September-October 2005

The Experts Speak: What You Need to Know About Noise and Vibration

Just when you thought things couldn't get worse, along comes noise control. Regulations are spotty and inconsistent, product performance standards are nonexistent, and sound attenuation must be carefully balanced with engine performance strategies.

"The noise spectra for power generators varies widely," says David Gries, NVH Applications Engineer for E-A-R Specialty Composites in Indianapolis, IN. "But the noise sources are typically the same. They include engine noise, engine exhaust, turbulent airflow, and blade passage associated with cooling and alternator fans.

"Typically," says Gries, "exhaust is your dominant noise. But there's also turbo noise, engine noise, firing frequency noise, and vibrational noise. In power generation, noise breaks down into two forms—airborne noise, noise that is propagated in the air; and structure-born noise, which is propagated through a solid medium."

Advice for how to deal with this inevitable by-product of power generation is consistent across the board. You, or your sound engineer, or your equipment packager's sound engineer, should know precisely and in detail what noise requirements your equipment will be subject to, as well as the precise characteristics of the environment in which the equipment will be operating. Then you design for the worst-case scenario.

"As our urban areas have become denser," says Larry Hansen, principle engineer for Engineered Aeroacoustics in Minneapolis, MN, "concern about noise has quadrupled. Also we are seeing more cities and states adopting the World Health Organization recommendations on environmental noise, which is 45 dBA at nighttime, 5 dBA reduced from what's been typical."

The Lightness of Building
Richard C. Berger, chairman of Vibration Mountings and Controls Inc. in Bloomingdale, NJ, thinks building construction is another factor that has influenced increased concern about noise. "Buildings are being built lighter," says Berger. "Given this cost-cutting decrease in weight, new buildings are incredibly sensitive to induced motion. Any vibratory energy that enters this lighter-weight construction easily has the ability to excite that building, and an emergency generator has a propensity to throw a great deal of energy into a building. And not just locally around the generator, but literally for hundreds of feet horizontally and vertically. And it's not just vibratory energy. It's also sound.

"Sound is a change in the atmospheric pressure level. When the molecular pressure excites the air in one particular part of a structure and that air has the ability to move to another member of the structure, the sound will transfer. There's the additional factor in that in order to withstand earthquake codes, there's a certain level of stiffness and resiliency a building must match, and this often gets you into trouble when you have equipment that has an ability to generate a lot of energy.

"With power generation, it's not just the generator that creates the problem, especially in an indoor application. It's everything that's attached to the engine, including mufflers and piping. It's literally everything that each time it makes contact with the building creates an additional vibratory and noise source for intrusion into the structure."

Berger also notes an increase in what he calls hybrid power installations. "When a public school is designated for secondary use as an emergency preparedness center, the requirement for emergency power is upped from minimal demands, like fire protection and lighting, to handling major pieces of equipment, like heating systems. I strongly suspect we'll be seeing a push for reduction in noise and vibratory energy in these kinds of installations."

Common strategies for noise attenuation typically combine quieting equipment and soundproofing the environment in which the equipment will be installed. As a precursor to design, the sound consultant—or acoustical advocate as Hansen puts it—should (a) research relevant sound regulations, remembering that local jurisdictions can be inconsistent and that local ordinances trump regional and state regulations; (b) establish a sound profile for the equipment to be used, through consult with the engine manufacturer and/or modeling the system under load (or if already installed, onsite); and (c) establish the sound characteristics of the environment, including ambient noise.

"One thing most people neglect," says Hansen, is that sounds are additive. Someone will be working with a specification for say 50 dBA at a particular distance and they will design all the components for 50 dBA including the engine exhaust muffler, the radiator discharge silencer package and the mechanical noise on the air intake. Then they're surprised when they don't meet regulations. The problem is that all three of these sounds will typically sum at a point, and what you're left with may end up being 5 dBA higher than the criteria you're trying to meet.

"Another issue is that a lot of designers, architects, and environmentalists will readily put their arms around 45 or 50 dBA nighttime criteria, but in many parts of this country that criteria was established up to 40 years ago when the world was a whole lot quieter. If you design your sound attenuation to meet these specifications you can end up with installations that are quieter than the surrounding ambient sound readings, which means you've spent more money than you need to. What we typically do in a case like this is we document the situation and get a variance."

"We do probably 300 units a year," says Mike Witkowski, vice president of sales and engineering for Pritchard Brown in Baltimore, MD, specialists in sound enclosures. "And at least two thirds of these are sound-attenuated. Sound and vibration are both vibrational energy. One begets the other. If you have something touching a structure or some other medium, it's going to set that vibrating, and if something is vibrating and there's air, it's going to produce a sound.

In Control
"Because you don't want to overkill a system with controls, it's critical that you know exactly what's required, the specific dBA, and at what distance. And when it comes to sound, it's also critical to remember that it's not just the level and amount of acoustic energy that matters, it's also frequency. A low-frequency noise is perceived much differently by the human ear than a high-frequency noise. A transformer humming—which is a very low rumble, for example—is not nearly as annoying as the same amount of energy in a range where the ear is most sensitive.

"Because everything associated with acoustics works logarithmically, as you attempt to achieve more and more sound attenuation, you will eventually reach a point of diminishing returns, which means that to get slightly quieter, you will create an exponentially larger, more complex, and more expensive system. This means that when it comes to sound attenuation more is not necessarily better, unless you absolutely have to have it to meet regulations. And because of the logarithmic nature of sound, when you're doing sound attenuation, it's critical to look at each element of the system.

"Picture it in your mind. You have four different elements to a system. Three of the noise sources are fans that are making an amalgam of different frequency noise, but the fourth is a whistle, a pure single frequency almost like a tuning fork. Each of those four elements might be 100 dBA, but if you picture yourself lying in bed trying to go to sleep and you have this system running at the neighbor's property line, that fourth element, that one particular frequency, is going to stand out and potentially be much more of an annoyance than the other three more subdued sounds, even though they're the same amount of sound energy."

"On each package," says Brad Fennell, sales manager for Chillicothe Metal Co. Inc. in Chillicothe, IL, "we have to deal with the noise that comes off the engine block itself, the noise coming off of the radiator fan, and the noise coming out of the exhaust system. We have to address each one of those to get the overall sound level down to what has been specified. On the engine exhaust noise, we don't rely on catalogue or cut-sheet information, especially if it's a very, very quiet package. We work with the muffler manufacturers and get direct input."

Mating Ritual
Fennell reminds us that each engine has a characteristic sound signature, the result of detonation of gases inside the individual cylinders, which he insists is what makes consult with engine manufacturers such an important step in system design. "It can become quite a cumbersome process to mate and match everything so that you get down to the ultimate sound level," says Fennell, "which is probably the reason a lot of people don't do it, and which makes me think there are a lot of ticking time bombs out there."

The kind of accurate mating of mufflers to engines Fennell describes can be complicated by cost-consciousness and the fact that when it comes to engine silencers there are no industrywide performance standards. "A manufacturer can make whatever it wants and call the product whatever it wants," says Steve Stoyanac, vice president of sales for Silex Innovations in Spokane, WA. "But these are just industry names that have developed over the years—'industrial,' 'residential,' 'critical,' 'super critical,' 'hospital,' 'super hospital,' 'super hospital plus.'

"Literature on exhaust silencers specifies, for example, that for a critical grade silencer the typical attenuation or insertion loss through the silencer will be 25-35 dBA. But there are no standards that either demand that or hold a supplier accountable. We would rather work with manufacturers who publish standards for their individual systems—who specify, for example, that a particular gen set will not exceed 75 dBA at 7 meters. And then we work with them jointly to provide a product that will do that. We know we're going to be measured; we know we're going to be tested. But let's say Sacred Heart Hospital puts out a bid for a 500 kW gen set and part of that bid calls for a super critical silencer, the likelihood is that the engineer who wrote the spec doesn't really know what he wants. He's just looking at somebody's literature. And what he doesn't understand is there are no performance standards that hold that product accountable for actually achieving the level of sound attenuation he's after."

In addition to lack of accurate information, Hansen suggests that cost-cutting as well as misinformation often factors into bad sound attenuation choices. "To be competitive, the muffler manufacturers might size the muffler small to cut costs. But, in doing so, they increase the flow velocity through the muffler. When you do that the muffler will generate its own sound power so that, when the generator goes to full load, the muffler will sound like a jet engine on takeoff."

The Biggest Fan
Fans are an ancillary aspect of equipment-generated noise that also need attending to, which brings in the additional element of air flow. "There was a time," says Hansen, "when people absolutely believed that a muffler on an exhaust system was the end all and be all of sound attenuation. Now we've added the dimension of air movement, which includes the air coming into and exiting the generator room and how it's mixed. First and foremost the generator must have sufficient air for its location, altitude, and operating perimeters. There's no sense in having a back-up generator or a peaking generator if it's going to trip off because of high temperatures."

"Quieting a fan is very challenging," says Witkowski. "A major difference between cogeneration and standby applications is that 95% of standby units have unit-mounted radiators and engine-driven fans. And especially on the bigger sets, addressing that noise and air flow is as challenging as the mechanical noise of the engine itself. The fan makes a lot of noise and requires a lot of air, which means the designer needs to take into consideration how he or she will get the required air in and out of whatever enclosure the equipment is located in, while keeping the mechanical noise from the engine and fan from escaping, and while maintaining the appropriate static pressure so the system still functions. What makes this aspect of noise control so challenging is that one of the most troublesome by-products of acoustically treating an enclosed space is you're thermally insulating the environment at the same time. Even in a co gen unit where you're using the heat, the enclosure designer has to make sure to provide auxiliary ventilation so the enclosure doesn't get so hot the engine begins to de-rate.

"If the purpose of an enclosure includes sound attenuation, the source of mechanical and fan noise under full load must be ascertained. As the fan moves faster, the frequency and intensity go up, so in designing the fan you have to carefully choose one that provides the appropriate air flow and minimizes acoustical energy. But the physics of it are challenging. We had a situation in which the combination of fan-blade design and speed created a low-frequency kind of ringing. What we did was slow the fan down. We made it run at a slower speed but increased the number of blades. This made the ringing, the natural frequency of the system, easier to treat.

"It's important to note that dimensions, noise, and air flow requirements can vary greatly from manufacturer to manufacturer for a given kW rating, which means that if more than one gen set is being considered. sizing an enclosure based on worst-case specifications is often a good idea. It's important to remember that perhaps the most overlooked aspect of choosing sound attenuation is that the quieter you make it, the larger the enclosure will become. It's not unusual as sound attenuation approaches the —40 dBA range—which is generally considered the maximum economically feasible reduction with a pre-fabricated container—to see as much or more enclosure space dedicated to air handling than to the equipment being installed."

Gries agrees, "When you're developing a new enclosure design, you should give careful consideration to where the noise will radiate from the enclosure. Typically it's best to minimize enclosure openings and to incorporate tortuous paths, such as louvers, where openings can't be avoided. Eliminating enclosure openings, however, is typically detrimental to cooling the generator. Ideally it's best to design additional space for noise control materials to be incorporated to preclude interference with the functionality of the power generator. Doing so assists in optimizing the openings in the enclosure to maximize airflow for thermal management, while at the same time maximizing openings for better noise control. One way to facilitate this is to build computer models of generators to determine the unit's noise sources before tooling up. We can model different noise control techniques and transmission loss from panels and other factors such as that."

Fitting the Container
Mickey Wilburn, director of sales for Maxim Silencers in Houston, TX, identifies another aspect of enclosures he sees affecting sound attenuation. Maxim makes exhaust silencers and Wilburn says that more and more co gen equipment is being provided containerized, which means designers are asking for system components to be smaller and more compact. "You get your noise reduction typically through the larger diameter or larger length muffler," says Wilburn, "and when you're containerizing these engines, the challenge is to get the same noise reduction and still be able to fit the unit in the container."

"Typically there are two main methods for controlling the airborne noise with an enclosed power generator," says Gries, "either blocking airborne noise via a weighted barrier, or absorbing airborne noise with acoustical absorbing insulation. You can achieve significant noise reduction by lining the generator's sheet metal enclosure with a weighted barrier or a decoupled weighted barrier (barrier over decoupling foam). The ideal is that 90% of the enclosure should be lined, and enclosure openings should be minimized. But a barrier does not take sound out of the system, it merely reflects it, which means the energy builds up inside. This is in essence what a metal enclosure does—the metal acts as a barrier and blocks sound from getting out. On the other hand, when sound hits an absorber the energy is dissipated as low-grade heat. Air is pushed into the absorbing material by the sound-pressure wave and viscous forces dissipate the mechanical sound energy as heat. Sound openings for air intake, exhaust, and heat release are generally detrimental to the performance of barriers and decoupled barriers because they allow noise to escape unhindered. But by incorporating acoustical absorbers as a lining for louvers or creating a tortuous path for airflow, noise can be absorbed before it escapes the enclosure."

Gries points out that increasing the thickness of the absorber increases the absorption of the lower, more difficult-to-control frequencies. Some absorbers come with protective facings that help protect them from grease and fluids. And aluminized polyester facings can reflect radiant heat. Gries cautions that, in addition to heat, designers need to take into account other environmental constraints on sound absorption materials, including moisture, humidity, and UV exposure.

Baffling Devices
"At Pritchard Brown, we typically manufacture our own acoustic baffling devices," says Witkowski, "and we do laboratory testing to determine how these products perform aerodynamically, as well as acoustically. If you have too much resistance in the system, not only does it make it harder for the engine to breathe, it can create problems for the fan to pull air through."

Apropos of co gen applications, Witkowksi cautions that designers must also be up to speed on heat-recovery silencers. "A silencer on a gen set is like the muffler on your car. The exhaust gases come out of the engine and pass through a set of chambers and baffles and absorptive devices—the idea being to slow that air and quiet it down. When you're selecting a heat recovery silencer, you have to make sure that it has the appropriate acoustic properties to match with the system, and understand that you may need silencers in series. We had one job, for example, where a large telecommunications company wanted their system to be quieter than the ambient noise in the industrial park where they were located. To accomplish this we included four silencers in series."

Good Vibrations
The experts agree that the functional relationship between sound and vibration needs more consideration. Sometimes what seems to be one thing is really another.

Between sound and vibration, Hansen considers vibration the simpler and less expensive to treat with low-cost vibration isolators as long as they are properly designed and applied. "The single-most problematic issue when a system is not working properly is that the installers never released the transfer latches and didn't set up the isolators properly. Or, connecting the alternator, they have used such hard wires and conduits they completely short-circuit the vibration isolators."

"For the most part, noise is an annoyance," says Witkowski. "With vibration, the concern is that left uncontrolled it will cause wear, and it's only secondarily that it could be perceived by someone as annoying. The types of vibration controls in standby applications aren't really critical because, from an acoustic standpoint, what's set in motion by the vibration—the mass of this thing compared to the earth or the concrete pad or the building it's going to be mounted to—is so small that it's not perceived. But now you take that same package and you say I'm going to reduce the sound by 45 dBA, and all of a sudden standing outside the thing it's so darn quiet that part of what you hear is the engine vibrating to steel in the base and giving off sound waves.

Isolating the Problem
"Pretty much every reciprocating engine co gen or standby power package has either some sort of rubber vibration pad or a more complex tuned spring isolation pad or isolation device between it and whatever it's mounted to. With an enclosure we always recommend that that be done inside, where the unit itself is mounted to the floor. It's very rare that sound sets something vibrating that creates the issue. It's usually the other way around. It's usually the vibrating in the engine giving off the sound."

Berger thinks manufacturers are going to have to change their isolators to be able to tune a system much more effectively to locations the isolators are attempting to protect against. "We have systems in stock that are standard and very effective, but some of them can't possibly meet individual community expectations. One very bad 'for instance' is when emergency generators are located on a roof.

"An isolation system's effectiveness is based on the rigidity of the structure the isolator is being supported by. In the basement the floor is very stiff; a roof location is springlike. If you couple the spring system the generator manufacturer typically packages with its unit with a floor that itself acts as a spring, you approach a very horrible condition called resonance, and this sympathetic motion of two moving bodies causes destruction. The remedy to this is not complicated; it's a softer spring," Berger continues.

"So far indoors there has been no strategy that has been able to replace the isolated room, wherein the inner and outer room are separated by an isolation system so that nothing physically touches it. For outdoors, better isolation type screens are being manufactured, although their capabilities may be limited by local height restrictions regulations," says Berger.

Sound Seal in Agawam, MA, manufacturers the type of outdoor barriers Berger is talking about. Sound Seal combines sound barriers with absorbers to replace metal panels in outdoor sound attenuation applications. "Not everybody can afford metal panels," says Applications engineer Mary Ellen Riemenschneider. "For people who either don't have low-frequency sound they want to attenuate or who don't need 40 or 50 decibel reductions we have an alternative." The Sound Seal panels mount on chain link fencing and last 10 years or more outside. "People confuse absorption with barriers, but if you only have a barrier the sound will bounce up and over.

"The other issue I see," says Reimenschneider, "is that people think they can get away with one type of treatment or another. In sound attenuation that's just not true."