bearing technology for the shaft. Thanks to the magnetic bearing, these chillers can be significantly less noisy compared to the other chiller types. They use a variable-speed direct-current (DC) motor with a direct drive. Absorption chillers, on the other hand, do not use mechanical compression; instead they use a heat source such as steam or natural gas.

Most chillers are capable of being enclosed and consequently the best method of mitigating chiller noise is enclosing them inside the building with high mass walls such as grouted concrete masonry units (CMUs) and acoustical door assemblies. The noise generated by chillers should be limited using specifications referencing sound measurements that are conducted in accordance with Air-conditioning, Heating, and Refrigeration Institute (AHRI) 1280, Sound Power Rating of Water-cooled Chillers.

Air-cooled chillers contain both the chiller and an induced draft cooling tower in a single system. It is sometimes best to avoid using this style of system when directly adjacent to noise-sensitive receptors because the mitigation options are limited. The tonal characteristic of a chiller can be prevalent, but because this system includes a cooling tower, it is not possible to enclose these chillers.

Some air-cooled chillers can include an acoustical package that wraps components of the chiller in mass-loaded vinyl. While these acoustical packages help reduce the noise, the reduction is modest. As this equipment needs to operate outdoors, the noise mitigation options are limited to careful space planning (location) and the design of acoustical barrier walls.

There are more options for acoustical mitigation for other rooftop equipment like exhaust fans and air-handling units (AHUs). In extremely sensitive cases, this equipment can be located inside the structure. If the fan is selected to be able to operate under higher static pressures, air inlet and outlets can include sound attenuators.

There are two types of exhaust fans that due to code typically cannot use sound attenuators. Kitchen exhaust and laboratory exhaust have particles in the exhaust air that can cause corrosion or collect in the sound trap—consequently, sound attenuators often cannot be used. In the case of kitchen exhaust, the mitigation options are very limited. Location of
a kitchen exhaust, careful selection of a quieter exhaust fan, and providing acoustical barriers between the exhaust opening and noise-sensitive receptors are the main options when addressing kitchen exhaust noise.

In addition to exhausting caustic particles, laboratory exhaust fans require fast velocities at the exhaust opening to expel the exhaust into the atmosphere, diluting the caustic particles. Increased exit velocities cause additional noise, beyond what is generated by the exhaust fan. Some high-plume exhaust openings are acoustically more efficient than others, but as with kitchen exhaust these fans should be located further away from noise-sensitive receptors and, in some cases, require acoustical barriers to provide additional attenuation.

Typical acoustical mitigation
Beyond locating noisy equipment as far from noise-sensitive abutters as possible, control of mechanical noise in the environment is typically achieved by enclosing the equipment, designing an acoustical barrier, or including sound attenuators or splitters at the inlet or outlet of the system.

Fully enclosing the mechanical equipment is the most effective option, provided it is indeed an option. At times, it may not be possible to enclose the equipment entirely, because some equipment needs access to outside air in order to function properly. In these cases, where air inlet and outlets are necessary, a sound attenuator, acoustical louver, or splitter can be used.

Sound attenuators are essentially ducts constructed of heavy-gauge sheet metal and perforated sheet metal baffles typically filled with an acoustically absorptive material such as mineral wool or fiberglass. In the cases where the air is caustic or needs to have higher levels of purity, the acoustical material in the baffles can be enclosed in Mylar bags, or removed from the sound attenuator completely (i.e. ‘packless attenuators’). Without an acoustical fill, the sound reduction provided by the attenuator is limited.

Acoustical louvers are similar to sound attenuators in that they have perforated sheet metal baffles with an acoustically absorptive fill, but are much shorter in length. When more attenuation is required, 
a louver can be replaced with a sound attenuator
(as space allows), or filled with acoustical ‘splitters,’ which are perforated sheet metal baffles without the encapsulating ductwork.

Acoustical barriers and walls provide attenuation by redirecting the noise. Walls are most effective when they are located either close to the source or close to the receptor. In the case of addressing community noise, it is generally better to build the wall near the source, since there are usually a number of receptors at various distances. That said, when designing the barrier, sufficient clearance must be maintained between the wall and the equipment to allow for proper airflow and for maintenance access.

[6]

Acoustical walls must be continuous to be effective; gaps at the bottom of the wall or along the wall significantly reduce the amount of attenuation the wall provides. Since the acoustical attenuation achievable with a barrier is limited by sound that bends over the top of the wall, the surface density of an acoustical wall can be as light as 10 to 20 kg/m² (2 to 4 lb/sf)—depending on its height—without sacrificing the acoustical benefit. Often, code requirements for wind and seismic loads require the wall to be more substantial than is required acoustically.

If there is an acoustically hard surface opposite the acoustical wall, or if the hard surface wall surrounds the equipment, there is the potential of creating a ‘canyon effect.’ The wall refocuses the sound energy toward the opposite structure, which, in turn, refocuses the sound energy back toward the acoustical wall. Since the sound energy has no place to escape, it bounces between the two structures until it reaches the top of the wall, allowing the same amount of acoustical energy to escape—in other words, the wall’s benefit is eliminated.

To avoid this problem, it is important to use an acoustically absorptive surface on the wall facing the mechanical equipment whenever there is another structure opposite. The surface should have a minimum Noise Reduction Coefficient (NRC) of 0.90, meaning 90 percent of the sound energy exposed to the surface is absorbed by the finish, and only 10 percent is reflected back off the wall. For most materials, this absorption specification requires a thickness of at least 50 mm (2 in.).

Careful planning and evaluation of the project site during design is the best method of ensuring the mechanical equipment does not adversely impact the community noise levels. By selecting the appropriate mechanical system and locating noisier mechanical components away from noise-sensitive receptors, the potentially significant cost of noise mitigation can be avoided. Specifying noise limits for each major mechanical piece of equipment can ensure quality equipment is installed on the project that will not require noise remediation after construction.

Aaron Bétit is a senior consultant in architectural acoustics and mechanical systems at Acentech, a multi-disciplinary acoustics, audiovisual systems, IT and security design, and vibration consulting firm. Based out of its Los Angeles office, his acoustical experience includes consulting during ground-up design, as well as providing diagnostic advice for existing facilities and structures with acoustical complications. Bétit’s environmental noise expertise encompasses acoustical design and computer modeling of power plants, roadways, warning sirens, and water treatment plants. His acoustical consulting services have been used on a number of landmark buildings in the Southern California community including healthcare facilities and laboratories, along with commercial, residential, and mixed-use towers. Bétit can be reached at abetit@acentech.com[7].

Source URL: https://www.constructionspecifier.com/controlling-mechanical-system-noise/

---