An Introduction to Noise Control

in Buildings

Course No: C02-016

Credit: 2 PDH

J. Paul Guyer, P.E., R.A., Fellow ASCE, Fellow AEI

Continuing Education and Development, Inc.

22 Stonewall Court
Woodcliff Lake, NJ 07677

P: (877) 322-5800
[email protected]
 
An Introduction  
to  
Noise Control  
In Buildings 

Guyer Partners J. Paul Guyer, P.E., R.A.

44240 Clubhouse Drive Paul Guyer is a registered civil engineer,
El Macero, CA 95618 mechanical engineer, fire protection engineer,
(530)7758-6637 and architect with over 35 years experience in
[email protected] the design of buildings and related
infrastructure. For an additional 9 years he
was a senior-level advisor to the California
Legislature. He is a graduate of Stanford
University and has held numerous national,
state and local positions with the American
Society of Civil Engineers and National Society
of Professional Engineers.

© J. Paul Guyer 2009 1

This course is adapted from the Unified Facilities Criteria of the United States government,
which is in the public domain, has unlimited distribution and is not copyrighted.

© J. Paul Guyer 2009 2

1. INTRODUCTION

This is an introduction to noise control in buildings. It is not an in-depth treatment, but it

will introduce designers to some important principles and terminology. In simple
applications on real projects the information provided here will give designers a good
start in addressing acoustic control issues. For more acoustically complex projects
designers will need to apply more detailed principles. There are excellent acoustical
engineering treatises available commercially and from government agencies to provide
practitioners with the necessary guidance in applying these more rigorous methods.

2. NOISE CRITERIA

2.1 General. This section includes data and discussions on generally acceptable
indoor noise criteria for acceptable living and working environments. These criteria can
be used to evaluate the suitability of existing indoor spaces and spaces under design.

2.2 Noise Criteria In Buildings. Room Criteria (RC) and Noise Criteria (NC) are two
widely recognized criteria used in the evaluation of the suitability of intrusive mechanical
equipment noise into indoor occupied spaces. The Speech Interference Level (SIL) is
used to evaluate the adverse effects of noise on speech communication.
2.2.1 Noise Criterion (NC) Curves. Figure 1 presents the NC curves. NC curves have
been used to set or evaluate suitable indoor sound levels resulting from the operation of
building mechanical equipment. These curves give sound pressure levels (SPLs) as a
function of the octave frequency bands. The lowest NC curves define noise levels that
are quiet enough for resting and sleeping, while the upper NC curves define rather noisy
work areas where even speech communication becomes difficult and restricted. The
curves within this total range may be used to set desired noise level goals for almost all
normal indoor functional areas. In a strict interpretation, the sound levels of the
mechanical equipment or ventilation system under design should be equal to or be
lower than the selected NC target curve in all octave bands in order to meet the design
goal. In practice, however, an NC condition may be considered met if the sound levels

© J. Paul Guyer 2009 3

in no more than one or two octave bands do not exceed the NC curve by more than one
or two decibels.
2.2.2 Room Criterion (RC) Curves. Figure 2 presents the Room Criterion (RC)
curves. RC curves, like NC curves, are currently being used to set or evaluate indoor
sound levels resulting from the operation of mechanical equipment. The RC curves
differ from the NC curves in three important respects. First, the low frequency range has
been extended to include the 16 and 31.5 Hz octave bands. Secondly, the high
frequency range at 2,000 and 4,000 Hz is significantly less permissive, and the 8,000
Hz octave band has been omitted since most mechanical equipment produces very little
noise in this frequency region. And thirdly, the range over which the curves are defined
is limited from RC 25 to RC 50 because; 1) applications below RC 25 are special
purpose and expert consultation should be sought and; 2) spaces above RC 50 indicate
little concern for the quality of the background sound and the NC curves become more
applicable.

Table 1 lists representative applications of the NC curves. The evaluation of the RC

curves is different than that for the NC curves. In general the sound levels in the octave
bands from 250 to 2,000 Hz are lower than those of the NC curves. Should the octave
band sound levels below 250 Hz be greater than the criteria a potential “rumble”
problem is indicated. As a check on the relative rumble potential, the following
procedure is recommended:

• Sum the sound pressure levels in the octave bands from 31.5 through 250 Hz on
an energy basis.
• Sum the sound pressure levels in the octave bands from 500 through 4,000 Hz
on an energy basis.
• Subtract the high frequency sum (step 2) from the low frequency sum (step 1).
• If the difference is +30 dB or greater, a positive subjective rating of rumble is
expected, if the difference is between +25 and +30 dB a subjective rating of
rumble is possible, if the difference is less than +20 dB a subjective rating of
rumble is unlikely.

© J. Paul Guyer 2009 4

Also indicated on the RC curves (Figure 2) are two regions where low frequency sound,
with the octave band levels indicated, can induce feelable vibration or audible rattling in
light weight structures.
2.2.3 Speech interference levels. The speech interference level (SIL) of a noise is the
arithmetic average of the SPLs of the noise in the 500-, 1000-, and 2000-Hz octave
bands. The approximate conditions of speech communication between a speaker and
listener can be estimated from Table 2 when the SIL of the interfering noise is known.
Table 2 provides “barely acceptable” speech intelligibility, which implies that a few
words or syllables will not be understood but that the general sense of the discussion
will be conveyed or that the listener will ask for a repetition of portions missed.

Figure 1
Noise Criterion (NC) Curves

© J. Paul Guyer 2009 5

 Figure 2
Room Criterion (RC) Curves

© J. Paul Guyer 2009 6

 Table 1
Representative Applications of the NC Curves

Table 2
Speech Interference Levels

© J. Paul Guyer 2009 7

3. SOUND DISTRIBUTION INDOORS

3.1 Sound Pressure level In a Room. The sound pressure levels at a given distance
or the sound power levels for individual equipment items can often be obtained from
equipment suppliers. Once the characteristics of the sound source have been
determined, then the sound level at any location within an enclosed space can be
estimated. In an outdoor “free field” (no reflecting surfaces except the ground), the
sound pressure level (SPL) decreases at a rate of 6 dB for each doubling of distance
from the source. In an indoor situation, however, all the enclosing surfaces of a room
confine the sound energy so that they cannot spread out indefinitely and become
dissipated with distance. As sound waves bounce around within the room, there is a
build-up of sound level because the sound energy is “trapped” inside the room and
escapes slowly.
3.1.1 Effect of distance and absorption. The reduction of sound pressure level
indoors, as one moves across the room away from the sound source, is dependent on
the surface areas of the room, the amount of sound absorption material on those areas,
the distances to those areas, and the distance from the source. All of this is expressed
quantitatively by the curves of Figure 3. Figure 3 offers a means of estimating the
amount of SPL reduction for a piece of mechanical equipment (or any other type of
sound source in a room, as one moves away from some relatively close-in distance to
any other distance in the room, provided the sound absorptive properties of the room
(Room Constant) is known. Conversely Figure 3 also provides a means of estimating
the sound reduction in a room, from a given source, if the distance is constant and the
amount of absorptive treatment is increased. Table 3 represents a simplification of
Figure 3 for a special condition of distance and room constant.

© J. Paul Guyer 2009 8

 Figure 3

1. Reduction of SPL in dB in going from normalized 3-feet distance and 800 sf Room
Constant to any other distance and Room Constant.
2. Negative value of reduction means an increase in sound level.

Table 3

© J. Paul Guyer 2009 9

3.1.2. Sound absorption coefficients. For most surfaces and materials, the sound
absorption coefficients vary with frequency; hence the Room Constant must be
calculated for all frequencies of interest. Even room surfaces that are not normally
considered absorptive have small amounts of absorption. Usually sound absorption
coefficients are not measured in the 31, 63 and 8,000 Hz frequencies. Where the data
at these frequencies are not available use 40% of the value of the 125 Hz for the 31 Hz
band, 70% of the 125 Hz value for the 63 Hz band and 80% of the 4,000 value for the
8,000 Hz octave band. Values of sound absorption coefficients for specialized
acoustical materials must be obtained from the manufacturer.
3.1.3 Estimation of Room Constant. In the early stages of a design, some of the
details of a room may not be finally determined, yet it may be necessary to proceed with
certain portions of the design. An approximation of the Room Constant can be made
using Figure 4 and Table 4. The basic room dimensions are required but it is not
necessary to have made all the decisions on side wall, floor, and ceiling materials. This
simplification yields a less accurate estimate than does the more detailed procedure,
but it permits rapid estimates of the Room Constant with gross, but non-specific,
changes in room materials and sound absorption applications. Then, when a favored
condition is found, detailed calculations can be made with an equation that is not
presented here.
3.1.4 Use of Figure 4. Figure 4 gives a broad relationship between the volume of a
typically shaped room and the Room Constant as a function of the percentage of room
area that is covered by sound absorption material. Room area means the total interior
surface area of floor, ceiling, and all side walls. The Room Constant values obtained
from this chart strictly apply at 1000 Hz, but in this simplified procedure are considered
applicable for the 2000- through 8000-Hz bands as well.

© J. Paul Guyer 2009 10

 Table 4

© J. Paul Guyer 2009 11

 Figure 4

© J. Paul Guyer 2009 12

4. SOUND ISOLATION BETWEEN ROOMS

Discussed here are data and procedures for estimating the changes in sound levels as
one follows the “energy flow” path from a sound source to a receiver, through building
components, such as walls, floors, doors etc. First, the sound pressure levels in the
room containing the source drop off as one moves away from the source. Then, at the
walls of the room, some sound is absorbed, some is reflected back into the room, and
some is transmitted by the walls into the adjoining rooms (this also occurs at the floor
and ceiling surfaces). The combined effects of this absorption, reflection, and
transmission are the subject of this discussion.

4.1 Sound Transmission Loss (TL), Noise Reduction (NR) And Sound
Transmission Class (STC). With the knowledge of the acoustical isolation provided by
walls and floors, it is possible to select materials and designs to limit noise intrusion
from adjacent mechanical equipment rooms to acceptable levels. The degree of sound
that is transmitted is influenced by the noise isolation properties of the demising
construction, the area of the demising wall, floor or ceiling and the acoustical properties
in the quiet room.
4.1.1 Transmission loss (TL) of walls. The TL of a wall is the ratio, expressed in
decibels, of the sound intensity transmitted through the wall to the airborne sound
intensity incident upon the wall. Thus, the TL of a wall is a performance characteristic
that is entirely a function of the wall weight, material and construction, and its numerical
value is not influenced by the acoustic environment on either side of the wall or the area
of the wall. Procedures for determining transmission loss in the laboratory are given in
ASTM E 90. This is the data usually given in most manufacturers literature and in
acoustic handbooks. Laboratory ratings are rarely achieved in field installations.
Transmission loss values in the laboratory are usually greater, by 4 to 5 dB, than that
which can be realized in the field even when good construction practices are observed.
ASTM E 336 is a corresponding standard method for determination of sound isolation in
buildings (in situ). There are many references that provide transmission loss

© J. Paul Guyer 2009 13

performance for building materials. In addition many manufacturers also provide
transmission loss for their products.
4.1.2 “Noise reduction” (NR) of a wall. When sound is transmitted from one room
(the “source room”) to an adjoining room (the “receiving room”), it is the transmitted
sound power that is of interest. The transmission loss of a wall is a performance
characteristic of the wall structure, but the total sound power transmitted by the wall is
also a function of its area (e.g. the larger the area, the more the transmitted sound
power). The Room Constant of the receiving room also influences the Sound Power
Loss (SPL) in the receiving room. A large Room Constant reduces the reverberant
sound level in the room at an appropriate distance from the wall. Thus, three factors
influence the SPL in a receiving room: the TL of the wall, the area of the common wall
between the source and receiving rooms, and the Room Constant R2 of the receiving
room. These three factors are combined in a manner that is beyond the scope of this
presentation.
4.1.3 “Sound transmission class” (STC). Current architectural acoustics literature
refers to the term “Sound Transmission Class” (STC). This is a one number weighting of
transmission losses at many frequencies. The STC rating is used to rate partitions,
doors, windows, and other acoustic dividers in terms of their relative ability to provide
privacy against intrusion of speech or similar type sounds. This one-number rating
system is heavily weighted in the 500- to 2000-Hz frequency region. Its use is not
recommended for mechanical equipment noise, whose principal intruding frequencies
are lower than the 500- to 2000 Hz region. However, manufacturers who quote STC
ratings should have the 1/3 octave band TL data from which the STC values were
derived, so it is possible to request the TL data when these types of partitions
are being considered for isolation of mechanical equipment noise. The procedure for
determining an STC rating is given in ASTM standard E 413.

4-3. Transmission Loss-Walls, Doors, Windows. Generally a partition will have

better noise reduction with increasing frequency. It is therefore important to check the
noise reduction at certain frequencies when dealing with low frequency, rumble

© J. Paul Guyer 2009 14

type noise. Note that partitions can consist of a combination of walls, glass and doors.
Walls can generally be classified as fixed walls of drywall or masonry, or as operable
walls.
4.3.1 Drywall walls. These walls consist of drywall, studs and, sometimes, fibrous
blankets within the stud cavity.
4.3.1.1 Drywall. Drywall is a lightweight, low-cost material, and can provide a very high
STC when used correctly. The use of Type X, or fire-rated drywall of the same nonrated
drywall thickness, will have a negligible effect on acoustical ratings. Drywall is generally
poor at low frequency noise reduction and is also very susceptible to poor installation.
Drywall partitions must be thoroughly caulked with a non-hardening acoustical caulk at
the edges. Tape and spackle is an acceptable seal at the ceiling and side walls.
Electrical boxes, phone boxes, and other penetrations should not be back-to-back, but
be staggered at least 2 feet, covered with a fibrous blanket, and caulked. Multiple
layers of drywall should be staggered. Wood stud construction has poor noise reduction
characteristics because the wood stud conducts vibration from one side to the other.
This can be easily remedied by using a metal resilient channel which is inserted
between the wood stud and drywall on one side. Non-load-bearing metal studs are
sufficiently resilient and do not improve with a resilient channel. Load-bearing metal
studs are stiff and can be improved with resilient channels installed on one side.
4.3.1.2 Fibrous blankets. Fibrous blankets in the stud cavity can substantially improve
a wall’s performance by as much as 10 dB in the mid and high frequency range where
non-load-bearing metal studs, or studs with resilient channels, are used. A minimum 2
inch thick, 3/4 lb/ft3 fibrous blanket should be used. Blankets up to 6 inches thick
provide a modest additional improvement.
4.3.1.3 Double or staggered stud walls. When a high degree of noise reduction is
needed, such as between a conference room and mechanical room, use double or
staggered stud wall construction with two rows of metal or wood studs without bracing
them together, two layers of drywall on both sides, and a 6 inch thick fibrous blanket.

© J. Paul Guyer 2009 15

4.3.2 Masonry walls. Masonry construction is heavy, durable, and can provide
particularly good low frequency noise reduction. Concrete masonry units (CMU) made
of shale or cinder have good noise reduction properties when they are approximately
50 percent hollow and not less than medium weight aggregate. Parging or furring with
drywall on at least one side substantially improves the noise reduction at higher
frequencies. The thicker the block, the better the noise reduction. An 8 inch thick, semi-
hollow medium aggregate block wall with furring and drywall on one side is excellent
around machine rooms, trash chutes, and elevator shafts.
4.3.3 Doors. The sound transmission loss of both hollow and solid core doors will
substantially increase when properly gasketed. Regular thermal type tape-on gaskets
may not seal well because of door warpage, and can also cause difficulty in closing the
door. Tube type seals fitted into an aluminum extrusion can be installed on the door
stop and fitted to the door shape. Screw type adjustable tube seals are available for
critical installations. Sills with a half moon seal at the bottom of the door are
recommended in place of drop seals, which generally do not seal well. Two gasketed
doors with a vestibule are recommended for high noise isolation. Special acoustical
doors with their own jambs and door seals are available when a vestibule is not
practical or very high noise isolation is required.
4.3.4 Windows. Fixed windows will be close to their laboratory TL rating. Operable
sash windows can be 10 dB less than the lab rating due to sound leaks at the window
frame. Gaskets are necessary for a proper seal. Some window units will have
unit TL ratings which would be a rating of both the gasketing and glass type. Double-
glazed units are no better than single-glazed if the air space is 1/2 inch or thinner. A 2-
inch airspace between glass panes will provide better noise reduction. Laminated glass
has superior noise reduction capabilities. Installing glass in a neoprene “U” channel
and installing sound absorbing material on the jamb between the panes will also
improve noise reduction. Special acoustical window units are available for critical
installations.
4.3.5 Transmission loss values for building partitions. Table 5 through Table 14
provide octave band transmission losses for various constructions, comments or details
on each structure are given in the footnotes of the tables. STC ratings are useful for

© J. Paul Guyer 2009 16

cursory analysis when speech transmission is of concern. The octave band
transmission losses should be used when a more thorough analysis is required, such as
when the concern is for mechanical equipment.

Table 5
Transmission Loss (in dB) of Dense Poured Concrete or Solid Core Concrete Block or Masonry

© J. Paul Guyer 2009 17

 Table 6
Transmission Loss (in dB) of Hollow-Core Dense Concrete Block or Masonry

© J. Paul Guyer 2009 18

 Table 7
Transmission Loss (in dB) of Cinder Block or Other Lightweight Porous Block Material with
Impervious Skin on both Sides to Seal Pores

© J. Paul Guyer 2009 19

 Table 8
Transmission Loss (in dB) of Dense Plaster

© J. Paul Guyer 2009 20

 Table 9
Transmission Loss (in dB) of Stud-Type Partitions

© J. Paul Guyer 2009 21

 Table 10
Transmission Loss (in dB) of Plywood, Lumber and Simple Wood Doors

© J. Paul Guyer 2009 22

 Table 11
Transmission Loss (in dB) of Glass Walls or Windows

© J. Paul Guyer 2009 23

 Table 12
Transmission Loss (in dB) of Typical Double-Glass Windows,
Using 1/4-in.-Thick Glass Panels with Different Air Space Widths

© J. Paul Guyer 2009 24

 Table 13
Transmission Loss (in dB) of a Filled Metal Panel Partition
and Several Commercially Available Acoustic Doors.

© J. Paul Guyer 2009 25

 Table 14
Transmission Loss (in dB) of Aluminum, Steel and Lead

© J. Paul Guyer 2009 26