acoustic leak detection in boilers

8/6/2019 Acoustic Leak Detection in Boilers

1/7

Babcock & Wilcox 1

Recent Advances in the Application of Acoustic Leak Detectionto Process Recovery Boilers

J.J. KovecevichD.P. SandersBabcock & WilcoxBarberton, Ohio, U.S.A.

M.O. RobertsonBabcock & WilcoxLynchburg, Virginia, U.S.A.

S.P. NusplBabcock & WilcoxAlliance, Ohio, U.S.A.

Presented to:TAPPI Engineering ConferenceSeptember 11-14, 1995Dallas, Texas, U.S.A.

BR-1594

AbstractFor over two decades, acoustic signal processing meth-

ods have been used to detect leaks in pressurized systemsof utility and industrial power plants. These methods havebeen proven to be sensitive to the sound waves resultingfrom the turbulence produced at the source of the leak by

the escaping fluid or gas. Through ongoing research andoperational experience, effective and reliable methods ofapplying acoustic leak detection have been developed.

In theory, leak detection using the acoustic method issimple and straightforward. Sound waves are detected bytransducers (sensors) that convert the mechanical wavesto electrical signals. These signals are then processed andthe resulting root-mean-square (RMS) voltages are tracked.An increase in these voltage levels is relied upon toindicate the onset and growth of a leak.

In practice, however, the application almost always in-volves several considerations, including significant amountsof interfering background noises, unacceptable sound waveattenuations, and inaccessibility of desired sensor locations.

For these and other reasons it is imperative that properspecialized techniques, generally unique to a given applica-tion, be employed to ensure optimum sensitivity.

This paper describes the basic principles of applyingacoustic leak detection to the various components ofutility and industrial power plants. It explains the differ-ences between airborne and structure-borne monitoringtechniques and how to choose the proper method for theapplication. It describes the various techniques of fre-quency discrimination and proper sensor selection foroptimizing signal-to-noise ratios, thus maximizing sensi-tivity to leak noise. It shows how the presence of corrosion

and/or insulation adhesion adversely affects sensitivity byseverely attenuating structure-borne sound waves. Final-ly, it describes the importance of acoustically isolatingsensors from the extraneous and non-leak-related noisesusually present at sensor locations.

IntroductionThe availability of utility and industrial power plants isof major importance and, during the last 20 years, sub-stantial efforts have been made to reduce down-time.Much of the focus has been concentrated on improvedtechniques for the inspection and evaluation of criticalplant components prior to failure.

A significant part of this evaluation capability dealswith early detection of leaks in pressurized systems, espe-cially boiler tubes. In utility boilers, early detection ofleaks is primarily a financial issue. High velocity steamescaping from a tube leak can quickly cause extensivedamage to the adjacent tubes, which increases repair costsand down-time. High cost replacement power may also be

needed if a forced outage occurs during peak demand.When a leak occurs in the furnace of a process recovery(PR) boiler, the potential for an explosive smelt/waterreaction exists, and the need for reliable leak detectioncapability becomes a safety issue. Property damage andpersonal injury are the primary concerns.

Detection of tube leaks using fluid-borne acoustic sen-sors was first developed during the early 1970s in Europefor feedwater heaters and later applied in an airborne formto large power generation boilers. This technology wassubsequently introduced into the U.S. market by TheBabcock & Wilcox Company (B&W) in the early 1980s.


2/7

Babcock & Wilcox2

Structure-Borne Leak DetectionThe structure-borne method of leak detection has found

applications in valves and pressurized pipelines. Under arecent Electric Power Research Institute (EPRI) spon-sored project, a high frequency structure-borne approachwas found to be the best method for detecting leaks infeedwater heaters. The structure-borne technique usespiezoelectric transducers coupled to acoustic emissiontype waveguides which are weld-attached as shown inFigure 2.

In this application, a single structure-borne sensormounted to the outlet side of the tube sheet will fully covera high pressure feedwater heater. Low pressure feedwaterheaters require an additional structure-borne sensor mount-ed to the inlet side of the tube sheet.

An alternative structure-borne technique uses acceler-ometers, rather than piezoelectric transducers, mounted tothe ends of the waveguides. While similar performancecan be obtained in some cases, the inherent nature ofaccelerometer function can produce misleading results inothers. This will be discussed further in a later section ofthis paper.

While being successful in utility boiler applications,the airborne approach to leak detection has limitations inthe application to PR boiler furnaces. These limitationsarise from several factors, including:

The attenuation of airborne acoustic signals is greaterbecause of particulate loading and flue gas watervapor content.

Black liquor and smelt coat the furnace walls, makingit extremely difficult to maintain an open acousticpath from the furnace enclosure to an airborne sensor.

The airborne technique does not work when soot-blowers are running, and the new high solids firedunits have been reported to require continuous soot-blowing with as many as three lances active at anygiven time.

To overcome these limitations, alternative approacheshad to be evaluated. During the past few years, B&W hasinvested in a major effort to develop a technique that reliesupon the monitoring of structure-borne sound waves whichpropagate within the furnace membrane tube walls of PRboilers. This paper summarizes this experience with thedevelopment and application of both airborne and struc-ture-borne acoustic leak detection methods in power plants.

Technical Discussion

OverviewAs a leak develops in a pressurized system, turbulence

created by escaping fluid generates pressure waves withinthe contained fluid itself, throughout the low pressuremedium (usually a gas) into which the fluid is escaping,and within the container structure. These are commonlyreferred to as fluid-borne, airborne, and structure-borneacoustic waves, respectively.

To detect leaks, the energy associated with these me-chanical waves can be converted into electrical signals

with a variety of dynamic pressure transducers (sensors)that are in contact with the medium of interest. Severalmethods of signal processing are available that allow thevoltages generated by these sensors to be evaluated for thepresence of a leak.

As mentioned above, leaks in a pressurized systemgenerate sound waves in three media. The decision re-garding which types of acoustic waves are most reliablydetected is important from both functional and economi-cal considerations. This decision, in some cases, is notsimple. Factors such as background noise level, soundattenuation within the medium, signal processing strate-gy, and installation costs play a role.

Fluid-Borne Leak DetectionThe most well known use of fluid-borne leak detectionis in the earlier work on feedwater heater applications. Itwas ultimately found, however, to be inadequate due towidely fluctuating background noise levels during loadchanges. In other applications, this method is rarely usedbecause the sensor must be in contact with the containedfluid, which requires mounting it through the wall of thepressurized container.

Airborne Leak DetectionSince 1974, airborne leak detection has been predomi-

nantly used in large commercial boilers. Airborne meth-ods are well established and have detected leaks as muchas a week before any other means available.

In airborne applications, microphones or low frequen-cy resonant piezoelectric transducers are coupled byhollow waveguides to the gaseous furnace medium. Thewaveguides are usually attached through penetrations ininspection doors, unused sootblower ports, or the casing.

The airborne waveguide, shown in Figure 1, servesthree purposes. It couples the sound waves from thefurnace interior to the transducer face, protects the trans-ducer from excessive heat, and allows easy access to thetransducer for inspection or replacement.

Preamp Enclosure NEMA 4

Stainless Steel Flange

Refractory

Gasket

Cast Door

Teflon Isolator

Waveguide 1 in. 304 S.S.

Figure 1 Standard installation of preamp, sensor and waveguide.

Carbon Steel Flange with11/4 in. Nipple

Pipe Cap 1/2 in. NPT

Cleanout/Purge Air Connection

Sensor


3/7

Babcock & Wilcox 3

in utility boilers. In fluid-borne applications involvingpipes or tubes, the presence of gas bubbles will severelyattenuate the fluid-borne signal. In structure-borne appli-cations, the presence of corrosion, surface adhesion ofinsulation, and/or welded attachments will severely atten-uate structure-borne acoustic waves.

Signal Processing StrategiesIn all methods of acoustic leak detection, the objective

is to be able to tell the white noise generated by a leak apartfrom the background noise at a reasonable distance. Sincehigh frequency sound waves are more severely attenuatedthan low frequency waves as they propagate through amedium, a general rule of thumb is to use the lowestfrequency range possible where the leak noise stands outfrom the background noise. This will ensure the greatestamount of coverage by the fewest number of sensors.

The lowest useable frequency range is determined bytesting. With most sensors, a high pass filter is used toeliminate the lowest of the detectable frequencies whichcan overload the inputs to the signal processor. This filteris usually located near the sensor in the preamplifier,which is used to boost the signal level to a point where itcan be sent over a long distance to the signal processor.

One method of signal processing involves bandpassfiltering of the input signal at the optimum frequencyrange and then applying an RMS conversion to establishan average direct current (DC) voltage level of the signal.This voltage level is then compared to a fixed thresholdlevel that represents the maximum normal backgroundnoise level in the optimum frequency range. When themeasured signal level exceeds the threshold level for apreset amount of time, a leak is indicated. Within a filteredband, the RMS conversion method can easily detect leakswith as little as a 3dB (1.4 to 1) leak signal-to-backgroundnoise ratio.

Another signal processing scheme is to take the entiresignal and filter it into several frequency bands. The noiselevels in these bands are then monitored and assigned aweighted value based on their likelihood to be leak related.As in the first method, when the appropriate thresholdlevels have been exceeded long enough, a leak is indicat-ed. In some cases, an alarm is not generated until one of thehigher frequency bands confirms a suspected leak, whichmay not be until it has grown substantially.

A third signal processing scheme is to immediatelydigitize the incoming signal and perform a Fast FourierTransform (FFT) on it to produce the frequency spectrum.This spectrum is then stored in memory and comparedeither visually or by software to a previous spectrum that

is characteristic of normal background noise. If a suffi-cient difference exists between these spectra and the nor-mal spectrum for a long enough time, a leak is indicated.

Installation CostsEconomics play a decisive role in selecting one leak

detection method over the other. For an application whereno boiler wall penetrations are available, the structure-borne method would be an obvious candidate. No open-ings are needed, no pressure part cutting and welding arerequired, and all installation work can be completed fromthe outside of the boiler.

Sensor Mounting Plate

1/4 in. Stainless Steel Rod

45 Field Weld Prep

Background NoiseIn leak detection applications, the most important fac-

tor to consider is background noise within the propagationmedium of interest. Almost all background noise can be

characterized as white noise combined with discrete fre-quency noise.White noise can be defined as containing components

at all frequencies within a range or band of interest. Bothnormal boiler noise and leak noise are considered to bewhite noise. Boiler noise is best described as low frequen-cy white noise (rumbling) while leak noise is best de-scribed as higher frequency white noise (hissing).

Discrete frequency noise is usually composed of afundamental single frequency (tone) and several associat-ed harmonic frequencies. These sounds are best describedas whistling or humming noises.

Sound Wave Attenuation

In any medium, low frequency sound waves will beattenuated less, and therefore travel farther, than highfrequency sound waves. Also, in most cases, sound willtravel farther in solids and liquids than in gases. These areprimary considerations when determining how far awayfrom the sensor a leak can be detected.

Acoustic attenuation characteristics of the various me-dia also play a role in influencing which leak detectionmethod to employ. For instance, the gaseous mediumpresent in PR boilers imposes significantly more airbornesignal attenuation because of high concentrations of watervapor, saltcake, and suspended char that are not present

Isolator Disk

Sensor

Figure 2 Structure-borne waveguide.

4 in. Diameter

Sensor Holding Cap Screw


4/7

Babcock & Wilcox4

On the other hand, if several structure-borne sensorsare required to cover the same amount of area as oneairborne sensor, then the costs of the additional sensors,signal processing hardware, and cable installation need tobe considered.

Development of the Application of Structure-Borne Leak Detection to the Furnace MembraneTube Walls of Process Recovery Boilers

Over the past three years, a significant developmentproject was undertaken to determine the optimum methodfor using structure-borne acoustic leak detection to coverthe furnace walls of PR boilers. This method would also beextended to utility boilers, where openings for airbornewaveguides are sometimes scarce and removal of flyashfrom the waveguides must be dealt with. The programbegan in 1990 and was completed late in 1993. The basicoutline for this developmental effort was as follows:

I. Develop and characterize a leak source simulatorII. Perform preliminary laboratory testing on a sec-

tion of PR boiler wall panelIII. Locate an Alpha test site and perform field testsIV. Locate a Beta test site and perform wide scale

testing

Leak Source Simulator DevelopmentTo successfully develop a structure-borne leak detec-

tion system, the firs t need was to be able to simulate a leakand quantify it. To meet these objectives, a two part studywas undertaken. The first goal was to develop a device tosimulate a leak signal using compressed air as the energysource. The second objective was to develop a data base tocorrelate leak signals with respect to fluid, pressure, andorifice diameter.

Cold water, hot water, saturated steam, and air leaksignals were studied at pressures ranging from 30 psig to

2,000 psig and orifice diameters ranging from 1/32 inch to1/2 inch. Information gained from these studies provideda mathematical model that relates signal levels generatedby high pressure leaks through small holes with lowpressure leaks through large holes.

This recently patented leak source simulator, shown inFigure 3, uses 90 psi plant air escaping through a 1/4 inchdiameter orifice to simulate a 1300 psi steam leak througha 1/16 inch diameter hole.

Laboratory TestingOnce the leak source simulator development was com-

plete, the device was used to investigate the attenuationcharacteristics of a PR boiler furnace wall panel. This was

accomplished by strategically locating a few fixed sensorsand then moving a portable leak source simulator frompoint to point in a grid.

It was found that structure-borne sound waves travelmuch further along a tube than across the membranes fromtube to tube. Welded obstructions, such as windbox at-tachments, were found to severely attenuate structure-borne signals that cross them. It was further found thatthe injected signals could not be detected with a sensormounted to the header, even with the source mounted ashort distance away.

A rake assembly was then attached across the tubes to

determine if the directivity ratio could be improved. Therake was composed of 1/4 inch diameter pins welded toeach tube and all welded to a common 1-1/2 inch by 3/16inch bar that spanned the tubes. It was found that by doingthis, the source was detectable as well across the tubes asalong them.

Figure 3 Structure-borne leak source simulator.

RaisedSensorBracket

Initial Field Testing

Following the laboratory tests, several sets of fieldtesting were conducted at a southern paper mill to confirmand broaden the understanding of structure-borne leakdetection in the furnace walls of a recovery boiler.

One tube of this 900 ton/day PR boiler was instrument-ed from lower header to upper header with leak sourcesimulator waveguides spaced an average of about thirteenfeet apart. At the liquor gun level, a four foot wide rakeassembly was installed centered on the selected tube. Asingle floor tube was then instrumented with two of thesimulator waveguides spaced fourteen feet apart. The testlayout is shown in Figure 4.

The first objective was to determine leak signal atten-uation rates with the boiler off-line and empty, off-line

and water filled, and on-line. The second objective was tocharacterize background noise levels with the boiler on-line at full load. A third objective was to establish the bestcombination of sensor type and filtering strategy to max-imize the simulated leak signal-to-background noise ratio.

A wide variety of commercially available sensor types,including accelerometers, were tested. Overall, a frequen-cy span from 5 kHz to 200 kHz was investigated. Thesensors were tested both with and without being coveredby an acoustic enclosure used for shielding out externalbackground noise. Results from this field testing includedthe following:

1/4 in. Diameter Rod

Sensor Brackets

Air Outlet

Air Inlet


5/7

Babcock & Wilcox 5

With the boiler off-line and empty, simulated leak likesignals could be detected at high frequencies as far as90 feet along a tube. The signals could not be detected,even at short distances, in the upper or lower headers.

Substantial attenuation was found, as expected, whenthe signal crosses a windbox attachment weld.

Background noise levels were found to be relativelylow during full load operation. The floor tube back-ground noise level was considerably lower than thatof the wall tube.

The best performance was achieved using a 150 kHzresonant sensor. This sensor could detect the simulatedleak signal in excess of 30 feet along the tube. It wasalso found that this sensor was virtually immune tosootblower noise, even without the acoustic enclosure.

External air coupling tests accomplished by opening90 psi air valves near the sensors showed the 150 kHzsensors to be affected if the valve was less than 10 feetaway. We concluded that the acoustic enclosureswould be a necessary part of a structure-borne leakdetection system for this application.

The accelerometers were found to be extremely sus-ceptible to externally coupled airborne noise, evenwith the acoustic enclosure in place.

It was immediately determined that so far, only onetube in one boiler had been investigated. The existing fourfoot long rake had provided no useful information. A planwas then devised to install two levels of rakes covering aspan of half the width of each of two adjacent furnacewalls. The lower level would be installed just above theliquor gun ports and the upper level would be installed justunder the nose level.

Figure 6 Beta test rake layout.

Wall Panel

Figure 5 Rake installation.

Man Door

3rd Floor

TertiaryWindbox

East Wall Source Grid

New Source ColumnTube #52

Upper Rake Level

Original Source ColumnTube #50

Tube #28

Tube #6

Secondary Windbox

Primary Windbox

South LiquorGun Port

5th Floor

6th Floor

Key to Symbols

Source SimulatorWaveguide Location

Sensor Location

Figure 4 Field test layout.

Beta TestingDuring the Spring 1993 outage, the East and South

walls of the PR boiler at the southern mill were outfittedwith rake assemblies. The rakes were then instrumentedwith standard structure-borne waveguides, 150kHz sen-sors, and acoustic enclosures. Nine leak source simulatorwaveguides were attached to tubes in a rectangular gridbetween the rake levels on the East wall. The rake attach-ment details are shown in Figure 5 and the test layout isshown in Figure 6.

Sensor

Waveguide

RakeAssembly

Rake AssemblySupport Bar

MembraneBar

Original Rake

Lower Rake Level

Primary Air Windbox

Secondary Air Windbox

Simulator SourceWaveguide (Typ)

ElevationFeet

182.0 1

164.8 2

148.4 3

132.1 4

118.6 5

99.3 6

83.0 7

76.1 8

70.5 9

60.1 10

WaveguideNo.

Upper Furnace Sidewall Header

Lower Furnace Sidewall Header

Rake Assembly

Tube #50


6/7

Babcock & Wilcox6

Several rounds of testing were performed, each basedon the results from the previous round. During the latterstages, a task force was formed from various disciplines toevaluate the test results and make recommendations forthe following round.

The goal of the initial testing was to simulate leaks atthe various points in the grid and determine the responsesin and effectiveness of the rake assemblies. The resultswere as follows:

Due to geometric spreading, severe signal attenuationoccurred in the first few feet of rake. It was found thatthe current rake design was unacceptable for thisparticular application.

The testing sequence involving the sources in the gridprovided unexpected results. While the sources on theright side of the rake near the initial vertical row ofwaveguides produced signals that could be detected inthe rake, the grid sources in the middle and left sidesof the rake could not be detected at all.

Detailed investigation showed that the wall panelson the right side of the rake had been recently re-placed. They were found to be painted and the insula-tion had also been replaced. The remainder of thefurnace wall was composed of older unpainted panelsthat were corroded to the point where the fibers of theinsulation were actually intermeshed with the tubesurface. It was concluded that the leak source signalwas completely attenuated in these areas before it hadtravelled as much as ten feet.

A query directed to field service personnel producedan answer that corrosion-adhesion of insulation to thetube walls is a common occurrence in boilers of alltypes.

With the effectiveness of the high frequency structure-borne approach in question, it was decided to strengthenthe simulated leak signal. This was accomplished by

mounting 1/4 inch pipe nipples to the membrane betweentubes and injecting air directly through the wall. We foundthat the 50% loss in signal strength due to the weldedsimulator source waveguide was eliminated, as was theexternal airborne noise due to the simulator itself.

Lower frequency approaches were reinvestigated byperforming side by side comparisons of the airborne meth-od, a low frequency AE type structure-borne sensor, andan accelerometer. The side by side tests included leaksource simulations, sootblower operation, and generationof external ambient noise.

Conclusions and Recommendations

Final Conclusions In units where corrosion-adhesion of insulation ex-ists, the high frequency structure-borne approach willnot work. The low frequency structure-borne approach,although affected by the attenuation of the insulation,will work at a reasonable range.

The low frequency (2-20kHz) structure-borne ap-proach detects simulated leaks in the same wall muchbetter than the airborne approach. From a single weldattached waveguide, a simulated leak can be detectedabout four times further along a tube than across thetubes.The airborne approach, however, detects simu-

lated leaks in adjacent and opposite walls much betterthan the low frequency structure-borne approach.

The rake improves the lateral distance that a simulat-ed leak can be detected, but not by a wide enoughmargin to justify the cost.

A low frequency AE type sensor responds better thanan accelerometer, but not by a wide margin. Theacoustic isolation enclosure reduces the sensitivity ofthe AE type sensor, but not the accelerometer, toexternal noise. This means that accelerometers shouldnot be used if the primary purpose is the detection ofboiler tube leaks.

In corroded regions, membrane attachment of wave-guides shows better performance than tube attachment.

The airborne approach is more sensitive to near soot-blowers as the low frequency structure-borne ap-proach. The low frequency structure-borne approachis more sensitive to far sootblowers as the airborneapproach. This occurs because the higher frequenciesare reflected and/or absorbed by the furnace wall.

If a small leak occurs in a tube internal to the boiler,such as in the convection pass area, the walls block mostof the noise from a structure-borne sensor. On the otherhand, the noise will pass through an open waveguide andreach an airborne sensor. As an analogy, you can overhearthe conversation in the neighbors yard a lot better if youopen the window instead of listening through a drinkingglass pressed against the window. The result is that whilethe structure-borne method is best for detecting leaks thatoccur in or near the wall, the airborne method is the bestfor detecting leaks that occur internally.

Strategy for ImplementationWhen devising a scheme for implementation of these

findings, several factors were considered, including ge-ometry. The first general guideline is the intent to be able

to detect a 1/16 inch diameter (what is considered to be apinhole) leak at 1300 psi in a furnace wall tube.A minimum of a 3dB (1.4 to 1) signal-to-noise ratio is

required for a leak to be easily detectable. Generally,sound losses are expressed in dB per foot. In reviewing thedata and extrapolating, it was found that the simulatedsource produces a signal-to-noise ratio of about 20dB (10to 1) at the source. It was also found that the attenuationcoefficients are about 1/2 dB/ft along a tube and 2 dB/ftacross the tubes. If this is plotted out to the 3dB edge, adiamond shaped area of coverage results. The verticalpoints are 60 feet apart and the horizontal points are 16 feetapart.

If the boiler tested is considered to represent a typical

PR boiler with about a 1000 ton/day capacity, then thetypical PR furnace is a box that is about 35 feet wide, 30feet deep, and 120 feet tall. In order to get completecoverage, the first level of sensors would be placed justabove the lower headers at a spacing of 16 feet apart. Thislevel would contain eight sensors. The next level of 8sensors would be placed 30 feet above the first level andstaggered 8 feet sideways.

This pattern would then be repeated for the four levelsrequired to provide complete coverage of the lower fur-nace. If full furnace wall coverage is desired, six addition-al sensors would be located just below the upper wall panel


7/7

Babcock & Wilcox 7

An automatic purge system should be installed tokeep the waveguides clear of saltcake.

Lower furnace walls. Low frequency structure-borne leak detection tech-

niques provide the best performance for furnace walltubes.

The optimum structure-borne monitoring frequencyrange is 2-20 kHz.

It is recommended that structure-borne acousticwaveguides be attached to tube membranes.

The area of coverage by a structure-borne sensor isdiamond shaped. Each level of sensors should becenter staggered with respect to the level beneath it.

Structure-borne sensors should be equipped withacoustic enclosures for isolation from external air-borne noise.

Either piezoelectric transducers or accelerometers areacceptable for structure-borne furnace wall applica-tions, but one should be aware that accelerometersmay not be isolatable from external airborne noise.

Both airborne and low frequency structure-borne leakdetection methods suffer from interference duringsootblower operation. A minimum of a five secondgap of quiet time between lance pairs is required forthese methods to function.

Furnace floor tubes. Although the low frequency structure-borne approach

is recommended, it is not believed that current signalprocessing methods are fully adequate for PR boilerfloor tube leak detection. It is thought that floor tubeleaks under the char bed will produce burst typeacoustic emissions popping rather than the contin-uous type hissing. Feasibility studies have beeninitiated into using structure-borne sensors and pat-

tern recognition acoustic emission technology to de-tect leaks in floor tubes and smelt spouts. This tech-nology is also expected to be useful in detecting theoccurrence of composite tube cladding cracks throughthe same sensors used for furnace wall tube leakdetection.

headers. A diagram showing sensor locations and cover-age for two adjacent walls is given in Figure 7.

Recommendations for Implementing AcousticLeak Detection in Process Recovery Boilers

Upper furnace and convection pass. Airborne leak detection techniques provide the best

performance in the upper furnace and convection pass

areas. The optimum airborne monitoring frequency range is2-15 kHz.

Eight to twelve airborne sensors are typically neededto cover the convection pass area and upper furnace.

Two to four airborne sensors will cover the penthouse.

Sensor Location

30

ft

30

ft

Area of Coverage

16 ft

Lower Wall Panel Header

Figure 7 Structure-borne furnace wall leak detection imple-mentation strategy.

acoustic leak detection in boilers

Documents