Understanding, implementing genset noise control
With the expanding use of diesel gensets for emergency standby power, peak shaving, and demand response, there is an increased focus on controlling the noise these generators create. Whether gensets are located in enclosures outside a facility, inside, or on the roof, design engineers are making more efforts to control genset noise and vibration to reduce the effects on neighbors as well as building occupants. Whether gensets run continuously in prime-power applications, intermittently in demand response applications, or occasionally in emergency standby situations or testing, their operating sound levels may require remediation (see Figure 1).
While local codes and zoning laws often require nominal noise reduction strategies, it is end users who are currently driving the demand for quieter genset installations within available budgets. The cost to make modest reductions in noise is generally quite low, and genset manufacturers have made the job easier by making engines that run quieter and with less vibration. In addition, a large aftermarket exists in advanced noise control solutions. However, the cost of noise control is not linear. The first 10 dBA of reduction may be relatively inexpensive, but the last 10 dBA may be prohibitively expensive. That is why the current strategy in genset noise control is to control as much noise as possible without losing control of the budget. This article explains the science behind genset noise control and describes solutions that can help consultants and design engineers achieve optimum noise control solutions.
What is noise?
Vibrating objects induce pressure waves that travel through the air and reach our ears as sound. By definition, noise is merely unwanted sound. When the amplitude of the pressure waves becomes too high, the amount of sound becomes uncomfortable. In addition to being annoying, excessive sound can cause permanent hearing damage. This is why OSHA established safety standards for workers exposed to loud noises. Local ordinances and zoning laws also establish rules regarding the amplitude of unwanted sound, but these standards are geared toward reducing public annoyance rather than promoting safety.
The human ear has such a wide dynamic range that the logarithmic decibel scale was devised to express sound levels in a convenient way. The ratio between the softest sound the ear can hear and the loudest sound it can experience without damage is approximately a million to one or 1:1×106. By using a base-10 logarithmic scale, the entire range of human hearing can be described by a more convenient number that ranges from 0 dB, which is the threshold of normal hearing, to 140 dB, which is the threshold of pain (see Figure 2).
There are two dB scales used to describe sound: L and A.
- The dB(L) scale is linear and treats all audible frequencies as having equal value. However, the human ear does not experience all frequencies the same way. Our ears are particularly sensitive to frequencies in the range of 1,000 to 4,000 Hz, and they are less sensitive to sounds in lower or higher frequencies.
- To adjust the measurement of sound pressure levels to more accurately reflect what the human ear perceives, engineers use an A-weighting filter. This results in the frequency-weighted dB(A) scale, which was adopted by OSHA in 1972 as the official regulated sound level unit.
Engine mechanical noise: With the advent of high-pressure common-rail fuel injection, advanced turbocharging, and better combustion control, manufacturers have significantly reduced overall mechanical noise from diesel engines. The amount of sound varies with the size of the engine and its load, and can be as high as 110 dB(A) measured at 39 in. (see Figure 3). High-horsepower engines are actually quiet for their size because the V-configuration of the cylinders makes them self-balancing. Engines with more cylinders have more power strokes per revolution and therefore deliver a smoother flow of power with less vibration. Smaller engines tend to be harsher in operation and produce more noise and vibration for their size.
Exhaust noise: Engine exhaust is a major contributor to overall sound levels and, when measured without an exhaust silencer, can be 120 dB(A) or more, depending on the size of the engine. The sound level can be reduced by up to 24 dB(A) and up to 40 dB(A) depending on the silencer.
Cooling fan noise: Sound emanates from turbulent air as the cooling fan moves it across the engine and through the radiator. The amount of sound varies with the speed and volume of air being moved as well as with the design and distortion of the fan blades. The amount of sound can be as high as 95 dB(A) at one meter.
Alternator noise: The alternator has an internal cooling fan, and the combination of cooling air movement and brush friction produces a sound level that is always small compared to the driving engine.
Induction noise: Current fluctuations in the alternator windings create mechanical noises that add to total noise when load demand changes.
Structural/mechanical noise: This is caused by mechanical vibration of
various structural parts and components that is radiated as sound. Elastomeric isolators between the engine, alternator, controls, and other components help to reduce the amount of vibration that gets converted to noise.
There are two major frequency bands that not only emanate from different moving components on the genset but also require different abatement methods. Frequencies below 300 Hz are generally considered low frequency; frequencies above 300 Hz are considered high frequency. Furthermore, there are no universal standards for the amount of permissible genset noise. Rather, each application is different, and each locale sets its own standards for noise emanating from a property.
During design and manufacture, gensets are tested at the factory and sound levels are recorded using a process defined in ISO 3744. Sound measurements are sometimes done in the field but usually only to verify compliance to a specific local code or noise-reduction objective related to the installation. In the factory, the manufacturer develops sound
measurement profiles for each genset model rather than testing every individual unit. Test results are recorded and kept on file for each model in case there is a need to revisit the information for a specific installation.
To get accurate sound data, measurements must be taken in a free-field environment. As distinguished from a reverberant-field environment, a free-field environment is a location in which there are negligible effects from sound reflected off obstacles or boundaries. In practical terms, this means being about 4 to 6 ft away from a wall. Closer surfaces reflect sound and cause higher, erroneous readings.
In a typical factory test, technicians take sound measurements at 12 to 19 locations (depending on the size of the genset) on the outside of an imaginary box, 39 in. larger than the genset profile (see Figure 4). This procedure standardizes the measurements so that end users can rely on the data for installation and site planning. Usually this is done with a handheld calibrated
microphone one point at a time, but more sophisticated arrangements with
multiple simultaneous measurements are also used. Measurements are recorded as sound pressure readings and are later converted into sound power levels from each of the positions.
While factory sound measurements are taken at 39 in. from the unit for the sake of convenience, the standard distance quoted in specification sheets and used by the industry is 23 ft. Unless otherwise stated, all published genset
sound data is calibrated to what one would measure at 23 ft. Manufacturers simply use an algorithm to convert the 39-in. readings to 23-ft readings.
Noise-reduction strategies vary depending on whether the genset is located in
a building or outdoors in an enclosure (see Figure 5). In any case, it is vitally
important to not let noise-control solutions interfere with the flow of cooling air
to the genset. Low-frequency noise is the most difficult to attenuate and is best controlled by rigid barriers that have substantial mass. High-frequency noise can be controlled by acoustic foam and other types of sound-absorbing insulation.
Genset manufacturers generally offer sound-attenuating enclosures for units up to about 2,000 kW. Typically, these factory drop-over enclosures are offered in three grades of sound control from Level 1, which is a basic sheet metal enclosure, to Level 3, which has substantial sound
attenuating capability. A basic Level-1 enclosure typically reduces noise by approximately 3 dB(A), while a Level-3 enclosure can achieve reductions of 14 dB(A). For gensets larger than 2,000 kW, many custom aftermarket sound-attenuated enclosure solutions are available.
Overall, the six basic strategies for controlling noise are:
1. Acoustic barriers: Acoustic barriers are rigid and have substantial mass and stiffness to reduce the transmission of sound energy. Examples include sheet steel typically used in enclosures and sand-filled block walls or poured concrete walls used in indoor locations. As there has been a trend to lighter-weight sheet steel for enclosures as a cost-saving measure, it is sometimes necessary to install reinforcing ribs when steel enclosure walls lack sufficient stiffness. Steel panels can also be covered with a barium-filled rubber mat that adds mass and is very effective at preventing the transmission of low-frequency sound. It is also important to eliminate sound paths by sealing seams around doors, panels, exhaust ports, and conduit channels.
2. Acoustic insulation: Sound-absorbing acoustic foam is effective for
controlling high-frequency noise and is used extensively in outdoor enclosures. In indoor installations, it can be very effective at reducing noise when used to line air ducts or when used as a wall or ceiling covering. Fire-retardant urethane foam is the most common material used in enclosures. However, fiberglass is also effective.
3. Vibration isolation: Vibrating generator components induce pressure waves as sound into the environment. Also, anything that is attached to the genset can cause vibrations to be transmitted into the building structure or foundation. These attachment points include skid anchors, radiator discharge air ducts, exhaust piping, coolant piping, fuel lines, and electrical conduit. Mounting the genset on isolation springs or on a base fuel tank helps reduce the transmission of vibration into the foundation. Pouring a separate mounting slab also helps isolate vibrations from the building. Flexible connectors on fuel lines and exhaust and electrical conduits effectively eliminate the transmission of vibration.
4. Attenuating cooling-air noise: The movement of cooling air is a significant source of high-frequency noise, but restricting its flow is detrimental to genset cooling efficiency. More than 20 cubic meters per sec of air is required for cooling a 50-liter diesel engine. In indoor installations, high-frequency noise can be reduced by making the airflow turn two 90-deg angles as it enters and leaves the power room. In Level-3 outdoor enclosures, cooling air is drawn in from the roof near the rear of the enclosure and turned 90 deg in order to flow over the engine and through the radiator. It is then turned 90 deg again and ejected upward out the roof of the enclosure. In this way, much of the airflow sound at ground level is reduced and is directed upward, away from people and other structures.
5. Exhaust silencers: Silencers are available in several different sound-attenuation grades, commonly referred to as industrial, residential, or critical/hospital. The standard industrial-grade silencer reduces exhaust noise from 12 to 18 dB(A). Residential silencers provide an 18 to 25 dB(A) reduction, and critical/hospital silencers cut noise up to 40 dB(A). In indoor installations with long exhaust piping, the length of exhaust pipe alone provides some additional level of sound attenuation.
6. Maximize the distance from the source: Noise zoning ordinances typically set noise limits based on what can be measured at the property line. Since sound diminishes as the square of the distance from the source, simply increasing the distance from the property line may be enough to meet local regulations.
Today, the science of noise control is well understood, and genset noise can be controlled to a significant degree in both indoor and outdoor installations. The trend toward greater genset noise reduction is driven by a combination of local ordinances and the desire of end users to have a quieter work environment. Genset noise control is mostly limited by economics. The first 10 dB(A) of sound reduction is relatively inexpensive, while the last 10 dB(A) may strain the budget.
When controlling noise, it is important to not compromise genset cooling or overall performance and reliability. Barriers that contain noise also tend to contain heat, and any restrictions in cooling airflow over a genset can reduce its performance or threaten its longevity. By working with genset manufacturers or aftermarket enclosure suppliers, end users can achieve the maximum sound reduction for the available budget. In addition, end users should involve local zoning and regulatory agencies to ensure that the installation meets applicable noise regulations.
Bloxsom is an applications engineer and a vibration and acoustics analyst at MTU Onsite Energy, Mankato, Minn. He has three degrees in mechanical engineering with graduate level emphasis in vibrations. Bloxsom is also a member of the adjunct faculty at Minnesota State University in Mankato.