How can I reduce the noise coming from my fan?

How can I reduce the noise coming from my fan?

This is a common question these days.  Like most things in life, the noise generated by a fan is often the symptom of an underlying problem.  Many times, reduction in noise levels starts with a basic understanding of how noise is generated, how to properly select and install fans, and how to treat the resultant noise.

A whole book could be written about how to reduce fan noise, but we’ll only address a couple areas today.  The first one is proper selection of the fan.  The list is quite long, but two of the most common things that happen during fan selection is that the customer wants to squeeze everything they can out of a small fan, thinking that will save money.  In some respects, it does (usually a lower cost fan, smaller footprint, etc.), but in other way, it costs more (higher power requirements, more maintenance, etc.).  One important aspect is that a smaller fan, running at higher speeds, will typically generate more noise.  This isn’t a blanket statement, of course, but is generally true.

The other item that affects noise creation that must be understood during fan selection is the location of the fan in relation to surrounding walls, other equipment, and human operators.  We often get calls about how loud a fan is, only to find out that the customer had installed a 24” wall fan in a room that was 12 ft. square, and then wondered why the fan is louder than the catalog data would indicate.

The catalog data is created under ideal laboratory conditions, with no obstructions to the fan inlet or outlet, no nearby sound reflecting walls, and no background noise.  Using this data, which is typically expressed in either eight octave band sound power levels, or in a single number like dBA or sones, requires a working knowledge of what the number mean.  The most common misconception is that the dBA value in the catalog represents what you should be able to measure with a hand held sound meter.

A dBA level in the catalog is based on a whole set of assumptions, which seldom correlate to real life.  They include: a distance from the sound source, the directivity of the sound, and the type of acoustic environment.  Let’s talk thru an example.

A 36” wall fan is advertised to produce 70 dBA at 5 feet.  The biggest assumption in the calculation is that the sound is generating into a free field.  This means that the sound is free to radiate in all directions, and there is no reflection of noise back to the sound measuring device.  In real life, reflecting surfaces are nearly always present (nearby equipment, walls, etc.)  Only if the fan is installed in the middle of football stadium will it approach the catalog rating. 

The other factor is the distance.  The dBA calculation basically assumes that the sound is being generated by a point source.  So, you can’t really place your hand held sound meter 5 feet away from the fan in this example and expect to get 70 dBA, unless you are very lucky.

The calculation also assumes that the sound source is small enough that by calculating the data at 5 feet, you will be far enough away from the source that the sound waves have fully developed.  This condition is called the far field.  It is easiest to understand with an illustration.  Let’s assume you are trying to measure the height of a wave of water when you throw a rock into the lake.  If you try to measure right next to the spot the rock enters the water, it is nearly impossible to get a consistent reading.  All you can measure is the splash, and you would have a hard time telling if the splash is moving upwards, or you were measuring the splash returning downward.  You can try to get a average by taking a whole bunch of them, but you will just end up with an average of a bunch of bad data.  This is similar to the situation with trying to get a sound reading too close to the fan, and is called the near field.

If, however, you watch the effect of the rock entering the water, and get far enough away from the entry point, you will observe that the wave pattern quickly develops, and you can get a repeatable measurement.  This is called the far field.  Of course, the height of the wave diminishes with distance, so knowing the distance accurately is very important.  The far field is characterized by a reduction of sound as a function of distance by the following formula:

Sound2 = Sound1 - 20log(Distance 1/Distance2)

This means that far field sound will reduce by 6 dBA every time you double the distance.  Our testing indicates that for the 36” fan in our example, you may need to be 40-50 feet away before you develop that condition.  When you measure the sound level too close to the source, you can get readings that are wildly different from what you expect.

Sound coming from a fan is also directional.  You can hunt around and find low and high sound points.  The laboratory test is designed to blend all of these values together, but in the field, the directivity comes into play.

Let’s talk a little bit about sound attenuation.  One of the simplest (if you have an installation with ductwork), is to use insulated ductwork.  Fiberglass duct lining does a pretty good job of reducing noise, especially in the higher frequencies.  When you talk about sound attenuation, most people immediately think of sound mufflers, or silencers.  These are basically sections of ductwork that have an expanded outer diameter filled with acoustic absorbing insulation.  The insulation is covered with perforated metal, which minimizes erosion due to the airflow.  This type of silencer is limited in its ability to reduce noise because the length to diameter ratio, and the openness of the design, will allow lower frequency sound to pass thru.

One way to increase the effectiveness of this type of silencer is to add a center body, with could be round or a series of baffles, depending on the silencer configuration.  This breaks up the sound paths into narrower channels, so for a silencer of the same length, you have a much higher length to width ratio, which absorbs more noise.  These center bodies have smooth, rounded ends, in order to reduce the turbulence and pressure drop.

Trying to push too much air through a silencer leads to a phenomenon known as self-generated noise.  When high velocity air passes over the holes in the perforated metal, it creates noise.  Reputable silencer manufacturers have tested for this, and will include it in the calculation.

There are two other technologies used to reduce noise, but they are not common on fan applications.  One is a tuned resonance chamber, which is specifically designed to eliminate a single or known set of frequencies.  The sound enters an empty chamber next to the main air path, and gets absorbed and dampened inside that chamber.  Some of them also include air paths that change direction multiple times, in an attempt to increase their effectiveness.  These are also used in applications where higher temperatures are involved, such as in engine exhaust. 

One last method we’ll mention is active sound attenuation.  Think of your noise cancellation headphones.  The technology uses a microphone (to measure the noise spectrum), a computer (to analyze it, and create an equal noise level that is 180 degrees out of phase), and a microphone (to re-broadcast the noise).  What you end up with is actually more noise, but since it is out of phase, it gets cancelled, and you can’t hear it.  These devices are most effective when the noise is repeatable, and will have little effect on random noises.

As you can see, there are many facets to noise reduction.  Each one has a spot in the marketplace.  But remember, the most effective noise reduction is to properly select and install the equipment in the first place.  You can easily spend many more times the cost of the equipment to solve a noise problem in the field.