Ultra-Low Sound Air-Cooled Chillers

Air-cooled chillers provide significant advantages for many facility operators and owners. The elimination of a cooling tower greatly reduces the complexity of the system and significantly eases the maintenance of the system. For smaller facilities without the resources of large institutional owners, this reduced maintenance can be critical.

Therefore it is very common to see air-cooled chillers on smaller facilities such as public schools and small office buildings. However, this suitability does carry with it some costs--Air-cooled chillers are usually significantly less efficient than water-cooled chillers and, now with the advent of super-low sound cooling towers, often much noisier. The noise problem can be very significant with this sort of equipment in that a large proportion of the facilities that utilize this technology are located near or in residential areas.

In an earlier post I discussed how the Smardt air-cooled Turbocor chillers significantly change the balance between water-cooled and air-cooled chillers with respect to efficiency and sound. However, it is worth investigating the sound issue in more depth.

Recently, manufacturers have spent some effort in addressing the sound issue on their air-cooled chillers. This has generally been approached by providing low sound fans and addressing the compressor noise. Two products that are currently being marketed as low-sound chillers approach compressor sound in different ways. One product utilizes screw compressors with a VSD to reduce compressor sound at low loads. Another utilizes scroll compressors with elaborate compressor sound enclosures. This approach yields published sound data which is excellent at all chiller load conditions. Until recently, this chiller has been considered by many engineers to be the quietest air-cooled chiller n the market.

The Smardt chiller, of course, uses the extremely low-sound Turbocor compressor, and variable-speed ultra-low sound fans. It is reasonable to believe that this combination would make for a very favorable comparisons with these other low-sound chiller options.

And this expectation is borne out by the data. Smardt air-cooled chillers compare extremely favorably against the variable speed screw chillers as this graph illustrates:


Two things should be noted: First, the comparison here is between a 177 ton screw and a 200 ton Smardt chiller, and second that even at 100% load, the Smardt chiller is far quieter than the screw at 25% load. The difference is even more significant in the lower octave bands that carry so well over distance.

Comparing the Smardt Chiller to the acoustically treated scroll chiller also yields an extremely favorable comparison:



This comparison of 120 ton chillers shows that while both approaches yield extremely low sound levels overall, the Smardt chiller beats the competitor in 5 of 8 octave bands. And the advantage for the competitor in two of the other bands is slight. It also shows the significant effect of A-weighting sound data. In this graph, the red line represents the published sound data from the manufacturer. Close reading of this data indicates that it is not bare sound power, but A-weighted sound power. This method of reporting sound data takes very significant credits into effect, especially in the lower octave bands:

Octave band center frequency (hz)	Weighting
31.5	-39
63	-26
125	-16
250	-9
500	-3
1k	0
2k	1
4k	1
8k	-1

The dark blue data show the raw, uncorrected sound data for this chiller.

If acoustics are a design consideration for your air-cooled chiller product, Smardt offers a solution that is unmatched in the industry.

Extra: Audio Comparison of compressor noise

85 dBA screw compressor
73 dBA Turbocor compressor

Both measured at 1.0 m away from compressor. Your speaker volume will affect the output, but the comparison should be clear if the volume is not adjusted between clips.

Extra extra:

This link has sound files that illustrate the amplitude of a decibel, to give perspective to the graphs above.

'Greening' Lab Design

Laboratory fume hoods are energy intensive. In order to provide safety for their operator, they need to ensure a constant face velocity of air at the sash--air that must first be conditioned to keep the space temperature acceptable for comfort, moved via mechanical means to the lab, and then exhausted out of the building.

A common comparison used to highlight the energy costs of these systems is to compare the energy impact of a single fume hood with that of a typical US household. On average a single lab fume hood uses as much energy as three typical US houses. And when you consider that a given facility may have many lab hoods in a single laboratory space, you can see how these energy impacts quickly add up.


In order to minimize the wasted energy associated with these laboratories, high-precision VAV lab controls have been developed to ensure operator safety, and to only provide the minimum amount of air necessary--And great savings have been realized by this sort of measure. But the energy efficiency of these systems can be improved even more.

Once the airflow has been taken down to a minimum, the energy associated with conditioning that air has been greatly reduced. But the energy associated with moving that air still can be reduced further. ASHRAE 90.1 states:

ASHRAE Standard 90.1 - 6.5.3.2.3:
“For systems with direct digital control of individual zone boxes reporting to the central control panel, static pressure setpoint shall be reset based on the zone requiring the most pressure; i.e., the setpoint is reset lower until one zone damper is nearly wide open.”

This calls for static pressure reset for VAV systems to minimize fan energy--ensuring that only the minimum amount of static pressure is provided to move the air. And this strategy is perfectly applicable to laboratory VAV systems as well as commercial air conditioning--as long as the system components are selected appropriately.

Tek-Air has published a white paper entitled Demand Based Static Pressure Reset Control for Laboratories That explores the energy benefits of this type of control scheme.


This paper analyzes system component selection, including control valves and sensors and illustrates the impact of these decisions on the overall energy use of the VAV system. In an analysis of a 50,000 cfm exhaust system, the reduced static from a pressure reset strategy can result in nearly $9,000 per year savings in fan energy (based on 0.75" savings, and $0.06/kwh electric costs).

These sorts of static pressure savings are easily attainable with a wise selection of air valve components. The commonly specified venturi-type valve has a minimum operating pressure that prevents these savings from being realized, and this added pressure drop often creates objectionable noise, which requires even more pressure drop for the system in the form of sound attenuators. This pressure reset strategy requires valves that can operate accurately and safely at low pressures.

The Tek-Air PRD valve provides unmatched pressure performance, and a quick examination of a cross section of the valve shows why:


Each blade of the damper is a smooth airfoil, greatly reducing turbulence and keeping the pressure and acoustic profile of the valve to a minimum.

If pneumatic air is not available, Tek-Air's new Accuvalve provides very similar performance with the convenience of electronic actuation. (And it won an innovation award at the 2008 AHR expo!)



A peek at the cross section of this valve shows how it attains these low pressure drops:


The airfoil shape of the valve assembly assures minimal pressure drop and sound generation for great efficiency in the fan system.

Energy savings cannot come at the cost of safety, and it is imperative that systems utilizing this method of pressure reset have sensors that can operate accurately and effectively in a wide range of pressure regimes. Tek-Air uses vortex shedding flow sensor technology to ensure the most accurate and linear control on the market.

Energy conservation is only going to become a bigger and bigger issue for designers of all building systems, and fume hood systems are a large opportunity for savings. It is important that designers and owners consider all the impacts of their design decisions and their system selections.

(Don't forget about checking the fan for stability: See this article for a review on this issue.)

Rethinking Air-Cooled Chillers

Air Cooled Advantages

Air cooled chillers offer many advantages to owners and designers. The first, and perhaps most compelling for many jobs is lower installed cost. Lower installed costs (compared to water cooled chillers) are driven by the following advantages:

• No Cooling Tower, Tower pumps, Tower and Pump Starters
• No equipment room required for the chillers
• Mounted starters

They also are easier to maintain, since the systems are significantly simpler than water-cooled systems:

• No on site Systems Engineer required
• No water treatment or make up water required
• No leaks on the roof
• No cooling tower, condenser pumps, associated starters

Generally, however, these advantages have come with significant trade-offs: Efficiency and Sound performance.

However, the introduction of Variable Speed oil-free air-cooled chillers by Smardt changes the balance.


First off, the Smardt Chiller is efficient. With IPLV's as low as 0.65 kw/ton, these chillers rival water-cooled system when the parasitic loads of the condenser pumps and cooling tower are considered. These chillers gain their efficiencies both from the inherent efficiency of the Turbocor compressor and the elimination of oil return issues that prevent other air-cooled chillers from capitalizing on the reduced head pressures available at low ambients.

This means these chillers use about 60-65% energy of other air-cooled chillers for the same load, and can nearly eliminate the energy benefit typically provided by moving to water-cooled systems. When you consider the cost of water (nearly $15/1000 gallons in Seattle, including sewer charges) this means the yearly cost of operation of these units is unrivaled. And energy conservation rebates are extremely attractive for these chillers.

The other major traditional trade off with using air-cooled equipment is sound. Screw chillers especially are known for their unfavorable sound characteristics. In most municipalities, sound ordinances are driven by occupancy and time of day. The most stringent criteria must be met during evening hours, typically when the units are not at their peak load. However, with constant-speed systems, the compressor is either on or off. This means it is either putting out its full sound or none at all. At full speed, such compressors can often exceed the evening sound criteria--even if they are on only momentarily. And the staging between on and off can be objectionable in its own right, regardless of sound level.

The Smardt chiller minimizes the problems with compressor sound in two ways. First, the variable speed drive allows the compressor to ramp slowly up and down to match the required output, eliminating the objectionable switching between compressors that constant-speed chillers exhibit. And secondly, they are just extremely quiet to begin with. Since no moving mechanical part is in contact with the chiller casing, very little mechanical noise is transmitted. Ninety-ton Turbocor compressors have been tested at 72 dBa at one meter, compared to screw compressors that can be as high as 80 dBa or higher in the same test. Five of these compressors operating together yield a sound level of 75 dBa at 10’.

More Benefits

But efficiency and sound are not the only benefits from using the Turbocor technology on air-cooled chillers. Other, less obvious ones exist.

Turbocor compressors have only one moving part, yielding un-matched reliability.

Reliability is enhanced by the elimination of oil in the refrigerant system. And the frictionless bearing requires almost no maintenance.

Since Turbocor compressors are variable speed driven, they provide an inherent soft-start on the compressor. Instead of kicking the motor up to full speed when power is applied to the system, the VSD slowly ramps the compressor up to the required speed for the load sensed by the system. This reduces stress on the already greatly simplified system to reduce wear and tear on the components.

But this soft start has another, very important advantage over standard air-cooled chiler systems--the use of the VSD eliminates inrush amperage. When an electrical motor is at rest, there is very little inductive resistance to current flow through the windings. As the motor starts to turn, this inductive resistance increases with the increase in RPM. What this means is when power is applied across the line (or even with a reduced voltage starter) to a stopped motor, there is a spike of electrical current far greater in amplitude than the design amp draw of the motor:

(example graph of inrush on a well pump motor)

This temporary increased amp draw heats the motor beyond where it is designed to operate for extended periods. This forces the chiller designer to provided anti-recycle timers to prevent rapid re-starts that could fatally overheat the motor. In practice, this usually means constant speed compressors cannot be started more often than every half-hour or so.

Additionally, this increased amp draw has effects that need to be addressed electrically. This becomes even more significant if the chillers are being served by emergency power. The emergency generators that serve the chiller must be sized to handle the inrush amperage. This can be a very costly addition, especially since the added amperage is only required for the first 30 second of operation or so.

Generators = $$$

Turbocor compressors on the Smardt air-cooled chillers eliminate inrush and provides a soft-start. This both heightens reliability and reduces electrical costs. For jobs where reliability is a primary concern, like data centers, this technology makes a lot of sense. First, it eliminates the need for increased generator sizing, it is an inherently more reliable compressor, and it frees the cooling system from reliance on a water utility service that could be disrupted.

Your Next Energy Conservation Measure May be a Quiet Fan

It might sound strange, but a super low sound axial cooling tower fan is an energy-saving device--But not because it uses less energy than the fan it replaces, because it doesn't. The reason is a little more complicated than that.

But first it makes sense to review a few basics about cooling towers.

The Basics

There are two major types of cooling towers and fluid coolers: Induced Draft and Forced Draft.


Forced-Draft towers utilize centrifugal fans to blow air through the tower. The air is forced into a pressurized plenum inside the tower and then through the fill. This means that access into these towers is limited, since doors must be able to resist pressure without leakage and tend to be small and difficult to use. This also makes it difficult to observe the basin of these towers while operating in order to troubleshoot problems if necessary.



Induced draft towers use an axial fan to pull air through the tower, creating a negative pressure within the tower. This allows the unit to be built in an open configuration, making access and observation far easier. In general, induced draft towers cost less, are easier to maintain and, importantly, require about half the fan horsepower to do the same cooling as a forced draft unit.

In fact, there are only a few reasons why you wouldn't use an induced draft tower in preference to a forced draft tower:
1. Height restrictions
2. Static pressure capacity for ducted installations
3. Noise Control

If you project requires an extremely short cooling tower or needs a tower to be installed indoors with ducted inlets and/or outlets, there is a good chance you will need to use the less efficient forced-draft tower. And, until recently, it used to be that the same was true of sound-critical installations. But not any more.

The acoustical benefit of forced draft units are twofold: First, they are quieter than induced draft units right out of the box. (Low-profile forced-draft units are especially quiet.) And, secondly, they can easily accept sound attenuators to make their already quiet performance even quieter. The price you pay, of course, is fan energy and dollars. Attenuators require that you expend even more money and fan energy than the already more expensive and less efficient bare forced draft unit.



Th super low sound fan (SLSF) changes the playing field. The addition of the SLSF on an Evapco induced draft cooling tower does not affect the efficiency at all--the performance is the same with and without the quieter fan. And since the fan knocks 9-15 dBa off of the sound power of the tower, suddenly induced draft fans are competitive in sound level with a forced-draft unit. Generally speaking (and each application is different) a SLSF induced draft unit is just about as quiet (if not quieter) than a forced-draft unit of the same capacity--and very competitive in first cost. And further sound abatement is available to shave a few more dB off of the sound level.

This development makes it very possible to meet demanding noise criteria and still retain the sizable energy benefits of the axial fan. And with innovative products like the Evapco ESWA, the lowest-sound option can even be the energy leader!

Hearing is believing, so Evapco has provided a few video clips to help you get an idea of how significant this sound improvement is [videos may require Internet Explorer to work properly]:

Video 1
Video 2

More information on low-sound options is also available here (pdf).

Why You Can't Buy an NC 35 Air Handler

No, it's not because we can't make quiet air handlers.

It's because NC isn't the right criteria to use to specify an air handler's sound level.

Why not? Well, to understand that, we have to discuss what NC is, exactly.

NC levels are defined by a series of curves that define the maximum sound level at a given frequency that an ambient sound can exhibit. Stated like this, it seems simple in the extreme to apply this rating to the sound level created by an air handler--but there is one important point missing: NC is a property of spaces, not equipment. Typically, allowable NC values are determined from charts like these:

It doesn't matter if the air handler whispers or is a screamer--if the sound levels in the space are below an acceptable NC curve, the sound level is acceptable. But this resultant sound level depends on a myriad of factors--the discharge sound level from the air handler, the duct layout, the selection of diffusers, attenuation devices, and, importantly, the room itself.

Hard surfaces and small volumes will tend to result in louder overall conditions, while large volumes and soft surfaces quiet a space. And, of course, sound generated in the space or from outdoor sources (traffic, etc.) will affect the overall NC level of a space.

How can you account for these effects? Well IAC has created a simple worksheet to determine the required insertion loss criteria needed in an attenuator array to meet a given target NC level. The SNAP sheet (Systemic Noise Analysis Procedure) is a simple method of calculating the resultant sound levels due to the HVAC system in a space. We've created a simple spreadsheet that helps keep the calculations straight here. Just simply copy the NC level you wish to meet from the green-tinted table at the bottom and paste those cells into the green bar at the top of the sheet. Then enter the discharge sound level from the air handler in the blue cells. If you follow the step-by step instructions on the SNAP form, you should be able to fill out the yellow cells with the sound attenuating characteristics of the system you are designing.

What you will have after putting in all this data is a required insertion loss criteria that should help you select a sound attenuator. Once you have picked one, just input its insertion loss performance into the first red bar, and the self-noise criteria in the next two bars. A successful selection will give you a sound attenuator that brings the sound level below the NC curve you are trying to hit, and does not generate enough self-noise to bring the sound levels back up above it!

This procedure only accounts for sound generated and transmitted by the HVAC system serving the space--You will need to account for other noise sources separately. But at least it takes care of the noise source you as an HVAC designer/contractor control!

How should you specify your air handler sound performance? By specifying the outlet, inlet and radiated sound power levels. As long as you have verified that the appropriate NC level in the space will not be exceeded with the specified values, you can be assured that any air handler meeting or beating your specification will be an appropriate fit.

Zero Pressure Drop Sound Traps?!

There's an old adage in life: There ain't no such thing as a free lunch.

In the world of sound traps, the "lunch" is insertion loss (the amount of sound attenuation provided) and the "bill" is pressure drop. Generally speaking, the more insertion loss you get at a given air velocity, the more static pressure drop you pay. And, when you consider that these pressure drops are greatly increased if the sound attenuator is located close to duct fittings (see page 8-9 of this document for examples), this bill can be very high indeed for any project where duct space is at a premium.

You know, all of them.

Enter the ZAPD™ series silencer from Industrial Acoustics (Z12A series cutsheet linked).


These silencers eliminate added static pressure drop by eliminating the air constriction that typical silencers impose upon the airstream:

Traditional Silencer

This is done by keeping the fill out of the airstream, so that the airflow sees no disturbance, and thus experiences no pressure drop:

New ZAPD™ Silencers

This design can provide significant insertion loss performance, with up to 12-14 dB in the first band for some 10' models, and 35-50 dB in the center bands for some 10' models--all with negligible self-noise and no additional pressure drop!

The exterior-baffle design does create functional limits to the size of these silencers, but they are great for use in systems where energy efficiency is a high priority, or where there is very little static pressure available, such as in VAV terminal boxes or water-source heat pumps.

Your free lunch just arrived. And it will pay you in energy savings for the rest of its life.