POWER SYSTEMS NOISE CONTROL SOLUTIONS

IES supplies complete packages and individual components and systems. IES power systems noise control solutions are supplied to OEMs, packagers, utilities, power and oil and gas contractors and end users.

IES solutions commonly provided include the following:

GAS TURBINE NOISE CONTROL

IES designs, manufactures, delivers and installs complete and partial noise control packages for high performance noise control applications worldwide for industrial gas turbines.

RECIPROCATING ENGINE NOISE CONTROL

IES designs, manufactures and supplies full packages, ISO containers, intake systems and reactive/absorptive engine exhaust silencers for worldwide application.

Acoustic enclosures for diesel power generation sets are fully internally fitted where required.

We offer a wide range of trailer mounted power generation units are supplied fully fitted including intercooling systems, electrics, pipework and exhaust systems.

On Noise Generation and Abatement

in Gas Turbine Installations

Jeff Lotton and Rainer Kurz

CONTENTS

1.0

INTRODUCTION TO TURBOMACHINERY COMPONENTS AND TERMINOLOGY

2.0

FUNDAMENTAL SOURCES OF TURBOMACHINERY NOISE

2.1

AXIAL COMPRESSOR

2.2

COMBUSTOR

2.3

EXHAUST COLLECTION / DIFFUSION

2.4

CASING

2.5

SUPPORTING EQUIPMENT

3.0

TYPICAL NOISE CONSTRAINTS ON TURBOMACHINERY INSTALLATIONS

4.0

STEPS TO ACHIEVE NOISE CONTROL LIMITS

5.0

VERIFICATION MEASUREMENTS

6.0

PRACTICAL LIMITATIONS OF VERIFICATION MEASUREMENTS

7.0

REFERENCES

8.0

APPENDIX

8.1

SAMPLE CALCULATION METHODS FOR DETERMINATION OF SITE SOUND

LEVELS

2

On Noise Generation and Abatement

in Gas Turbine Installations

1.0

INTRODUCTION TO TURBOMACHINERY COMPONENTS AND TERMINOLOGY

The typical arrangement of a gas turbine engine (Figure 1) involves several discrete steps to achieve

useable output power. The air is compressed, fuel is mixed in, and the mixture is combusted, forced

through a turbine, and then is exhausted to atmosphere.

Figure 1. Cross-Sectional Schematic of Typical Industrial Gas Turbine

Each of these steps involves large changes in gas enthalpy, which usually is the source of large pressure

changes. Pressure fluctuations are the basis of sound and its subjective counterpart, noise.

Although within the practice of noise control engineering, one must be very careful about the distinction

between the terms “noise” and “sound.” It is hoped that some informality taken within the context of this

paper will make the task of absorbing the subject matter easier.

2.0

FUNDAMENTAL SOURCES OF TURBOMACHINERY NOISE

2.1

AXIAL COMPRESSOR

The axial compressor uses stages of rotating airfoils usually with stationary sections of positioning airfoils

in between the rotating stages. Often, these airfoils are simply called blades. For a stationary observer

upstream of the inlet to the compressor, each time a blade passes the observer, there will be a large

fluctuation of pressure associated with the blade pressure field passing by (Figure 2).

3

Primary Noise Generation Mode of

Axial Compressor with Cascaded Blades

The pressure fluctuation superimposed over the entire inlet face of the compressor is the basic noise

generating mechanism from the axial compressor. Since the pressure fluctuations occur each time a

blade passes a fixed position and since the blades are mounted to the rotating shaft, the inlet noise levels

exhibit significant amplitude at a specific frequency. This is referred to as a “tone” or "blade passing

frequency” (BPF) and is a characteristic element of turbomachinery noise signatures.

The first stage of the axial compressor in the industrial gas turbine is the primary contributor to the

compressor noise. If the BPF and shaft speed are known, it is relatively simple to determine this

experimentally by measuring the BPF and dividing by the shaft rotating speed. The result should be the

number of blades on the first rotating stage.

The primary contributions of the axial compressor are the two “haystacks” on the right side of the inlet

sound pressure levels shown in Figure 3.

4

Axial Compressor Sound Level Frequency Spectrum

Further practical complications arise when active guide vane controls are considered. The total sound

pressure level of all of the axial compressor effects is not always highest at the highest speed or output

power, (Tyler and Sofrin, 1961).

2.2

COMBUSTOR

The process of combustion generates pressure fluctuations, which are the primary source of noise; the

pressure fluctuations then excite the natural frequencies of the working fluid volume itself. The resulting

pressure fluctuations are carried with the working fluid and are often transmitted to any surface that

contains the working fluid. Finally, when the working fluid is exhausted to atmosphere, the pressure

fluctuations are transmitted to the atmosphere. These pressure fluctuations are the source of acoustic

noise.

The combustor housing (Figure 4) and exhaust collector or ducting can have the acoustic noise

transmitted to them by the working fluid and then reradiate the noise to the surroundings. Ultimately, the

noise is fluid borne and transmitted to the surroundings by the flow conduits and the structures that

support them.

Because there is a large amount of low frequency acoustic energy in the working fluid during and after

combustion, transmission to any surface that comes into contact with the working fluid is a common mode

of noise generation in the combustion section of the industrial gas turbine. Generally low frequency

energy transmits through mechanical connections better than high frequency energy and, as a result, flow

conduits, connecting structures and equipment foundations can become significant sources of

combustion-generated noise.

Although there is some debate on the ability of the combustion-generated noise to travel through near

sonic flow nozzles, the contribution of combustion-generated noise in the exhaust flow has been

conclusively demonstrated.

5

2.3

EXHAUST COLLECTION / DIFFUSION AND EXIT TO ATMOSPHERE

The bulk mass flow of the working fluid through the various turns and cross-sectional changes

encountered while the working fluid is exhausted to atmosphere can cause flow-generated pressure

fluctuations. These pressure fluctuations can be from flow vortices around edges, flow separations from

transitions and bends, or flow interaction with stationary bodies. Flow-generated pressure fluctuations

can be significantly stronger than the pre-existing pressure fluctuations from the combustor, although the

frequency of flow-generated noise in exhaust systems is normally very low. A theoretically perfect

diffusion of the exhaust gasses to atmosphere can eliminate this source of noise.

The nature of the combustion process and the exhaust diffusion to atmosphere (Figure 5) in a practical

installation invariably causes significant turbulence and pressure fluctuations. These pressure

fluctuations are an additional source of acoustic noise over the existing combustion noise.

Figure 5. Typical Exhaust Diffuser / Collector, (J. Liu and C. Twardochleb, 2001)

COMBUSTOR

HOUSING

INJECTOR

FLANGES

6

Jet noise occurs because the boundary of stationary and moving flow generates vortices and

corresponding pressure fluctuations. The relative difference in speeds of the two gasses is the significant

factor. In most instances, there should be very little contribution of jet noise for Mach numbers in the

range of values typically seen in the exhausts of industrial gas turbine installations. Jet noise can be a

contributor when a high-pressure reservoir is bled to atmosphere, such as during blow down or bleed

valve actuation.

Flow-generated noise can become significant at typical exhaust velocities when extensive silencing is

employed and the residual silenced noise levels are low. Noise is “regenerated” by the flow as it passes

through the silencers, if the residual noise levels after silencing are low enough there may be some flow

generated contribution to the net noise levels at the exhaust to atmosphere.

As stated previously, the exit to atmosphere introduces the existing exhaust noise into a reacting medium

and noise is transmitted to atmosphere. This is often the primary mode of noise transmission from the

exhaust.

2.4

CASING

The engine "casing" is usually referred to as a significant source of noise and included in noise source

disclosures by equipment manufacturers.

In general, the casings contain the working fluid and fuels of the gas turbine and, as a result, are in direct

contact with the pressure fluctuations that are generated in the working fluid and fuels by the means

described previously. The casings are then a means of noise transmission through their own structural

responses. The casings often serve as bearing housings (or load paths) and can transmit noise

generated in the bearings.

Practically, there is less significance to the casing contributions of noise in industrial gas turbines

because, while the basic mechanisms of noise generation are present, the overall contribution to the net

noise levels is usually small. The primary reason that the casing contribution is small is that the relatively

heavy-walled casings have significant transmission losses, while the openings at the inlet and the

exhaust have little or no transmission loss. The result is that the contribution of the openings of the inlet

and exhaust are large in comparison. Also, it is common in industrial gas turbines to have a thin gauge

sheet metal housing on the inlet and exhaust. This thin sheet transmits noise more effectively than the

heavy castings of the casing.

True casing noise levels in industrial gas turbines are very difficult to obtain and the overall noise levels

around the entire package (less the direct contribution of inlet and exhaust openings) are considered as

the casing contribution.

2.5

SUPPORTING EQUIPMENT

Industrial gas turbines use supporting equipment that can itself be the source of noise. The top

contributors are oil coolers, pneumatic starters, driven equipment, and lube oil pumps.

In addition, the equipment associated with the gas compression process (piping, coolers and valves) are

also significant contributors of noise.

Coolers, whether they cool oil or gas, usually have fans, which are generally the primary source of noise.

The coolers can also radiate noise from their structures, which is generated by the fluid that is within the

heat exchanger. Most often, the net contribution of noise is by both sources. The fans and their drives

are the primary source of noise in most cases.

Pneumatic starters can be significant noise sources, but are transient in nature. Lube oil pumps generate

noise by causing significant fluid-borne pressure fluctuations, which radiate through piping and supports.

The pressure fluctuations are caused by the basic working mechanisms (gear meshes, for example) of

the pumps and are the source of fluid-borne noise. Pumps can also cause cavitation, although this is

usually not the most significant source of noise generation.

7

Piping that carries the process gas transmits noise generated by the compression process. The gas

carries with it the pressure fluctuations and transmits the fluctuations to any surface it comes into contact

with. Piping mounts can transmit the vibrations to the foundations or mounting supports causing the entire

foundation to be a source of noise.

Process valves generate noise associated with their operation. The noise is typically generated within the

body of the valve, transmitted to the casing and then into the surroundings. The valve generated pressure

fluctuations can be carried with the fluid itself and in the same fashion as the process gas in the above

paragraph, can become a significant noise source in addition to the valve body itself.

Driven equipment such as centrifugal gas compressors or gearboxes can be significant sources of noise.

Gas compressors generate noise in the diffusion of the gas (particularly if the diffusers are vaned). The

noise from centrifugal gas compressors is mostly tonal and fluid borne, the cases are usually heavy

enough so that they are less significant contributors. Exceptions do exist however.

Gearboxes can be the source of noise primarily resulting from the bearings and the gears themselves.

These types of machinery elements involve unsteady forces induced in part by the practical limits of their

machining tolerances. These unsteady forces get transmitted to whatever load paths exist. When the

forces are exposed to a reacting medium (such as air), the result is pressure fluctuations, which are the

source of airborne noise.

3.0

TYPICAL NOISE CONSTRAINTS ON TURBOMACHINERY INSTALLATIONS

The Federal Energy Regulatory Commission or FERC regulates any domestic interstate pipeline. FERC

regulates compression installations to a sound pressure level of 55 dBA, Ldn. The term "Ldn" refers to a

day and night weighted average that, when continuous sources are considered, equates to a sound

pressure level of 48 dBA. This criterion is typically applied to the property boundary or the nearest noise

sensitive area (NSA) or both. If the facility is not part of an interstate network, it can be regulated on local

or state levels and these criteria are normally applied to the property boundary.

International criteria are similar in concept, but vary widely in the required sound levels. The basic

underlying principle in all cases is to avoid disturbing the surroundings beyond some "acceptable" limit.

In addition to property boundary or NSA type criteria, there may be the need to control the noise levels at

specific locations within the installation. Almost invariably, the basic 85 dBA sound pressure level at one

meter distance criterion is the basis used to establish equipment noise limits. The 85 dBA criterion stems

from a health and safety aspect. It is based on an acceptable dosage concept and applies only to

workplace sound levels. It is mistakenly used as an equipment noise criterion in the hopes of achieving

compliance in the workplace, (Johnson, 2000).

These two methods of establishing criteria are the common recurring themes in the practice of noise

control of industrial equipment.

The two previously mentioned methods of establishing criteria are sometimes taken in conjunction and

the requirement is simply stated to meet the lowest of all criteria for any particular location in the

installation. The importance of establishing the correct criteria for each particular installation cannot be

overemphasized. Problems often arise when all aspects of increasing the site noise levels are not

considered.

4.0

STEPS TO ACHIEVE NOISE CONTROL GOALS

The steps used to control the various sources of noise are all based on the fundamental relationships

between each source and its path to the receiver. In addition, a criterion or noise limit (for all sources) at

the receiver needs to be identified.

8

The source-path-receiver concept of noise control is applied to each of the contributing sources in a

method consistent with the characteristics of each source and its path to the receiver.

The primary path in property boundary criteria is usually airborne noise that emanates from the gas

turbine intake and exhaust, oil cooler, process gas piping and valves and process gas coolers. The gas

turbine casing is enclosed in either a building or a close-fitting acoustic enclosure. Driven equipment is

often enclosed in a similar manner.

Variations on the theme revolve around additional treatment methods. Close-fitting acoustic enclosures

may themselves be placed within acoustically treated buildings. Walls or vegetation barriers may be

used. Occasionally, the driven equipment is unenclosed for maintenance reasons.

Table 1 lists typical sources encountered in an industrial gas turbine installation and the most common

treatments associated with each.

Table 1. Typical Sources and Treatments for Industrial Gas Turbine Noise

Source Description

Typical Treatment

Axial Compressor

·        Package Enclosure

·        In-Duct Dissipative Silencer

·        External Lagging

Combustor Housing

·        Package Enclosure

·        External Lagging

·        Mechanical Isolation

Exhaust to Atmosphere

·        In-Duct Dissipative Silencer

·        Direct Stack Exit Away from Receiver

·        External Lagging

·        In-Duct Reactive Silencer

·        Mechanical Isolation

Gas Turbine Casing

·        Package Enclosure

·        Offskid Building

·        Both of the Above

Lube Oil or Process Coolers

·        Reduce Fan Tip Speed

·        Enclosure

Process Piping

·        Bury

·        External Lagging

·        In-Line Dissipative Silencer

·        Mechanical Isolation

Gearboxes

·        Package Enclosure

·        External Lagging

The most significant elements of noise control are proper source identification and source noise level

evaluations. Source noise levels need to be established prior to implementing abatement measures. The

“unsilenced” (data taken without the presence of a silencer) noise data must be of high quality to achieve

reliable results.

The most useful and reliable source noise data are those obtained under the strict adherence to

recognized standards.

9

5.0

VERIFICATION MEASUREMENTS

Verification of the steps used to control noise is generally a component of meeting the regulatory

obligations made to the controlling authority.

Verification measurements should follow the agreed upon criteria at the agreed upon locations. Typical

measurements (Figure 6) are made with integrating sound level meters of various types. Spectrum

analyzers may be required in cases where the frequency content, as well as the overall sound pressure

level, is part of the criteria. Spectrum analyzers are also indispensable when attempting to isolate

sources and contributing factors.

The general practice of taking measurements should be based on recognized standards wherever

possible. Examples of existing standards for gas turbine installations can be found in the references

listed in Section 8.0 in this paper. In many cases, no real standard will exist to cover the specific needs of

an installation and, in these cases, the verification measurements need to be taken in a previously agreed

upon fashion.

Figure 6. Typical Instrumentation. From left, B&K 2144, _” microphone, B&K 2230 and B&K 2260

10

It is not unusual for verification measurements in general to have some predefined tolerance. For some

unknown reason, noise criteria are typically taken as an upper limit. It is often unacceptable to claim that

the requirements have been met if the agreed upon limit is even slightly exceeded by the verification

measurements. The process of collecting noise data inherently involves tolerances. The reality of the

conflicts between the "upper limit" mentality and the inherent tolerances involved in collecting the data is a

frequent cause of concern. Inevitably, the practicing noise control engineer learns how to cope with the

tolerances while meeting the "upper limit" expectations, (Johnson, 2001).

6.0

PRACTICAL LIMITATIONS OF VERIFICATION MEASUREMENTS

Unfortunately, noise measurements can be difficult or impossible to be used verbatim to show

compliance. There can be several reasons for this, but the biggest reason is normally that the equipment

supplier is only interested in verifying their own equipment and the measurements almost always include

some environmental influences. Obviously, verification measurements could be made in a controlled

environment, such as an anechoic chamber, but that is impractical and expensive.

Industrial gas turbines are no exception to being impractical candidates for anechoic measurements and

the applications they find themselves in typically do not warrant the extra costs involved, even if anechoic

measurements are possible.

Measurements are usually taken when the equipment is installed and all other supporting equipment is

fully operating. As discussed above, the sound fields become complicated as a result of additional

sources, reflecting surfaces, weather effects, other human activity, and wildlife.

The task of the noise control engineer in these situations is to make as much effort as possible to obtain

useful data that can be used to show compliance (or to offer recommendations for remediation)

analytically. This is an inherently difficult proposition because the client may or may not be inclined to

accept the analysis when measured data are available. Carefully isolating each component of a

contributing source is often necessary.

Several test standards exist for field measurements of industrial gas turbines, [1,2,3,4,5]. It is necessary

to agree up front upon the types of measurements to be made and the limitations that can arise from

them. If a manufacturer declines to use a recognized standard in its equipment noise declarations, the

manufacturer should at least propose an explanation of how the data were derived. Verification should

then proceed in the same manner as was used to generate the data.

If the criteria were incorrectly chosen, there may be no possible way to make verification measurements.

For example, the criteria may be such that the sound levels to be verified were significantly below the

ambient sound levels. It is rare to find a place that has ambient sound levels lower than 35 dBA at any

time of the night or day.

Several techniques are available to the noise control engineer when trying to make good verification

measurements. A few of these techniques are described in the standards. Perhaps the most useful

technique is to simply get closer to the source that needs to be isolated while taking narrow band data, (at

least 1/3 octave). This can be done in several steps so that the change in frequency spectrum, as well as

the overall change in amplitude, can be observed as the measurements get closer to the source.

Shielding sources can be a valuable tool in determining the effects of the source. A position can usually

be found that shields one or more of the contributing sources and, by a process of elimination using the

shielded frequency spectrum of various positions, the relative contribution of each source can then be

determined.

Another promising method is the use of sound intensity evaluations. A commonly used standard for

making sound intensity evaluations is ISO 9614, [2]. Sound intensity evaluations involve two

microphones making simultaneous readings. The difference in the readings is then analyzed to

determine the direction as well as the magnitude of the sound field. Using sound intensity evaluations

can help isolate individual sources. The main drawbacks to intensity evaluations are the inherent

complexity of the method and the need for evaluation of several data quality indicators. If the data quality

indicators are not monitored carefully, the usefulness of the measurements is doubtful. Sound intensity

11

evaluations have not been as widely used as “normal” sound pressure level measurements and there

does not seem to be a great deal of momentum towards using sound intensity evaluations over sound

pressure level measurements.