NBSIR 75-653
EPA 550/8-76-002
Measurement Methodology and
Supporting Documentation
for Portable Air Compressor
Noise
Curtis !. Holmer
Institute for Basic Standards
National Bureau of Standards
Washington. D. C. 20234
January 1975
Final Report
Prepared for
U. S. Environmental Frotect.on Agency
Office of Noce Abatement and Control
Weshingtoi.. D C. 204SO
-------
NBSIR 75-653
EPA 550/8-76-002
MEASUREMENT METHODOLOGY AND
SUPPORTING DOCUMENTATION
FOR PORTABLE AIR COMPRESSOR
NOISE
Curtis I. Homer
Institute for Basic Standards
National Bureau of Standards
Washington, D. C. 20234
January 1975
Final Report
Prepared for
U.S. Environmental Protection Agency
Office of Noise Abatement and Control
Washington, D. C. 20460
/yN
U.S. DEPARTMENT OF COMMERCE, Rogers C.B. Morton, Secretary
James A. Baker. Ill, Undersecretary
Dr. Betsy Ancker-Johnson, Assistant Secretary for Science and Technology
NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Acting Director
-------
TABLE OF CONTENTS
Page
Summary 1
1. Background 2
1.1. Source Description 2
1.2. Parties Affected by Noise 3
1.3. Present Accepted Noise Measurement Methodologies 3
2. Rationale for Methodology
2.1. Introduction 5
2.2. Determination of Sound Power 5
2.3. Use of Weighted Sound Levels 7
a. Limitations of Weighted Sound Levels 7
b. Representative Noise Spectra of Portable Air
Compressors 8
c. Use of A-Weighted Level 8
2.k. Discussion of Accuracy and Precision 17
3. Review of Methodology 18
U. Portable Air Compressor Methodology 22
U.I. Purpose and Applicability 22
U.2. References 22
U.3. Measurement Uncertainty 22
k.k. Acoustic Environment 22
a. General Requirements 22
b. Criteria for the Adequacy of the Environment 23
c. Ambient Noise 23
d. Other Ambient Factors 23
I*. 5. Instrumentation 23
a. General Requirements 23
b. Additional Instrumentation Requirements 2k
k.6. Installation and Operation of the Source 2k
a. Source Operation 2k
b. Source Monitoring 2k
U.7 Measurement Procedure 25
a. Microphone Positions 25
-------
k.B. Information to be Recorded 25
1*.9. Calculation Procedures ' 31
5. Acknowledgement 36
6. Bibliography 37
Appendix A. Procedure for Predicting A-Weighted Sound Level from
A-Weighted Sound Power 39
-------
Summary
This report presents recommendations and supporting rationale on a
measurement methodology for portable air compressors. The methodology is
believed to be general enough to allow determination of noise emission
from other outdoor stationary noise sources. The methodology provides for
the determination of A-weighted sound pover level or the equivalent
weighted sound pressure level at a reference distance. A-weighted level
is used because of its strong, positive correlation with community
response to noise from internal combustion engine noise. It is
recommended, however, that the spectra associated with the regulated
source be monitored in some manner to insuEe that the spectra remain
similar to those for which A-weighted sound level retains good correlation
with community response. The methodology uses weighted sound level
measurements at eight positions on a curved surface surrounding the source
at a distance of one metre from the surface of the machine. Data recorded
at these positions are used to calculate the average A-weighted sound
pressure level of the machine on the measurement surface. This is
combined with the area of the measurement surface to give the sound power
level of the machine. From this value, a rating sound pressure at a
rating distance may be calculated by subtracting a constant value.
While a specific methodology is proposed which is believed to be
consistent with achievable accuracy and precision of field measurements
conducted for regulatory purposes, we believe that the flexibility
inherent in regulation by use of sound power level should be exploited by
permitting the use, for quality control purposes, of any measurement
methodology which provides estimates of the sound power level which has
demonstrable precision and accuracy consistent with the difference between
the actual level radiated by the source and the regulated level.
Procedures which permit the rapid estimation of A-weighted sound
level are presented in Appendix A. These are applicable for estimation of
A-weighted sound level in a variety of circumstances when the sound power
or equivalent sound pressure level at a reference distance is known.
Key Words: Acoustics; air compressor; internal combustion engine; noise;
sound power level; sound pressure level.
-------
1. BACKGROUND
The construction site is an omni-present fixture of present life in
U. S. cities, and represents a significant source of noise to occupants
and residents in nearby areas. While the noise arises from many sources
within the site, the U. S. Environmental Protection Agency has identified
portable air compressors as a major noise source under authority of
Section 6 of the Noise Control Act of 1972. At the request of EPA, the
National Bureau of Standards, with the. cooperation of industry, has
derived a test procedure applicable to portable air compressors. In this
report we describe this measurement methodology and review the rationale
behind the decisions which led to its development.
1.1 Source Description
Portable air compressors are self-powered devices which produce
compressed air for the operation of other equipment. They are portable in
the sense that they are mounted on trailers or truck beds, to facilitate
moving from site to site. The machine consists of a driver, i.e., a
gasoline or diesel engine (or possibly an electric motor), and a driven
element which is a reciprocating, or rotary vane or rotating screw pump.
Currently, units are available which may be described as standard and
quieted (the latter category generally implys some degree of noise control
treatment added to an otherwise standard unit). Also included is an air
receiving tank, an automatic regulation system which matches air output to
demand, and suitable performance indicating gauges such as air pressure
and engine speed. The output automatic control system may be of an on-off
or continuously variable type. In the former case, depending on air
pressure demand from the receiving tank, the engine will be operated
either at idle or rated (i.e., governed) speed. In the continuously
variable case, engine speed is varied through a control loop to match
compressor output to air demand. Because of the relatively fine gradation
in model size (there are approximately twenty-five model sizes to span the
output range from 85-2000 cubic feet per minute (cfm)) compressors tend to
be selected to closely match the demand requirements for the tools which
they power, and are probably operated very close to rated output nearly
all of the time when they are in use. In physical size, the units vary
from about O.Tmxl.Omxl.Sm1 with a 30 horsepower engine to
approximately 3mx3mx6m with a 650 horsepower engine. Within the
range of compressors manufactured, there are logical divisions, for
example, as a function of driver type and pump type. There is a less
clearcut division on size. Virtually all gasoline powered units have a
rated capacity below 250 cfm (U2U m /s), while all diesel units are larger
than 125 cfm (213 m /s) capacity. No similar categories for pump type
occur, since.all three types are used throughout the range of sizes
produced.[l ]-/
Numbers in brackets indicate literature references at the end of this
report.
-------
The predominant noise sources associated with the compressor opera-
tion are those associated with the operation of internal combustion engines
(i.e., engine intake, exhaust and casing radiation plus cooling system fan)
and also radiation from the compressor casing and air receiving system and
associated valving. No single routine operation leads to the production
of noise from warning devices or from pressure relief valves.
There does not appear to toe Justification for fundamental variations
in the measurement methodology to' accommodate the above mentioned machine
parameters.
There is no typical use pattern established for portable air
compressors since the requirements and site descriptions vary substantially
from construction Job to construction Job. The site may vary from a large
controlled area (such as a site for a major building or power plant), to
a city street where pavement is "being removed. In a typical urban site,
the compressor is frequently located on or near the street where passers-
by and nearby building occupants may be expected to receive substantial
exposure.
1.2 Parties Affected by Noise
The two principal groups affected by compressor noise are construction
workers on the Job site, and the surrounding community. The nature of
the effects on construction workers range from task interference and
annoyance to possible hearing loss. Potential hearing loss, task interference,
communication interference, and annoyance will constitute the most significant
effects in the community. Construction workers presumably will be protected
by occupational health and safety regulations, with the possible exception of
small construction firms which do not engage in interstate commerce. The
proportion of exposed workers in this latter category is unknown. For these
reasons the principal concern in this methodology is to equitably assess the
noise emission which contributes to community noise levels.
1.3 Present Accepted Noise Measurement Methodologies
The Compressed Air and Gas Institute and its European counterpart
(CAGI/PNEUROP) have promulgated a test code for the measurement of
noise from pneumatic equipment[2], including portable air compressors,
which is now nationally and internationally accepted[3]. This code
consists of determining the average octave band and A-weighted sound
pressure level at ten positions near the ground plane (1.2-1.5 meters
above the ground) at 1 and 7 meters from the surfaces of the machine.
Current standardization activity includes proposed modifications to these
standards which incorporate sound power level determinations[U].
The data resulting from use of the CAGI/PNEUROP methodology are not
equitable measures for characterizing the noise radiated to the community
for two major reasons. The first problem is that the sound pressure level
(SPL) measured at a fixed distance from the surface of a machine is not
-------
simply proportional to the level at another distance from a machine, since
the proportionality constant depends on machine size. Consider, for
example, two machines (whose sizes represent the extremes mentioned earlier)
with the same SPL at 1 meter from the surface. At 30 meters from the center
of the machines the SPL from the larger machine will be about 5 dB higher;
while for the same SPL at 7 meters, the difference in SPL at 30 meters will
be about 2 dB with the larger machine again higher. Thus it is clear that
without some allowance for machine size,- the present methodology will not
produce an unbiased estimate of noise in the community. A second problem
arises because the CAGI/PNEUROP code provides for no measurement above
the unit, so that noise radiated to upper stories of nearby buildings
would not be controlled. Both of these problems have been recognized by
the industry which is one reason for the present work on modification of
the international standards.
Both of the above problems may be rectified by a change of determined
quantity from average sound pressure level to radiated sound power or
sound power level. The proposed methodology is based on this latter
concept.
Present methods for determining sound power involve the calculation
of average sound pressure level from measurements on large radius hemispherical
surface, centered on the source which is placed on a hard reflecting plane[5],
combined with an added constant which is proportional to the logarithm of
the surface area of the hemisphere. An alternate procedure[6] involves
the determination of average sound pressure over a surface surrounding a
device at a uniform distance from the surface of that device, times a
constant which is equal to the area of that measurement surface. The
former method involves extensive test site requirements which make the
procedure very difficult to use directly as a quality control or field
enforcement test. The latter test has good possibilities for such use
because it does imply the possibility of a measurement at short distances
from the machine with associated easing of test site and background noise
requiremen^s. The major problem associated with this latter test is the
relatively lower experience with use of the test by the acoustics community,
which creates difficulties in estimating, a priori, how many measurement
positions are required to achieve a given level of precision, or what
accuracy may be expected.[7]
-------
2, RATIONALE FOR METHODOLOGY
2.1. Introduction
Section U of this report consists of a detailed description of the
measurement methodology for characterizing the noise from portable air
compressors with respect to the resulting noise exposure in the
community. In the following paragraphs of this section we review the
information which led to the proposal of this methodology, and which
in fact constitutes the rationale for the decisions made during its
preparation. The remainder of this introduction will provide a brief
overview of the methodology, while subsequent paragraphs will address
the significant decisions concerning the use of sound power and weighted
sound level to describe the source, the questions of accuracy and precision
in the measurement, and finally a detailed discussion of the important
paragraphs within the methodology itself.
The recommended methodology involves the determination of the A-weighted
sound power output of the source. The sound power output is estimated
from the average of eight measurements of sound pressure level (SPL) on
a measurement surface which is one metre from the surface of the source,
with the source placed on a hard reflecting plane.
2.2. Determination of Sound Power
The sound power output of a device is a quantity which characterizes
the power radiated by that device in a manner which is, to first order,
independent of the surrounding environment. There is a strong analogy
between thermal and acoustical problems which may assist in visualizing
this situation. In both cases power is used to characterize energy output
per unit time. Temperature is the variable in the thermal case which
corresponds to sound pressure in the acoustic case. Both variables are
measures of the intensity of the radiation from the source at a given
position. In either case, the value variable is directly related to the
source strength and is also influenced by the environment. As energy
propagates away from the source, it may be thought of as spreading over
a surface of equal distance from the source. In the case of a small
source, for example, which radiates in all directions, placed on a non-
absorbing surface, the power radiated from that source is distributed
over a hemispherical surface.
The acoustic power W radiated by a simple source such as a pulsating
sphere is determined from the equation[5l
W = S/pc (1)
2
where p is the density of air, c the velocity of sound in air, is
the space and time average sound pressure, and S is the area of the surface
on which is determined. This equation is accurate within one decibel
for complex sources when the effective radius of the measurement surface
-------
is greater than (120/f meters) vhere f is the center frequency of the
frequency band of greatest contribution to the weighted sound level. It
should be noted that this equation appears to produce an upper bound estimate
of the sound power determined by other standard procedures under typical
conditions[8].
Since the area of the hemispherical surface is proportional to the
square of the radius, the intensity (i.e., power per unit area) is
inversely proportional to the square of.the distance from the source.
Thus if the distance from the center of the source is doubled, the
intensity is reduced by a factor of four. Since a 3 dB reduction in sound
pressure level (SPL) corresponds to a halving of sound pressure, a
doubling of distance in this case will produce a 6 dB reduction in SPL.
The total power radiated may be determined from the mean square sound
pressure averaged over a measurement surface, the area of that measurement
surface, and the characteristic impedance of the ambient air.
The sound radiated from a device of complex shape spreads out on a
wavefront which is always at a uniform distance from the surface of the
source. Thus, in principle, the determination of sound power can be made
on any measurement surface which is a uniform distance from the source
surface. At large distances from any source mounted near a reflecting
plane, the wavefront surface becomes nearly hemispherical in shape, and a
hemispherical measurement surface is appropriate. Thus the determination
of sound power output may be made on any hemisphere of any arbitrary
radius greater than some appropriate minimum radius subject to other
limitations of a practical nature which affect the accuracy or precision
of the measurement. The major practical limitations are imposed by
background noise, atmospheric propagation effects, the necessity of
testing over reflecting plane which is larger than the test hemisphere and
accuracy of'microphone positioning. The first three factors all become
increasingly important as distance increases, and thus minimizing the
measurement radius is desirable. The last factor is more significant for
smaller radii, and thus also suggests a minimum radius.
The foregoing information suggests that, in principle, it does not
matter whether or not the measurements are made on a hemispherical surface
whose radius is several source dimensions, or on a surface surrounding the
source at a fixed, relatively small distance as long as the effective
radius is greater than the minimum specified for complex sources. The
same information is obtainable using either surface, but the number of
measurements required to obtain information to the same degree of accuracy
may differ. This is due to the fact that close to a complex source (i.e.,
one composed of a number of independent subsources), the sound level is a
stronger function of the contribution of the nearest subsource, and thus
additional measurement positions may be required to appropriately include
the contributions from each subsource. The results of recent research[8]
suggests, however, that for compressor noise the number of measurement
positions required is nearly the same and is independent of the distance
from the source (outside a region very close to the source).
-------
2.3. Use of Weighted Sound Levels
a. Limitations of Weighted Sound Levels
A-weighted sound level has been identified "by many researchers as
a simple, useful measurement which correlates with human response to noise
in a number of situations. It has been noted that A-weighted levels show
fair correlation with speech interference and annoyance for a wide variety
of sources, and, in particular, A-weighted levels are one of the better
available descriptors of community response to internal combustion engine
noise. These considerations strongly indicate that A-weighted level is
an appropriate measure for present air compressors. It should be noted,
however, that two significant factors which contribute to this fortuitous
situation are that a) compressors which are currently manufactured do not
radiate dominant discrete tones at mid and high frequencies and b) the
spectra of energy from presently manufactured compressors are "regular"
in the sense that they do not contain large amounts of low frequency
(i.e., below 125 Hz) energy. We believe it is appropriate that the
methodology be a suitable measure of community response not only for
compressors as presently manufactured, but also for compressor designs
which can reasonably be anticipated to result in response to noise emission
regulation.
The lack of discrete tones at mid and high frequencies from present
compressors is attributed to the fact that it is physically difficult (with
the exception of poorly designed compressor air handling systems) to pro-
duce such sounds with present compressors internal combustion engines and
cooling fans with a relatively small number of blades. The only anticipa-
ted mechanism for changing this situation would be the future use by the
industry of other types of engines for driving the compressor, such as the
gas turbine, which may generate strong high frequency tones. This is
considered to be a reasonable possibility in view of the rapidly devel-
oping technology of small gas turbines. In this event, it is conceivable
that a more complex descriptor such as perceived noise level may be
required in order that the measured value will correlate well with
community response.[9]
The other major question which must be addressed is the frequency
spectra of noise from compressors which is to be expected when a internal
combustion engine powered compressor is quieted to meet an A-weighted noise
regulation. It should be noted that the A-weighting puts very little
emphasis on low frequency sounds, and low frequency sounds are not reduced
as much, and may in fact be increased, by noise control treatments which are
designed to control high frequency sound. Thus there are both physical
and economic grounds for expecting low frequency noise from portable air
compressors to at best remain the same and potentially increase under the
influence of an A-weighted noise regulation.
2/
' It was noted earlier in this report that the engine driving the compressor
is the principle noise source on presently manufactured compressors.
-------
Given the possibility that lov frequency noise from compressors
could increase, what might be the resulting response from the community?
One major potential source for community disturbance is the vibration
induced in structures by low frequency sound . Another source of
complaint has been identified by one author[10] as "disturbance or unease"
in the presence of noise which peaks in the 31.5 Hz octave band or lower
frequencies. In particular he identifies "the indadequacy of dB(A) as a
predictor of adverse reaction to sound with most of their energy at low
frequencies," which he attributes to the fact that "dB(A) was originally
designed for use at fairly low intensities and makes no allowance for the
rapid growth of loudness with intensity which occurs in this [low fre-
quency] region." The spectra identified as disturbing exhibited a
difference between C- and A-weighted levels of 22 dB or more, and substant-
ially exceeded the threshold of hearing (^60 dB SPL) in the 31.5 Hz
octave band. A typical range of difference between C- and A-weighted
level for presently manufactured portable air compressors is 5-13 dB.
If we assume that A-weighted noise from compressors is reduced by 10 dB,
and assuming C-weighted noise, which is determined by low frequency com-
ponents of the spectrum, remains unchanged, we find that the C-minus A-
weighted difference will be in the range 15-23 dB, thus creating a potential
problem.
Should the need for reduction of the low frequency noise emission
from a compressor be explicitly identified, then engineers can design to
incorporate low frequency noise control treatments. The principle source of
possible increase in low frequency noise is associated with the large surface
area enclosure (large with respect to the size of the source) which is needed
to contain the noise radiated by the engine and compressor. If the enclosure
is directly connected to the vibrating surface of the compressor, it will
produce a new, far more efficient, path for radiation of low frequency
energy. This problem can be reduced or avoided by Judicious selection
of attachment points for the enclosure.
The proposed methodology incorporates no measure of this low frequency
noise. It is recommended that some appropriate mechanism be created which
permits the detection of significant increases in radiated noise in this
frequency region, so that this may be regulated if it occurs. The Justifi-
cation for this approach is that the lack of current community response
data for this low frequency noise problem makes it extremely difficult to
define a suitable measure, or an appropriate limit, even though we expect,
based on limited experience that A-weighted level alone will not completely
describe the community response to signals containing very high proportions
of low frequency energy. Additional research on community response to low
frequency noise is clearly needed in order to provide a basis for regulating
sources with these characteristics.
Several acoustical consultants (G. W. Kamperman, Kamperman and Assoc.,
R. M. Hoover, Bolt Beranek and Newman, In., F. M. Kessler, Dames and
Moore, Inc.) have observed that sound levels on the order of 75-80
dB in the 31*5 Hz octave band are sufficient to rattle windows and
dishes in wood frame housing.
-------
b. Representative Noise Spectra of Portable Air Compressors
Figures 1-8 depict 1/3-octave band sound pressure level spectra for
rotary vane, rotary screw, and reciprocating compressors with diesel and
gas power plants, in standard and silenced configurations[11]. All
spectra represent either single position or average levels at a seven
metre distance from the surface of the machine in question. The data
indicate, within the limitations of the analysis bandwidth, the lack of
high frequency pure tones from compressors. Of the seven spectra, only
the data for the 750 cfm, silenced, rotary screw compressor exhibits
potential high frequency tones. In this case it was verified that the
signal at 1600 Hz represented a narrow band noise radiated by the muffler
shell, and not a tone per se. The spectra are representative in the sense
that they include a diversity of machine types and manufacturers, but
should not be considered as typical or indicative of the sound levels
produced by any given type or model of compressor, or of the limitations
of noise control treatments of machines. Additional spectral data may be
found in reference 8, which further confirms these generalizations.
The presence of significant, apparently-discrete-frequency energy at
low frequencies is attributed primarily to exhaust discharge at the firing
frequency and its first few harmonics for the case of relatively minimal
exhaust silencing.
An inverse A-weighting curve is superimposed on each spectrum in
such a fashion as to identify the important spectral components for that
weighting. The most important portions of the spectra are those which are
within five to ten decibels of the weighting curve as plotted. It should
also be noted that noise in the low frequency region below 100 Hz is such
that mid frequency noise would have to decrease by 15 dB or more for this
group of compressors before the low frequency noise would significantly
affect the A-weighted sound level, indicating lack of positive response of
the A-weighting in this low frequency region. Alternatively, this low
frequency noise could increase by as much as 15 dB in some cases in the
present situation without affecting A-weighted sound level.
c. Use of A-Weighted Level
One major correlative study on the use of various physical measures
to describe human response to noise has identified the A-weighted sound
level as "...the only measure [of motor vehicle noise] having high correl-
ation with subjective reaction that can also be read directly on a
commercially available meter having standardized performance"[12]. This
conclusion is not at variance with other published studies which have
considered noise produced by vehicles. However there has been little
study of stationery sources. The lack of more general studies of
community response to noise is unfortunate from at least two points of
view. The first difficulty results from the fact that the above mentioned
study was specifically concerned with the driveby situation (both uniform
speed and accelerating) which is a transient event, while compressor noise
is a steady state event. The second cause for concern is the change in
spectrum shape (already discussed in this section) which may result when
an A-weighted criterion is applied. Neither of these questions can be
resolved from presently available information in the open literature.
-------
IANDMO.
03
P-.
O
OJ
«o
joo
So
O.
O
V
3 i
h*
O <
1 ^
-f
a i
[1-
W-i
i
1 .V,
T-
W-l
1
_,
i
_s<
1
25 -t
s.
-
~
40
. 1 . . .
100
-
-r
-2
V-
I
T
00
2
-V
-2
-
4
50
5
-3
-t
'
*
*;
1
li-4
"
I
...
1-
00-
-J
--I
1
h
i
*
g
-«
-i
H
,
~
1
t,
^\-
30-3
if
* «e
t
if
^
i
r
r
00-'
-
i
b
50-1
J
I
25
1000
-1
i
40
f?
-J
1
4
00-2
1
40- a
"I
* J
=te
I5-*
^
»-i
TT
»-<
fU
30- 8
00-!
r
2 5
10000
Frequency, Hz
Figure 1. 85 cfm Standard, Gas Povered (2050 rpm), Rotating Vane
Compressor (7 meter Measurement)[11].
10
-------
03
PL,
O
OJ
5J
o
o
1000
10000
Frequency, Hz
Figure 2. 175 cfm Standard, Gas Powered (2200 rpm), Rotary Screv
Compressor (T meter Measurement)[11].
11
-------
IANONO.
<2o
no
o
C\J
o;
/»0
PQ
H
So
?a
^
111
~^
M
rr
5.
' 11
\
SO (
11
SJ
*
*
3
rr
JO-I
71
T
-r
00-1
J
N
,±
~
Vi-l
! 7:
M,
1
40-2
1 J'
*.
I
00-2
4=
M
n' 4 ' " ' i
100
1
-\
-\
-,
-3
?/
i
.....I
k
p
|
t=
li-4
, r
I
-J
1
i
,
1
P=
oo-s
1 ?l
1
1
,
'
H
4-
X>-6
» »
=t
_.T
. .r=
5
=4
r^._
rr4J*
-4-
-T---
^4
!__!__
3O -800-1
3
-^
-
n^
E
yj-
f
|
1
-
!-> .
=
T -
1
J
b
j
4-
25-1
T
-1
~~
-
+ £
a>~i
i
c
r-
t
oc
i i 2
1000
J.
-2
J.
1
1
SO- 3
»
1
15-^
> I
1
-V-
J
00-5
X
1
u
00-
-------
O
CVJ
0)
o
o
IAND NO.
16 17
18
71 Yl n 24 M U V n V> 30 31
M 37 38 3»
I j j ~
.^t:
V-
^FF
H=rt=t4r±±
«H
t~zh
-1
±3E
p
-4~
i.i::
n3-
^4^
=-tr:
5XIZ43i3=h^-±=li±d=iii
0-JO«»40-100-l^i-140-200
' P-
1'
2JO-3?3-
-------
Jlo
Ho
^ )dO
o
CM
(L)
lAMDKO.
16 17 II 1»
So
o 7o
i
Co
^T-
w^»l
[. II
n 1* H It 77 M 19 30 31 « M M » M Vf M » 40 41 O
±&
I±:
<0SO 4*
Efe
,
100
1000
10000
Frequency, Ez
Figxire 5. 185 cfm, silenced, Gas powered, Rotary Screw Compressor
(7 meter Measurement)[11].
-------
two MO.
16 17 II 19 M 21 n 13 M M U 77 M t? M 31 M M M M M » M M 40 41 4»
(3
(^
o
o
o
H
M6
/OO
°>0
ao
?o
to :
^o
f-
^E
:nt:
^
-I 1._4
-f-4~
:zp:
4_.
4f
t
-1ri-
«
rt
4=t
44:
-V
^rq=t
fc«=±
5
55
osoaeo-ioo-i:.5-i«o-3
HI i1 i I i I''l-
-^
s*
-jr-^-ifa-
| J l.^y' I 1
rtfct3S:
30-400-too-iri
i
100
-r1Iir1
i 'i i i
i
it
200-JI50
315-^00-
T^ r-
J00-«0
i I1
eco-ioo-iM
litti
1000
10000
Frequency, Hz
Figure 6. 330 cfm Silenced, (engine unidentified), Reciprocating
Compressor (7 meter Measurement)[11].
15
-------
a
o
CM
(U
h
17 11 19 *> 21 22 13 M 15 M 17 t» » 30 31 M » M » * » » »,» ** «
100
1000
10000
Frequency, Hz
Figure T. 750 cfm, Silenced, Diesel Powered, Rotary Screw Compressor
(T meter Measurement)[11].
16
-------
We conclude at this point that A-weighted sound level is an adequate
predictor of community noise for "regular spectra" of internal combustion
engine noise, but that some additional measure of low frequency noise
levels may be required in the future in order to adequately protect public
welfare. C-weighted level could provide this latter measure, since it
is also the only single scale which is available on standardized instru-
ments which adequately reflects sound levels in this low frequency region.
The lack of suitable correlation of community response with C-weighted
level alone[12] is the reason that measurement of C-weighted level is not
included at this time. Further research is required in order to quantify
community response to low frequency noise and evaluate the validity of
various objective measures of this noise.
2.U. Discussion of Accuracy and Precision
The proposed methodology is intended to provide a measure of
A-weighted sound power radiated to the far field. The estimated precision
(one standard deviation) of the measurement methodology using eight
measurement positions is +1.7 decibels with respect to A-weighted power of
the same source determined under laboratory conditions utilizing a large
number of measurement positions and laboratory grade equipment[8]. The
principal sources of inaccuracy of the field measurement are attributed
to:
(a) instrument tolerances
(b) sound field sampling error.
The instrument tolerances for American National Standard Sl.U-1971
(Specification for Sound Level Meters) Type 1 precision instruments, as
required by the measurement methodology, are no more than +1 dB for meas-
urement of A-weighted sound level at any given frequency from 50 to UOOO
Hz. This uncertainty, coupled with the +0.5 dB uncertainty for the
calibrator, produced the principal sources of instrument error in the
measurement, as well as the principal source of imprecision when differ-
ent instrumentation systems are used.
If the measurement methodology uses five measurement positions, the over-
all measurement imprecision is expected to increase to about 2.h dB.
The overall estimated standard error (one standard deviation) of
the recommended methodology (i.e., reproducibility of measurement
using different observers, test sites and instrumentation, but sources
with the same sound power output) is estimated (on the basis of
limited experience with similar measurements, assuming normal distributions
of all variables and ignoring truncation error) to be on the order of 2 dB
with 95% confidence[8]. (That is, 68% of the measurements on a given
source are expected to be within +2 dB of each other with a 95$ level
of confidence in the value 2. Similarly 95% of the measurements will
lie within +1» dB.) If the number of sample points for sound pressure
level measurement is reduced to five measurement positions, the estimate
of standard error of the methodology will increase to 2.7 dB (one standard
deviation). (That is to say 68J5 of the measurements will lie within +2.7 dB
with 95% confidence.)
17
-------
3. REVIEW OF METHODOLOGY
In the following paragraphs, we will discuss individual sections of
the methodology in the order of their appearance in Section k, and identify
the rationale behind detailed decisions that have not been discussed under
previous more general headings.
a. Acoustic Environment. Requirements for acoustic environment for
the measurements are derived from presently accepted national and
international measurement standards [2," 3, 5, 6]. The requirements for outdoor
measurements are based on reasonable considerations of contributions from
reflections in the context of the accuracy requirements outlined earlier.
No indoor measurement qualification procedure is provided, since procedures
presently being considered have serious defects, and no new procedure could
be developed within the time constraints of this program. Such a procedure
would be highly desirable, since indoor testing would eliminate weather
restrictions. The outdoor site, used for tests by NBS[8], was checked using
one proposed indoor site qualification procedure[6], which resulted in a
calculated correction of more than 2.5 dB when in fact none was required.
Because of such problems, we believe considerable further study is
required before indoor testing for regulatory purposes with an
environmental correction, can be considered. Since the facilities
required to test indoors without a correction are much more expensive than
is believed to be Justifiable, this matter has not been pursued further.
(The acoustic treatment alone, for one such test facility, suitable to
frequencies as low as say 50 Hz, is estimated to cost in the region of
$250k-500k, which is more than the estimated cost of research needed to
fully clarify the correction problem.)
b. Instrumentation. The instrument specifications are derived from a
minimum required performance to achieve desired accuracy. More sophisticated
or faster systems involving multiple microphone arrays and/or computerized
data reduction may also be implemented within this specification.
c. Source Operation. These requirements are derived from present
standards for these devices, [2, 3] and are necessary to assure a reproducible
measurement.
d. Measurement Procedures. The measurement procedure described here
is adapted, with very minor improvements, from the currently proposed ISO
Draft International Standard[6]. Based on our experimental data[8], this
methodology offers the maximum precision and minimum bias for total measure-
ment effort expended. Reducing the number of measurement positions from
eight (as used here) to five as suggested for some survey class measurements
significantly increases the estimated 95% confidence interval for the
measurements from on the order of +U dB to +5.5 dB (two standard
deviations).
The modifications include small changes in the heights of the lower
microphone positions to bring the microphone height closer to the center of
the area it is intended to sample, thereby reducing the potential of correla-
18
-------
tion betveen the measurement points. The specification of the height of the
upper microphone positions are changed to bring them onto the measurement
surface .
No vertical traverse of the microphone at a measurement position has
been incorporated, since it is believed, based on the experimental data, that
the interference pattern arising from ground reflections is not as significant
at small measurement distances from the source. At large distances from the
source, such traversing is believed to be necessary to maintain the same level
of precision with a given number of measurement positions.
The precision estimates for the methodology are based on the
assumption that the variation in the sound level with position on the
measurement surface will be comparable in magnitude or less than that for
the sample group of compressors. The requirement for additional
measurement positions has thus been added for cases where the variance is
large in order to maintain the precision of the measurement. This
requirement for additional positions is only expected to be significant
for some large compressors, however, none of the seventeen compressors
measured in a previous study [8] exceeded this requirement.
e. Information to be Recorded. The data included as line items in
the tables are believed to be those which are both necessary and
sufficient to insure reproducibility of a given measurement at another
time within the above precision limits. Note that the tables are provided
as a guide only, and are not in any way intended to be specific recording
formats .
f . Calculation Procedures. The calculation procedures described
are straightforward applications of the intended measurement. The
included corrections for barometric pressure and temperature are not
required for measurement on a standard day at altitudes below 1 km, as
defined by the pressure and temperature range in section U.9, but are
included to anticipate the potential variations implicit in application of
the methodology throughout the world. The reference conditions are those
which produce an acoustic impedance in air (pc) equal to 1*00 MKS Rayls +_
10% (^0.5 dB).
The expression used in the equation for computation of measurement
surface area is exact. The following approximate expression, given in a
draft ISO Standard, may be found to be computationally more convenient
than the equation used
= Mab + ac + bc)(a + b + c)
(a + b + c + 2r)
where a, b, c and r are as defined in Table I. The expression gives a
value which is always larger than the true value, by an amount which
varies from 0 at r = 0, up to +15$ (0.6 dB) at large r. The error for r =
1 meter and a, b, c comparable with typical compressor dimensions is on
the order of +&% (0.3 dB).
In view of the number of questions that have been raised concerning
the potential difficulties with calculating sound power level, the follow-
ing example is included.
19
-------
Example computation of sound power, using the proposed methodology (Data
taken from test 9 of reference 11).
A compressor is measured on a surface of one metre distance from the
source.
The physical size of the compressor is 3.66 m long by 1.82 m wide by
2.lU m high. Test conditions are 25°C, 755 mm Hg. The coordinates of the
measurement positions are as follows (r = 1m)
a - (|+ P) - 2.8 m, |= 1.1* m
b - (f + P) - 1.9 n
c = (h + r) = 3.1 m
II.JL = |(a + b -]
i
a, = ^(a + r/2) = 1.7 m (less than
1 d
1 - -
than b)
The measured A-weighted sound level at the eight measurement positions were
found to be (rounded to the nearest half dB):
Position
Source Sound
level (dB)
Background Sound
level
Mean square 2
pressure (pascals)
(from Table IV)
1
79-5
50
.036
2
78.5
51
.028
3
76.5
U8
.018
U
78.5
57
.028
5
79.5
1*9
.036
6
77.5
35
.023
7
76.5
51
.018
8
79-5
51
.036
Since the spread of the source data is less than 9 dB, only eight positions
were required according to the methodology.
The background noise was found to be negligible (more than 10 dB below
the measured levels). The sound levels were converted to mean square pressure
using Table IV. In using Table IV, all the data were noted to be in the range
7!* to 80 dB. The upper portion of Table IV shows that the mean square pressure
should be in the range 0.01 to O.OU (pascals) . The lower (expanded) scale
permits easy, accurate interpolation to find values more precisely.
If equation 1 had been used to calculate the values (very easily
handled with a pocket calculator that includes log and power (x ) functions),
the calculation would go as follows:
I*/
'Equation numbers refer to the equations given in Section U.9
Calculation Procedures.
20
-------
.«i . 107-95-9.U = 10-l.«i
= 0.0351*813389
with similar simple calculations for the remaining values.
Averaging (adding up and dividing "by eight) gives:
p , = >028
*avg n 8
The measurement surface area is calculated from equation (3) as
follows .
S = 2(3.66)(2.ll») + 2(l.82)(2.lU) + (3.66)11.82)
+ ir(l)(2(2.llt) + 3.66 + 1.82) + 2ir(ir
= 67.06 sq. meters.
The sound pover is computed from equation (U). Since temperature and baro-
metric pressure are within the acceptable range, the value of C is unity and
W = 2.5 (1}(67.06)(0.028)
= U.69 milliwatts
The sound power level from equation (5) is
L = 10 Log (U.69 + 90)
= 6.1k + 90
= 97 dB
rounded to the nearest dB. This number is expected to be within +U.O dB of
the true sound power with 95$ confidence. An independent, more accurate
measurement gave L equals 98.7 dB.
If the compressor rating were in terms of average sound level at ten
metres from the center of the source, this could be computed using equa-
tion (6).
L (r) = 97 - 8 - 20 Log 10
P =97-8-20
= 69 dB.
-------
1*. PORTABLE AIR COMPRESSOR MEASUREMENT METHODOLOGY
U.l. Purpose and Applicability
This measurement methodology defines an acceptable procedure by which
the noise from portable air compressors may be measured to ascertain whether
the noise emitted is in compliance with the noise emission standards promul-
gated by the U. S. Environmental Protection Agency pursuant to Section 6 of
the Noise Control Act of 1972 (Public Law 92-571*). The method is applicable
to air compressors of a portable type intended for outdoor use. The measure-
ment method provides for the determination of the A-weighted sound power
output of the device in the free-field over a reflecting plane based on
A-weighted sound level (SPL) measurements determined on a measurement
surface one metre from the surface of the machine.
H.2. References
The following standards are referenced within this methodology.
a. American National Standard Sl.U-1971 "Specifications for
Sound Level Meters".
b. International Electrotechnical Commission Publication 179,
"Precision Sound Level Meters".
c. American National Standard SI.2-1962 "The Physical Measurement
of Sound".
d. American National Standard SI.1-1966 "Specifications for Octave
Half Octave and One-third Octave Filters".
e. American National Standard SI.13-1971 "Methods for the Measure-
ment of Sound Pressure Levels".
U.3. Measurement Uncertainty
The achievable uncertainty for a measurement of A-weighted sound power
level of a portable air compressor according to this methodology using
commercially available instrumentation meeting the requirements of Section
U.5 is estimated to be a precision of 2 dB (one standard deviation with
95% level of confidence) with negligible bias.
U.U. Acoustic Environment
a. General Requirements
The test site shall be such that the compressor radiates sound into a
free field over a reflecting plane. This condition may be considered fulfilled
if the test site consists of an open space free of large reflecting surfaces
within the distances specified below, from any microphone position or equip-
ment location (See U.2(b)). The minimum measurement area shall consist of a
22
-------
circular, flat, hard-surface which extends from the source at least 1 meter
"beyond the most distant measurement point (see Section U.5.2, Microphone
Positions).
b. Criteria for Adequacy of the Environment
1. Reflecting plane. The reflecting plane shall be flat (^0.1 meters),
and of smooth concrete or sealed asphalt, or other hard material. Materials
other than concrete or sealed asphalt shall have normal incidence absorption
coefficient of less than 0.06 over the frequency range 20 Hz to 10 kHz.
2. Outdoor measurement. No large reflecting surface such as a sign-
board, building, hillside, trees, etc. shall be located within 10 meters of
a microphone position or source location.
c. Ambient Noise
It is strongly preferred that the background ambient noise at the test
site shall be more than ten decibels below the A-weighted sound levels of
the unit under test, especially if the noise is fluctuating. In no case shall
a site be used in which the maximum background ambient noise is within four
decibels of the levels to be measured. The background ambient noise shall be
recorded before the start of the test. In the event that the average ambient
background noise is within ten decibels of the measured sound levels, the
ambient background noise after the test shall be recorded, and the smaller of
the before and after levels shall be included in the calculation of sound
power.
i
See also Section k.6.(e.) for ambient noise from sources other than the
compressor during compressor operation.
d. Other Ambient Factors
The temperature and barometric pressure at the time and place of the
test shall be recorded for reference purposes. No measurements shall be
taken when the wind speed exceeds 5 m/s (12 mph). (See paragraph U.5(a)
for windscreen requirements.)
1*. 5 Instrumentation
a. General Requirements
The measurement system used shall conform to the requirements of
American National Standard SI.13 "Methods for the Measurement of Sound
Pressure Levels", Section 5, Field Method, except that the sound level meter
or equivalent instrumentation shall meet the frequency response requirements
of a Type 1 meter as stated in American National Standard Sl.U-1971 "Speci-
fications for Sound Level Meters" over the frequency range 50 Hz - 10,000 Hz.
A windscreen shall be used when the average wind noise is within 20 dB of
the levels to be measured. The windscreen shall not affect the measured
23
-------
A-weighted sound levels from the noise source in excess of +0.5 dB. This
nay be experimentally evaluated by comparing the measured A-weighted sound
level with and without the windscreen, at a position close enough to the
source that wind noise is not a factor.
b . Additional Instrumentation Requirements
a) A sound level calibrator accurate within +0.5 dB.
b) An anemometer or other device for measurement of ambient
wind speed and direction accurate within 10$ at 5 ra/sec.
c) An engine speed indicator, accurate within +2%.
d) An air pressure guage, accurate within
e) A thermometer for measurement of ambient temperature accurate
within +1°C.
f ) A barometer for measurement of ambient pressure accurate within
±1*.
U.6. Installation and Operation of the Source
a. Source Operation
The machine under test shall be operated continuously at design full
speed with the compressor on load, delivering its rated airflow and pressure.
The compressed air discharge should be piped clear of the test area or fed
into an effective silencer. The air discharge line shall be provided with
a throttling device remote from the compressor such that no significant
pressure drop need be maintained at the compressor air distribution valve(s).
The sound level from the compressed air discharge shall be at least 10 dB.
below the machine sound level at all microphone positions. All cooling
air vents in the engine/compressor enclosure shall be full open during all
sound level measurements. Service doors that should be closed during normal
operation (at any and all ambient temperatures) shall be closed during all
sound level measurements.
All sound level measurements reported shall be obtained after the
machine is operating at normal engine temperature.
b- Source Monitoring
The compressor speed and air discharge pressure shall be monitored and
recorded at the beginning and at the completion of the sound level tests .
21*
-------
U.7. Measurement Procedure
a. Microphone Positions
The measurements are made at eight positions on a measurement surface
which is a uniform distance from the reference surface. The reference
surface is the smallest hypothetical rectangular box of dimensions L x W x H
with top surface parallel to the reflecting plane which encloses the source.
The minimum number of measurement positions shall be eight, including one
near the center of each of the four sides of the source and four above the
top of the source near the corners of the measurement surface. The
coordinates of these positions are given in Table I. The measurement
positions are shown in Figure U.I. Note the height for positions 5 to 8 is
determined by the requirement that the distance from the closest point on
the reference surface to the measurement point is equal to the measurement
distance r.
Accuracy of microphone positioning with respect to the reference sur-
face shall be r +_ 0.1 m. Accuracy of microphone positioning on the measure-
ment surface shall be within 0.2 r of the position defined by the three
coordinates of Table I.
The minimum allowable value for r to be used shall be 0.75 m. The
preferred value of r is 1.0 m. The maximum allowable value for r shall be
that value which is consistent with background noise level and test site
dimension requirements given in Section k.h.
The microphone shall be oriented with respect to the measurement
surface, in the manner which it is calibrated for optimum flat frequency
response assuming that the direction of sound field propagation at each
measurement location is perpendicular to the measurement surface at the
measurement position. Observers shall be at least one meter away from the
microphone, and r + 1 meters away from the measurement surface.
The time average A-weighted sound level using the "slow" response mode
of the sound level meter shall be determined to a precision of +0.5 dB at
each measurement location with the compressor operating and the compressor
off.
If the range of the eight sound-level values exceeds 8 dB for the
operating compressor condition, additional sound level data shall be
recorded at eight additional positions. The coordinates of the additional
positions shall be as given in Table IA.
U.8 Information to be Recorded
The maximum steady observed A-weighted sound level using the slow
response characteristic, shall be recorded for each measurement location to
the nearest 0.5 dB. The average A-weighted background noise levels shall
be recorded at all locations after the measurement, if within ten decibels
of the measured value.
If levels are resolved to the nearest 0.1 dB, then levels in the range
x.8 to y.2, shall be recorded as y.O; values in the range x.3 to x.T
shall be recorded as x.5 (y = x + l).
25
-------
Measurement surface
Reference paroilelpiped
Reference porollelpiped
i- O-
-x xf
///////////
I
Measurement surface
"" Y~ in*
O Key measurement position
x Secondary (additional)
measurement position
Figure U.I. Location of measurement positions on the measurement surface.
-------
TABLE I
Microphone Position Coordinates
Source reference surface dimensions are L, W, H.
a = |+r, b=|+r, c = H + r
1^ = |(a + b - |), if 1^ is greater than H then take ^ = H
a. = :j(a + T)» if a., is greater than -z then take a., = L/2
b. = (b +§)> if b is greater than b then take b = b
Position Number ' X
1 a
2 0
3 . -a
k 0
5 al
6 -ai
i *
7 -^
8 a,
Y - Z ! Dist. from
reference surface
o hl
b h]L
o hl
-b h;L
b.. greater than H
b. greater than H
-b, greater than H
-b.. greater than H
r
r
r
r
r
r
r
r
Origin for the coordinate system 'is the point on the ground plane under
the geometric center of the source.
27
-------
Position Number
TABLE IA
Coordinates of Additional Positions
X Y Z
greater than H
i
greater than H
greater than H
greater than H
9
10
11
12
13
lit
15
16
a2
0
-a2
0
&1
'"l
-a
^
al
0
bl
0
-bl
b
*>
-b
1
-b '1
Distance from
reference surface
r
r
r
r
r
r
r
a« =
+ 21), if &- is greater'than a, then take a« = a
28
-------
TABLE II
EXAMPLE DATA FORM FOR MEASUREMENT OF SOUND FROM PORTABLE AIR COMPRESSORS
Manufacturers Test Report No. Date of Test
SUBJECT
Model:
Manufacturer:
Serial No:
Rated Speed & Capacity:
Description:
Dimensions U,w,h)
TEST CONDITIONS:
Manufacturers Test Site Identification
Test Radius (r) for Measurement:
rpm
_rpm
Distance from Observer to Microphone:
Operating Speed as Tested: (Beginning)
(End of test) _
Reflecting Plane Composition:
Ambient Temperature: °C
Ambient Maximum Wind Speed During Sound Level Measurements:
Wind Direction (see Sketch Following Page):
Atmospheric pressure:
Remarks:
.meters
meters
km/h
jnmHg
INSTRUMENTATION:
Microphone:
Sound Level Meter:
Calibrator:
Other:
Ser. No.
Ser. No.
Ser. No.
Ser. No.
SOUND LEVEL DATA
A-weighted Sound Level (dB re 2 x 10 pascal)
Full
Load
Condition
Background
(compressor
off)
Pos 1-8
Pos 9-16
Before
(pos 1-8)
After
(pos 1-8)
Position
1
2
3
U
5
6
7
8
Tested by
Date
29
-------
TABLE II (Continued)
Sketch: Indicate source orientation, wind direction,
Computed Measurement Location Coordinates
a - (L/2 + r) _____ meters observed value of height for
positions 5-8 ___ meters
b = (W/2 + r) ____ meters observed value of height for
positions 13-16 __ meters
c = (H + r) ____ meters
h, = (ika + b - £)) meters (Not greater than H)
JL e. c. -
a. = (p-(a + "p")) ___ meters (Not greater than L/2)
b, = (ir(b + §)) _ meters (Not greater than b)
J. c. d " "~
a« = ('p(a * p^) ) ___ meters (Not greater than a)
30
-------
The additional data identified in Table II shall be recorded.
A calibration history of all equipment Used, meeting the requirements
of American National Standard SI. 13-1971. Section 5 shall be maintained.
1|.9- Calculation Procedures
The average A-weighted sound level on the measurement surface, and the
A-weighted sound power level shall be computed in a manner similar to that
shovn in Table III. Use of the specific format shown is not required.
The steps of the computation of sound power level are as follows.
(l) Determine the corrected sound pressure (or pressure level).
(2) Compute the average mean square pressure for the measurement
surface.
(3) Compute the area of the measurement surface.
(U) Compute the sound power for the measurement surface from the
product of the mean square pressure and the area of the surface.
(5) Compute the sound power level from the logarithm of the sound
power.
The equations which may be used for this calculation are shown below.
(l) Corrected sound pressure.
The corrected mean square sound pressure at a measurement position may
be calculated using the following equation.
r\
where p(i) is the mean square pressure at position i (pascals )
L (i) is the sound level of the source with background noise at
P position i (dB)
L1 (i) is the sound pressure level of the background noise at posi-
P tion i (dB).
(Table IV may be used to convert pressure level to squared pressure.)
If the background noise sound level at a position is more than ten
decibels below the source sound level then the correction will be negligible.
31
-------
(2) Average sound pressure.
The average mean square sound pressure (P) may be calculated from
the equation
P(2) * ... +P(n))
N
(3) Measurement surface area.
The measurement surface area is calculated using the equation:
S = 2 LH + 2 WH + LW + irr (2H + L + W) + 2irr2
where S is the area of the measurement surface (square meters)
L,W,H are the source reference surface length, width and height
respectively (metres)
r is the distance from the source to the measurement surface
(metres )
(H) Sound power.
The sound power is computed from the equation
W = 2.5 x C x S x P
where W is the sound power (milliwatts)
C is the value of - .
pc
where p is the density of air at the time of the test (kg/m ) and c is
the speed of sound in the air at the time of the test (m/s). (See
note below. )
S is the measurement surface area from eq. 3>
P is the average sound pressure from eq. 2.
Note: If the test temperature is in the range -15°C to hO°C and the barometric
pressure at the test site is in the range 680 - 780 mm Hg the value for C shall
be taken as 1.
Outside this range the following expression shall be used for C.
Where B is the barometric pressure (mm Hg)
T is the air temperature (degrees Celsius)
32
-------
(5) Sound power level.
The sound power level is calculated from
Lv = 10 Log1Q W + 90
12
L is the sound power level (dB re 10~ watt)
W is the sound power (mW), from eq. h.
The sound power level shall "be rounded to the nearest dB.
(6) Average sound level.
The average sound level at any distance from the center of the source
over a reflecting plane (with no nearby vertical reflecting surfaces) may be
computed from:
L (R) = LW - 8 - 20 LogR dB
R Is the distance from the center of the source (meters).
33
-------
TABLE III
EXAMPLE FORM FOR SOUND POWER CALCULATION FOR NEAR FIELD MEASUREMENTS
Manufacturers Test No.
Machine Identification
1. Test source dimensions
Test Date
L
W
H
2. Test radius (r)
3. Test surface area S = (2LHp+ 2 WH + LW +
+ irr (2H + L + W) + 2 irr )
1|. Barometric Pressure (fi)
5. Temperature (T)
6. Number of measurement positions n_
7. Average A-weighted Sound Pressure
jn
jn
jn
m
m
mmHg
~
Position
1
2
3
It
5
6
7
8
A-weighted SPL
(dB)
(l) Source
On
(2) Source
Off
A-weighted 2
(Sound Pressure)
(3) Source
On
(10 Source
Off
Corrected ?
(Sound Pressure)
(3) - (1»)
8.
9-
*2
Sum of (Sound Pressure) position 9-1.6 (Enclose tahle)
p
Sum of (Sound Pressure) (sum of tatle plus Line 8)
10. P. = Average Sound Pressure (line 5 divided by n)
A
-------
TABLE IV
CONVERSION OF WEIGHTED SOUND PRESSURE LEVEL TO SQUARED SOUND PRESSURE (pascal)2
2
A solution of the equation: P = 10
avg
' (pascal)
100-
m
.
90-
:
80-
PRESSURE
LEVEL (dB)
70-
U
2
1
- 0.1*
0.2
' 0.1
- .OU
SQUARED
. 02 PRESSUREg
(pascal)
' .01
- U.10~3(.OOU)
60 -li*.10 (.oooi*)
V
-2
Ti
50 U.io"5 (.00001*)
EXPANDED SCALE (TYPICAL FOR ALL 10 dB INTERVALS)
PRESSURE LEVEL
-ij 1 1 H 1
.1* .5 -6 .8 1 1.25 1.5 2 2.5 3 1*
SQUARED PRESSURE
35
-------
5. ACKNOWLEDGEMENT
The author recognizes the significant contributions made "by members
of the NBS Task Group which assisted in the formulation of earlier drafts
of the measurement methodology and indicated some of the significant
problems in need of investigation in the experimental program. Members of
the task group included:
NAME ORGANIZATION
Mr. Harry Baker Worthington CEI, Inc.
Mr. George M. Diehl Ingersoll-Rand Co.
Mr. R. S. Geney Atlas Copco, Inc.
Mr. L. T. Hayes Schramm Inc.
Mr. F. Wm. Heckenkamp Gardner-Denver Co.
Mr. F. M. Kessler Dames & Moore
Mr. Richard Ostwald Gordon Smith Co.
36
-------
6. BIBLIOGRAPHY
[l] Kearney, A. T., Inc., A study to determine the economic impact of
noise emission standards in the construction equipment industry
Draft portable air compressors report, Contract No. 68-01-15^2 (U.S.
Environmental Protection Agency, Washington, D. C., December 1973).
[2] CAGI-PNEUROP Test Code for the Measurement of Sound from Pneumatic
Equipment (Compressed Air and Gas Institute, New York, New York,
1969) (Also American National Standard S5.l).
[3l American National Standard Test Code for the Measurement of Sound
from Pneumatic Equipment, S5-1-1971 (American National Standards
Institute, New York, New York, 1971); International Organization for
Standardization Recommendation for the Measurement of Airborne Noise
Emitted by Corapressor/Primemover Units Intended for Outdoor Use,
R2151-1972 (American National Standards Institute, New York, New
York, 1972).
[U] Work by International Organization for Standardization Technical
Committee U3 (Acoustics)/ Subcommittee 1 (Noise)/Working Group 9
(Pneumatic Equipment).
[5] American National Standard Method for the Physical Measurement of
Sound, Sl.2-1962 (revised 1971) (American National Standards
Institute, New York, New York, 1971).
[6] International Organization for Standardization Draft Engineering
Method for the Determination of Sound Power Levels of Noise Sources
for Free- Field Conditions over a Reflecting Plane, DIS-37UU
(American National Standards Institute, New York, New York).
[7] For example: Hubner, G., Analysis of errors in measuring machine
noise under free-field conditions, J. Acous. Soc. Am. jjMU), 967-977
(1973).
[8] Holmer, C. I., Procedures for Estimating Sound Power from
Measurements of Sound Pressure, to be published as an NBS report,
IR-652/EPA 550/8-76-001.
[9] International Organization for Standardization Recommendation for
Acoustics, Assessment of Noise with Respect to Community Response,
R1996-1971 (American National Standards Institute, New York, New
York, 1971).
[10] Tempest, W., Loudness and annoyance due to low frequency sound,
Acustica 29_, 205-209 (1973).
[ll] Private communication, W. Patterson (Bolt Beranek and Newman, Inc.,
Cambridge, Massachusetts) to C. I. Holmer (National Bureau of
Standards, Washington, D. C.), April 197^. Data were recorded
during the preparation of reference [13]. Also from data supplied
by members of the compressor noise task force.
37
-------
[12] Galloway, W. J., et al., Highway noise Measurement, simulation and
mixed reaction, National Cooperative Highway Research Report No. 78
(NAS-NRC Highway Research Board, Washington, D. C., 1969).
[13] Patterson, W.N., Ely, R., and Huggins, G., Portable air compressor
noise measurement, Report No. 2795a (U. S. Environmental Protection
agency, Washington, D.C., March 1971*).
38
-------
APPENDIX A
PROCEDURE FOR PREDICTING A-WEIGHTED SOUND LEVEL FROM A-WEIGHTED SOUND POWER
A.I. Introduction
This appendix is intended as a guide for users to estimate the A-weighted
sound level to be expected from a portable air compressor in practical
situations under rated operating conditions. The following applications
are considered: (l) estimating the A-weighted sound level at various dis-
tances from a portable air compressor operating in a relatively open area,
(2) estimating the A-weighted sound level at various distances from a
portable air compressor operating in a confined space, such as in an alley
or on a street between tall buildings, and (3) estimating the effective
A-weighted sound level reduction due to introducing a "noise shield" or
barrier wall between the portable air compressor and the observer. This
appendix is only concerned with the relative A-weighted sound level contribu-
tion from one specific source. To determine the total A-weighted sound level
from multiple sound sources, refer to American National Standard SI. 2-1962.
A. II. Sound Level Decrease With Distance
The following procedure is to be used for estimating the A-weighted
sound level from the geometric center of a portable air compressor, operating
under rated conditions more than ten meters (33 feet) away. For this situa-
tion it is assumed that the area including both the portable air compressor
and the observer is relatively free of large sound reflecting surfaces (except
the ground), such as buildings or berms, and there is a direct line of sight
from the observer to the air compressor.
From the A-weighted sound power level of the source (A), determine A _ dB
the corresponding A-weighted sound level for the reference distance -28 dB
of 10 meters (B), by subtracting 28 dB from (A). B. _ dB
Measure the direct distance from the geometric center of the air
compressor to the observer (in meters) Cj _ m
Using Nomogram No. 1 read the corresponding sound level correc-
tion D: __ dB
Subtract the value found in Step D from that found in Step B. Xj ___ dB
The final result above assumes no excess sound attenuation due to atmos-
pheric conditions which may affect the actual A-weighted sound level at
distances greater than about 100 meters (330 feet).
39
-------
NOMOGRAM NO. 1
A-WEIGHTED SOUND LEVEL DECREASE WITH DISTANCE
C: Distance From
Center of Air
Compressor in
Meters
C D
1000-j
_
500-
mm
200-
100-
_
,
50--
30-
20-
10-
-40
-39
-38
-37
-36
-35
-34
-33
-32
-31
-30
-29
-28
27
-26
-25
-24
-23
-22
-21
-20
-19
-18
-17
-16
-15
-14
-13
-12
-11
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
J-b
D: A-weighted Sound
Level Seduction
(as)
Uo
-------
NOMOGRAM NO. 2
APPROXIMATE SOUND LEVEL ATTENUATION USING A BARRIER
A'
P = PATH LENGTH DIFFERENCE) = A1 + B1 - C (Meters)
P. PATH LENGTH
DIFFERENCE (METERS)
(MULTIPLY BY 3.3
TO OBTAIN P IN FEET)
0
.031
.1 J
.2 -
.3 -
5 -
.8 .
1.0 -
2.0 '
i».o -
6.0
5
- 7
9
11
. 13
. 15
. 17
' 19
" 21
. 23
^ 25
A-WEIGHTED SOUND
LEVEL DECREASE DUE
TO BARRIER ATTENUATION
(dB).
PRACTICAL LIMIT OF
BARRIER ATTENUATION
-------
A.m. Sound Level Reduction From Barriers
The following procedure can be used for estimating the minimum
attenuation due to a barrier, or the height of a barrier to be constructed, or
the distance the portable air compressor or observer must ba from a barrier to
achieve a specified A-weighted sound level reduction. (The A-weighted
spectrum of the operating compressor is assumed to have maximum energy near or
above 500 Hz for this noise reduction estimating procedure.) The length of
the barrier is assumed to be at least two times the barrier height or two
times the largest compressor (source) dimension, whichever is greater.
A. To determine the attenuation for a given barrier height and compressor
location, we must determine the effective path length for sound propagation
over the barrier.
The effective height of the barrier is the height of the top of the
barrier above the line of sight from the top of the compressor to the
observer. The barrier changes the sound level by blocking the sound and
forcing the sound to "bend around" the top of the barrier. The amount of
bending is measured by the "path length difference" (P) between the direct
path (if there was no barrier) and the distance along the shortest path over
the top of the barrier. If the compressor and the observer have approximately
the same elevation, then a rough estimate of the path length difference is
calculated by:
Measuring the height of the barrier above the top of the compressor H m
Determining the longer of the distance from the barrier to the
compressor or barrier to the observer R m
Calculate P = H x H/R P m
If the compressor and observer are on significantly different elevations,
then the simplest method is to make a scale sketch of a section showing the
top of the compressor, top of the barrier and the observer, and measuring the
dimensions A1, B1, and C indicated in the nomogram. P is simply computed from
these dimensions.
With the value of P, the nomogram is entered to find the A-weighted level
difference (G). The sound level with the barrier is then
A-weighted sound level at observer = X - G = dB
Note that the barrier provides some attenuation even if it Just meets the
line of sight (but this attenuation reduces rapidly toward zero within a short
distance above the barrier).
B. To determine the height of a barrier to be constructed to achieve a given
A-weighted sound level reduction:
Measure the compressor to barrier distance in meters or
the observer to barrier distance and enter whichever is
larger m
-------
Note the desired A-weighted reduction dB
Note the path length difference indicated for the desired
level reduction P m
Enter the line of s:' p,ht distance C m
Estimate the effective height of the barrier Heff = C x P/2 Heff m
Estimate the elevation of the line of site (H.. ) above the
ground at the proposed barrier site. (Use scale sEetch if
necessary) H m
The total height of the barrier is H ff + H. = m
This is the total barrier height necessary to achieve the desired
"A-weighted reduction" assuming no other large sound reflecting surfaces in
the vicinity of the compressor.
The value of H ff is the approximate maximum height required if the
barrier is midway between the compressor and observer. To check the result,
perform the calculation described in A above.
The above procedure for predicting noise reduction effectiveness of a
barrier placed near a compressor assumes that both the compressor and the
observer are in a relatively open, level area. In many practical situations
this is not the case and the effectiveness of the barrier will be somewhat
reduced because of sound reflecting off of other surfaces. Also, if the A-
weighted sound level reaching the observer before installation of a barrier is
determined by low frequencies (such as the fundamental firing rate of the
engine) due to an inadequate or poorly maintained compressor engine muffler
the barrier will provide less A-weighted reduction.
A.IV. Summary
This appendix is a simplified application guide for estimating noise
level from sound power rated portable air compressors. The guide can prove
very useful for determining whether the sound level of the compressor is in
compliance with local or state noise control regulations. Alternately, the
guide may be used to determine what techniques are available to the user of a
portable air compressor for reducing the compressor sound level exposure at
specific locations of interest.
USCOMM-NBS-DC
-------
NBS.I14A (Rev 7.731
U.S. De.Pt. OF COMM.
BIBLIOGRAPHIC DATA
SHEET
I. I'UHI.IC AT ION OK KI-.l'OKl NO.
NBSIR 75-653
2. Ciov'i Acii'ssiun
No.
3. RiL-ipiiTii's Ac <.(.' i on NIL
EPA 550/8-76-002
4. TIILl'! AND SUIO ITU.
Measurement Methodology and Supporting Documentation for
Portable Air Compressor Noise
5. Publication Date
Jan. 1975
6. Performing Or^m/'JIion ( ode
7. AUfHOR(S)
Curtis I. Holmer
Ropor, No.
9. PERFORMING ORGANIZATION NAMI-. AND ADDRESS
NATIONAL BUREAU OF STANDARDS
DEPARTMENT OF COMMERCE
WASHINGTON, D.C. 20234
10. Proifit/Tmk/Work Unit No.
2130152
11. Contrail/Grant No.
12. Sponsoring Organization Name* and Complete AJdrest (Street, City, Stale, ZIP)
Joint NBS and Environmental Protection Agency (EPA)
Office of Noise Abatement and Control
Washington, D. C. 20460
13. Type of Report & Period
_. Cohered
Final report
Feb.1974 -Dec. 1974
14. Sponsoring Agency Codi
15. SUPPl.LMENTARY NOTES
16. ABSTRACT (A 200-word or less tactual summary of most significant information If document includes a significant
bibliography or literature survey, mention it here ) This report presents recommendations and supporting
rationale on a measurement methodology for portable air compressors. The methodology
provides for the determination of A-weighted sound power level or the equivalent weightec
sound pressure level at a reference distance. A-weighted level is used because of its
:orrelation with community response to noise from internal combustion engine noise. It
Ls recommended, however, that the spectra associated with the regulated source be moni-
tored in some manner to insure that the spectra remain similar to those for which
^-weighted sound level retains good correlation with community response. The methodology
jses weighted sound level measurements at eight positions on a curved surface surround-
Lng the source at a distance of one metre from the surface of the machine. Data recorded
it these positions are used to calculate the average weighted sound pressure level of the
nachine on the measurement surface. This is combined with the area of the measurement
surface. This is combined with the area of the measurement surface to give the sound
>ower level of the machine. From this value, a rating sound pressure at a rating dis-
:ance may be calculated by subtracting a constant value. Procedures which permit the
rapid estimation of A-weighted sound level are included. These are applicable for
istimation of A-weighted sound level in a variety of circumstances when the sound power
>r equivalent sound pressure level at a reference distance is known.
17. KEY WORDS (six to twelve entries, alphabetical order, capitalize only the first letter of the first key word unless a proper
name; separated by semicolons)
Acoustics; air compressor; internal combustion engine; noise; sound power level;
sound pressure level.
18. AVAILABILITY [£ Unlimited
I ' For Official Distribution. Do Not Release to NTIS
I I Order From Sup. of Doe., U.S. Government Printing Office
Washington, D.C. 20-102, SO Cat. No. CIS
l Order From National Technical Information Service (NTIS)
Springfield, Virginia 22151
19. SECURITY CLASS
(THIS REPORT)
UNCLASSIFIED
20. SECURI1 Y CLASS
(THIS PAGE)
UNCLASSIFIED
21. NO. OF PAGES
47
22. Price
$3-75
USCOMM.OC 2B042.P74
------- |