KANSAS CITY, Mo. — Can air terminal units be quiet? Yes, if proper attention is given to their placement and installation. This was the opinion of four speakers at “Controlling Noise From Terminal Boxes,” a seminar given at the 2003 Annual Meeting of the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE).

Don’t pay proper attention, however, and you risk significant noise problems. Once considered a minor point of air system noise, research points to terminal units as being a major source of noise radiation into the room.

Quiet, Please

Ted Carnes, Ph.D., P.E., is an ASHRAE associate with Pelton Marsh Kinsella (Dallas), a professional consultancy in acoustics and related fields. Carnes also used to work for a terminal manufacturer in the mid 70s. “Air terminal units can be quiet,” he stated in his presentation of the same name.

In a system overview, he explained that there are two types: constant and variable air volume (VAV).

The typical air terminal unit starts with a basic, single-duct design, with a primary air valve, plenum inlet, fan motor, and unit discharge. If heating is needed, Carnes said, the unit can use either supplemental heat or the primary heat source. Units for parallel-flow configurations are yet another variation, as are modulating diffuser designs. There are a variety of terminal configurations (single-duct, induction, bypass, and dual-duct) and damper designs (integral diffuser with inlet damper and with outlet damper).

Noise is a factor when the terminal unit is sitting above the ceiling, Carnes explained. Duct noise travels down to the diffuser.

Different sound paths can be a concern, he said; these are detailed in ARI Standard 885, “Procedure For Estimating Occupied Space Sound Levels In The Application Of Air Terminals And Air Outlets.” Of course, you can also have sound breaking out of ductwork.

Discharger sound is the biggest concern, he said, in fan-powered systems. Motors and the type of speed control are also concerns, Carnes said. “Variable-speed control is good for balancing, but it adds to the sound,” he said, particularly high-frequency noise.

“Terminal units radiate sound into the room,” he explained. “They act as a large radiating source, not a small local source.” In addition, “Ceiling height, room sound levels, people, and time of day are all big factors,” when it comes to room noise. (See Figure 1.)

Air goes through the damper that is modulating in a basic unit, he explained. Sound is critical because the dampers are generally placed over people for their thermal comfort. However, Carnes said the associated noise results in people saying things like, “Why did you put that noisy thing over where I sit?”

Figure 1. Terminal units can radiate sound from multiple sources. (From ARI Standard 885-1998.)

Keeping Sound Down

There are many standards related to sound. Carnes ticked off those from ANSI, ARI (880 and 885), and ASHRAE (a testing standard that is being revised). “Most everyone is using Standard 885 — or should be,” he said.

“Be very careful about using NC [noise criteria] data out of catalogs,” Carnes warned. This data “is representative of the device only, and does not include any external system effects. You may wind up needing to buy silencers.”

Installers need some straight ductwork at the inlet, Carnes said, to help keep static pressure down. Velocities at the outlet should be kept down, although “You will need some downstream static pressure.”

The terminal unit itself re-quires proper support and considerations of height with and without the ceiling. Don’t pull the unit up tight against the structure during installation, Carnes said. “This can cause severe rumble.” In general, stay below 2 inches of static pressure. (He noted that 1.5 inches wc [water column] is “a suggested upper limit.”) He suggested designing sound levels for the highest load condition.

Basic considerations for radiated noise troubleshooting are the ceiling/plenum, leakage at the tile and grid, and return air paths. Improvements can include a sound attenuation blanket, sheetrock, and return air grilles with sound boots.

Those examining noise from an acoustical room effect should look at the room itself. The ceiling height and overall volume of the room should be taken into account, as well as whether its surfaces are hard and reverberant or soft and absorptive. The activity levels/uses of the room also need to be considered.

Finally, “Do not rely on terminals for speech masking. This is sometimes done with disastrous results.”

Parallel Vs. Series

Gus Faris is an ASHRAE associate with Nailor Industries Inc. (Houston). He offered a “Comparison of Room Noise Control For Series And Parallel VAV Units At Different Operating Conditions.”

The first units were parallel “with a modified control sequence to prevent stale air,” Faris explained. Because discharge sound from the damper is affected by unit discharge, the unit’s operation was based on minimum airflow requirements, controlled by a pressure-dependent system. Because of this new control sequence, “There is no more condition that is ‘fan only,’” Faris said.

In the series units, primary air was provided upstream of the fan. The fan had an effect on radiated NC levels, Faris said. So did primary discharge sound. The primary radiated sound affects radiated NC.

Testing of both systems examined pressure drop. The researchers took the reverberated numbers and converted them to NC.

The results: “Radiated sound set the room NC in all the applications,” Faris reported. There was not much difference between parallel and series systems’ NC levels, though the parallel system suffered more from peaks.

Actual Conditions, Troubleshooting

Dan Int-Hout III, Member ASHRAE, is with Krueger (Richardson, Texas). He covered “Specifying Terminal Boxes.” Among his basic concepts, he pointed out, “Actual conditions will always be louder than reverb room tests indicate.”

Int-Hout said that in his job, he often gets to decipher specs that other people can’t figure out. For instance, specifications often state that the system “Shall provide an NC 35.” “How do they qualify that?” Int-Hout wondered. “It’s like some magic number.”

On the other hand, “Operating conditions [design inlet static pressures] are seldom spelled out. This allows vendors to pick whatever assumptions favor their equipment.

“Requiring ‘sound power less 10 dB room effect’ will result in grossly oversized boxes and un-necessary (and probably ineffective) duct silencers,” he stated.

Ductwork plays a considerable role in room sound levels. For instance, duct lining is probably required if room sound levels of less than room criteria (RC) 35 are required. Flex duct can be “a wonderful attenuator,” Int-Hout added. However it can be, and often is, misapplied.

Int-Hout abused designers for what he perceives as a lack of knowledge about current design standards. “How many of you have seen ARI 885?” he asked.

As far as diffuser specs go, many that pass his desk “have been assuming 10-dB room effect. Most rooms are 8 or 9 dB, but 10 isn’t bad,” he said.

Radiated sound is more complicated. “Standard 885-90 presents attenuation for common materials,” Int-Hout said. “Acousticians tend to stand under the box, where it’s somewhat higher.” Current NC data has been changed. Without going into the specifics, he stated, “These new numbers are pretty good.”

If you want to test your room absorption and sizing figures, do the math — then run it backwards, Int-Hout advised. “Then you’re not making assumptions.”

Specifying an NC with specific path-attenuation elements, based on acceptable application data, will result in applicable data, he stated. After construction, problems are difficult and expensive to fix.

Field Testing And Construction Details

Gaylon Richardson, Fellow ASHRAE, of Engineered Air Balance Co. (Houston), looked at construction details and other sound considerations in “Field Testing Concerns Of Terminal Units.”

It may seem to be a given, but are the installed terminal units the same size that was specified? Check other construction details, Richardson advised, such as materials. Check the branches and return ducts. If you’re troubleshooting a system for noise problems and sound levels are greater than desired, “Check the construction design and operating conditions,” Richardson said.

He listed some typical problems and some possible solutions.

  • Actual operating conditions not as designed. Confirm that the system is operating at or near the airflow and pressure drops used in the estimation process. This is often a large source of error. Verify that the static pressure control and controls that regulate flow are functioning properly. Make installation adjustments as needed.

  • Fan noise in a fan-powered mixing terminal. Reduce the fan speed if possible or reselect the terminals for critical areas.

  • Valve noise. Reduce the inlet pressure, if possible. Otherwise, replace the terminal with a lower pressure drop terminal and then reduce the inlet pressure.

  • Flexible duct breakout. Re-place metal duct or lag the flex duct.

  • Diffuser noise. Check the diffuser inlet to make sure that the damper is not almost fully closed and that there is an acceptable duct connection (flexible duct not crimped, etc.). Verify whether the diffuser noise is self-generated. An easy check is to remove the diffuser core. If the diffuser sound is self-generated, consider adding additional diffusers to achieve a lower airflow per diffuser or re-select the diffuser. If the noise is duct noise and is not generated by the diffuser, add internally lined duct attenuation upstream of the diffuser. Exterior lining provides little acoustical benefit.


    Air leakage may result in airflows different than designed, resulting in higher than expected sound levels and pressures. Check and seal leaks.

    If the air distribution system noise source cannot be significantly reduced or relocated, then it is necessary to use path attenuation to achieve desired acoustic goals. For air terminals or other sources above the ceiling tile (not diffusers), the following path attenuation modifications may be considered:

  • Increase the absorption of the plenum cavity in the immediate area near the VAV terminal.

  • Relocate return air ducts, grilles, etc.

  • Select a higher insertion loss ceiling tile system.

  • Use an absorptive ceiling barrier under the noise source to provide some absorption and prevent direct radiation of terminal noise to the ceiling tile.

  • Straighten flexible duct sections and eliminate unnecessary bends and sagging.

    Publication date: 07/28/2003